UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Neurobiological correlates of overweight and obesity in people with bipolar disorder Bond, David Joseph 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2013_fall_bond_david.pdf [ 2.37MB ]
Metadata
JSON: 24-1.0103356.json
JSON-LD: 24-1.0103356-ld.json
RDF/XML (Pretty): 24-1.0103356-rdf.xml
RDF/JSON: 24-1.0103356-rdf.json
Turtle: 24-1.0103356-turtle.txt
N-Triples: 24-1.0103356-rdf-ntriples.txt
Original Record: 24-1.0103356-source.json
Full Text
24-1.0103356-fulltext.txt
Citation
24-1.0103356.ris

Full Text

  NEUROBIOLOGICAL CORRELATES OF OVERWEIGHT AND OBESITY IN PEOPLE WITH BIPOLAR DISORDER  by  David Joseph Bond  BSc, Memorial University of Newfoundland, 1992 MD, Memorial University of Newfoundland, 1996   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIRMENTS  FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  The Faculty of Graduate and Postdoctoral Studies  (Neuroscience)   THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)  October 2013  ? David Joseph Bond, 2013 ii  Abstract  Up to 75% of people with bipolar disorder (BD) are overweight or obese, and these patients suffer more severe psychiatric symptoms than normal-weight patients, including more frequent depressions, more suicide attempts, lower response rates to pharmacotherapy, and greater cognitive impairment. Obesity is a chronic inflammatory condition that damages numerous body organs and is causally linked to the development of diabetes, heart disease, and cancer. BD is fundamentally a brain illness, and this, along with converging evidence from human and animal studies suggesting that the brain is a target organ for obesity-related damage, compelled me to investigate obesity-related neurobiological changes early in BD.    I found that at recovery from their first manic episode, there was no difference between BD patients and age- and gender-matched healthy subjects in mean body mass index (BMI) or rates of overweight or obesity. Nonetheless, magnetic resonance imaging (MRI) demonstrated that overweight/obese patients had reduced white matter and temporal lobe volumes compared to normal-weight patients. WM reductions are characteristic of early-stage BD, while temporal lobe reductions are frequently reported later in the illness. These findings thus suggested a testable hypothesis: that the neuropathology of BD is exacerbated with elevated BMI. Subsequent investigations supported this hypothesis. A voxel-based analysis of regional brain volumes revealed that BMI-related volume reductions primarily affected frontal, temporal, and subcortical emotion-generating and ?regulating brain areas implicated in BD. Moreover, MR spectroscopy showed that overweight/obese patients had reduced hippocampal N-acetylaspartate concentrations compared to normal-weight patients. Similar findings were not detected in overweight/obese healthy subjects, who had reduced occipital lobe grey matter volume and no neurochemical alterations.   These are the first data to establish a relationship between elevated BMI and neurobiological alterations in BD, or any psychiatric illness. They demonstrate that elevated BMI is associated with unique brain changes early in BD that negatively impact regions believed to be vulnerable in the illness. This immediately suggests an explanation for the more severe illness course experienced by obese BD patients, and creates a compelling argument for examining the iii  neurobiological impact of obesity in other mental illnesses with high obesity rates, such as major depressive disorder and schizophrenia. iv  Preface  This thesis summarizes research I carried out during my doctoral studies at the University of British Columbia Department of Psychiatry from Sept 2007 ? August 2013, under the supervision of Dr. L. N. Yatham, Professor of Psychiatry at UBC. The insight to investigate neurobiological changes associated with obesity in bipolar disorder, as a possible explanation for the more severe psychiatric illness course experienced by obese as compared to normal-weight patients, was mine. I tested my hypotheses using data from the Systematic Treatment Optimization Program for Early Mania (STOP-EM), a prospective clinical and research program for people recently diagnosed with bipolar disorder, which was initiated by Dr. Yatham in 2004 and led by him since then. STOP-EM and the procedures described were approved by the UBC Clinical Ethics Research Board, Certificate Number H04-70169.  I have written a number of manuscripts based on the data and conclusions presented here. For all manuscripts, I formulated the scientific question and hypothesis, designed the research approach, conducted the statistical analysis, wrote the first draft of the paper, and revised the paper based on my co-authors? comments. Several of the manuscripts have been published in or submitted to peer-reviewed journals, or are in preparation for submission, including:  A version of Chapter 2 was published as: Bond DJ, Kauer-Sant?Anna M, Lam RW, Yatham LN. Weight gain, obesity, and metabolic indices following a first manic episode: prospective 12-month data from the Systematic Treatment Optimization Program for Early Mania (STOP-EM). J Affect Disord 124:108-117.  A version of Chapter 3 was published as: Bond DJ, Lam RW, Yatham LN. 2010. The association of weight gain with mood symptoms and functional outcomes following a first manic episode: Prospective 12-month data from the Systematic Treatment Optimization Program for Early Mania (STOP-EM). Bipolar Disord 12: 616-626.    v  A version of Chapter 4 was published as: Bond DJ, Lang DJ, Noronha MM, Kunz M, Torres IJ, Su W, Honer WG, Lam RW, Yatham LN. 2011. The association of elevated body mass index with reduced brain volumes in first-episode mania. Biol Psychiatry 15:381-7.  A version of Chapter 5 is in press as: Bond DJ, Ha TH, Lang DJ, Su W, Torres IJ, Honer WG, Lam RW, Yatham LN. Body mass index-related regional gray and white matter volume reductions in first-episode mania patients. Biol Psychiatry, in press.  A version of Chapter 6 is currently in the final stages of preparation for submission as: Bond DJ, da Silveira LE, Macmillan EL, Gigante AD, Lang DJ, Su W, Torres IJ, Honer WG, Lam RW, Yatham LN. Body mass index x diagnosis interaction for hippocampal glutamate/ glutamine and N-acetylaspartate in remitted first-episode mania patients. In preparation.  I also wish to acknowledge the following for their additional contributions: ? Dr. D. J. Lang (Radiology) for generously sharing her knowledge about MR imaging and data analysis, and for allowing me to include data from her healthy subjects in my analyses in Chapters 4 and 5. ? Dr. T. H. Ha (Department of Psychiatry, Seoul National University) and Dr. M. Gonzalez (Radiology) for their assistance with the theory and practice of voxel-based MR data analysis.  ? Mr. W. Su (Psychiatry) for extracting the region-of-interest structural brain data used in the analyses in Chapter 4, and assisting me with assigning anatomical labels to areas of reduced white matter volume in Chapter 5.  vi  Table of contents Abstract ........................................................................................................................................................ ii Preface ......................................................................................................................................................... iv Table of contents ........................................................................................................................................ vi List of tables................................................................................................................................................. x List of figures ............................................................................................................................................. xii List of abbreviations ................................................................................................................................ xiv Acknowledgements .................................................................................................................................. xix Dedication .................................................................................................................................................. xx 1.  Introduction and literature review ....................................................................................................... 1 1.1  Prevalence and burden of bipolar disorder......................................................................................... 1 1.2  Prevalence of obesity in BD............................................................................................................... 1 1.3  Health consequences of obesity in BD .............................................................................................. 4 1.4  Obesity and the brain ......................................................................................................................... 6 1.5  Psychiatric consequences of obesity in BD ....................................................................................... 7 1.6  The Systematic Treatment Optimization Program for Early Mania (STOP-EM) ............................. 8 1.7  Thesis overview ............................................................................................................................... 10 2.  BMI at recovery from the first manic episode in patients with BD, and weight gain during the initial 12 months of maintenance treatment ........................................................................................... 13 2.1  Introduction ...................................................................................................................................... 13 2.2  Methods............................................................................................................................................ 15 2.2.1  STOP-EM ................................................................................................................................. 15 2.2.2  Magnetic resonance imaging (MRI) and proton MR spectroscopy (1H-MRS) ......................... 17 2.2.3  Assessment of weight, BMI, and metabolic indices ................................................................. 17 2.2.4  Data analysis and statistics ........................................................................................................ 18 2.3  Results .............................................................................................................................................. 20 2.3.1  Demographic and clinical data .................................................................................................. 20 vii  2.3.2  Weight gain, overweight, obesity, and metabolic indices ......................................................... 20 2.4  Discussion ........................................................................................................................................ 24 3.  The association of weight gain with mood symptoms and impaired psychosocial functioning in BD patients during the initial 12 months of maintenance treatment after a first manic episode ...... 35 3.1  Introduction ...................................................................................................................................... 35 3.2   Methods........................................................................................................................................... 36 3.2.1  STOP-EM ................................................................................................................................. 36 3.2.2  Assessment of time spent with mood symptoms and relapse rates ........................................... 37 3.2.3  Assessment of weight and BMI ................................................................................................ 38 3.2.4  Data analysis and statistics ........................................................................................................ 38 3.3  Results .............................................................................................................................................. 40 3.3.1  Sociodemographic and clinical characteristics ......................................................................... 40 3.3.2  Baseline and 12-month weight and CSWG .............................................................................. 41 3.3.3  Pre-manic clinical outcomes ..................................................................................................... 41 3.3.4  Prospective 12-month clinical outcomes .................................................................................. 42 3.3.5  Functional outcomes ................................................................................................................. 43 3.4  Discussion ........................................................................................................................................ 44 4.  The association of overweight/obesity with reduced white matter volume and temporal lobe volume at recovery from a first manic episode ...................................................................................... 58 4.1  Introduction ...................................................................................................................................... 58 4.2  Methods............................................................................................................................................ 59 4.2.1  STOP-EM ................................................................................................................................. 59 4.2.2  MRI protocol and data extraction ............................................................................................. 60 4.2.3  Data analysis and statistics ........................................................................................................ 60 4.3  Results .............................................................................................................................................. 62 4.3.1  Patients with BD and healthy subjects ...................................................................................... 62 4.3.2  BMI and brain volumes............................................................................................................. 63 viii  4.4  Discussion ........................................................................................................................................ 64 5.  The association of overweight/obesity with reduced regional GM and WM volumes in frontal, temporal, and subcortical limbic brain structures at recovery from a first manic episode ............... 80 5.1  Introduction ...................................................................................................................................... 80 5.2  Methods............................................................................................................................................ 81 5.2.1  STOP-EM ................................................................................................................................. 81 5.2.2  MRI protocol and data extraction ............................................................................................. 81 5.2.3  Image preprocessing and analysis ............................................................................................. 81 5.2.4  Statistical methods .................................................................................................................... 82 5.3  Results .............................................................................................................................................. 83 5.3.1  Patients with BD and healthy subjects ...................................................................................... 83 5.3.2  BMI and regional GM and WM volumes ................................................................................. 83 5.4  Discussion ........................................................................................................................................ 85 6.  BMI ? diagnosis interaction in hippocampal N-acetylaspartate at recovery from a first manic episode ...................................................................................................................................................... 103 6.1  Introduction .................................................................................................................................... 103 6.2  Methods.......................................................................................................................................... 104 6.2.1  STOP-EM ............................................................................................................................... 104 6.2.2  MRI protocol and data extraction ........................................................................................... 104 6.2.3  Acquisition and processing of MRS signals ........................................................................... 105 6.2.4  Data analysis and statistics ...................................................................................................... 105 6.3  Results ............................................................................................................................................ 107 6.3.1  Patients with BD and healthy subjects .................................................................................... 107 6.3.2  The relationship between BMI and hippocampal NAA .......................................................... 107 6.4  Discussion ...................................................................................................................................... 108 7.  Summary and discussion ................................................................................................................... 118 7.1  Summary of research findings ....................................................................................................... 118 ix  7.2  Clinical implications ...................................................................................................................... 119 7.2.1  Weight gain ............................................................................................................................. 119 7.2.2  Mood symptoms and psychosocial functioning ...................................................................... 121 7.3  Implications for understanding the neurobiology of BD ............................................................... 123 7.4  Limitations ..................................................................................................................................... 131 7.5  Relevance and future research directions ....................................................................................... 132 7.5.1  Mediators in the relationship between BMI and brain structure and function ........................ 133 7.5.2  Randomized clinical trial of a weight loss intervention in people with BD............................ 136 References ................................................................................................................................................ 140  x  List of tables  Table 2.1: Baseline sociodemographic characteristics of 47 BD patients and 24 healthy subjects enrolled in STOP-EM ............................................................................................................................................... 28 Table 2.2: Baseline clinical characteristics of BD patients ......................................................................... 29 Table 2.3: Medication use by BD patients at baseline, 6 months, and 12 months ...................................... 30 Table 2.4: Mean BMI and rates of overweight and obesity at baseline, 6 months, and 12 months in BD patients and healthy subjects ....................................................................................................................... 32 Table 3.1: Baseline sociodemographic characteristics of BD patients with and without CSWG from baseline to 12 months .................................................................................................................................. 48 Table 3.2: Baseline clinical characteristics of BD patients with and without 12-month CSWG ................ 49 Table 3.3: Medication use over 12 months in BD patients with and without 12-month CSWG ................ 51 Table 3.4: Pre-manic illness characteristics in BD patients with and without baseline overweight     obesity ......................................................................................................................................................... 52 Table 4.1: Baseline sociodemographic and clinical characteristics of BD patients and healthy subjects taking part in a region-of-interest MRI study ............................................................................................. 69 Table 4.2: Associations between baseline BMI and a) normalized and b) absolute brain volumes, detected with linear regression analyses, in BD patients and healthy subjects ......................................................... 71 Table 4.3: Mean a) normalized and b) absolute brain volumes (mL) in overweight/obese and normal-weight BD patients and healthy subjects .................................................................................................... 77 Table 5.1: Baseline sociodemographic and clinical characteristics of BD patients and healthy subjects taking part in a VBM MRI study ................................................................................................................ 88 Table 5.2: Baseline BMI and rates of overweight and obesity in BD patients and healthy subjects in a VBM MRI study ......................................................................................................................................... 90 Table 5.3: BMI-related GMV and WMV reductions in BD patients and healthy subjects, detected in linear regression models (p < .005 uncorrected, spatial extent threshold ? 500 voxels) ...................................... 93 xi  Table 5.4: The effect of diagnosis and diagnosis-BMI interaction on GMV and WMV, detected with factorial models (p < .005 uncorrected, spatial extent threshold ? 500 voxels) ......................................... 97 Table 5.5: WMV reductions in overweight/obese BD patients compared to overweight/obese healthy subjects (p < .005 uncorrected, spatial extent threshold ? 500 voxels) .................................................... 101 Table 6.1: Sociodemographic and clinical characteristics of BD patients and healthy subjects in a single-voxel hippocampal MRS study ................................................................................................................. 113 xii  List of figures Figure 1.1: Prevalence of obesity in studies comparing BD patients and healthy subjects ........................ 11 Figure 1.2: Prevalence of MetS in studies comparing BD patients and healthy subjects ........................... 12 Figure 2.1: a) Mean weight gain (kg) and b) proportions with CSWG over 12 months in 47 BD patients and 24 healthy subjects ............................................................................................................................... 33 Figure 2.2: Weight changes over 12 months a) BD patients and b) healthy subjects (as a percentage of baseline weight) .......................................................................................................................................... 34 Figure 3.1 Mean number of days with mood symptoms during 12 months of maintenance treatment in a) BD patients with baseline normal weight vs overweight/obesity, b) patients with 12-month normal weight vs overweight/obesity, and c) patients with and without 12-month CSWG ............................................... 53 Figure 3.2 : Mean 12-month MSIF scores in BD patients with baseline normal weight vs overweight/ obesity a) measuring functioning over the previous 30 days, and b) measuring functioning over the previous 6 months ....................................................................................................................................... 55 Figure 3.3 Mean 12-month MSIF scores in BD patients with 12-month normal weight vs overweight or obesity a) measuring functioning over the previous 30 days, and b) measuring functioning over the previous 6 months ....................................................................................................................................... 56 Figure 3.4 Mean 12-month MSIF scores in BD patients with and without 12-month CSWG a) measuring functioning over the previous 30 days, and b) measuring functioning over the previous 6 months ........... 57 Figure 4.1: Segmentation of MRI image into a) grey matter, white matter, and cerebrospinal fluid (CSF) and b) frontal, parietal, occipital, and temporal lobes, using FSL 4.1 ........................................................ 68 Figure 4.2: Relationship between BMI and a) normalized TBV in BD patients, b) normalized TBV in healthy subjects, c) normalized GMV in patients, d) normalized GMV in healthy subjects, e) normalized WMV in patients, f) normalized WMV in healthy subjects, g) normalized temporal lobe volume in patients, and h) normalized temporal lobe volume in healthy subjects ...................................................... 73 Figure 4.3: Mean normalized grey and white matter volumes in a) patients with BD and b) healthy subjects ........................................................................................................................................................ 78 xiii  Figure 4.4: Mean normalized frontal, parietal, occipital, and temporal lobe volumes in a) patients with BD and b) healthy subjects ................................................................................................................................ 79 Figure 5.1: BMI-related brain volume reductions (p < .005 uncorrected, spatial extent threshold > 500 voxels), demonstrated on T1-weighted sections and glass brains, in a) GM in BD patients, b) WM in BD patients, and c) GM in healthy subjects ...................................................................................................... 91 Figure 5.2: Effect of diagnosis on a) GMV and b) WMV (p < .005 uncorrected, spatial extent threshold > 500 voxels). c) Effect of BMI-diagnosis interaction on white matter volume ............................................ 95 Figure 5.3: WMV reductions in overweight/obese BD patients compared to overweight/ obese healthy subjects (p < .005 uncorrected, spatial extent threshold > 500 voxels) .................................................... 100 Figure 6.1: Sample MRS spectra from a) a BD patient and b) a healthy subject ..................................... 112 Figure 6.2: Mean hippocampal NAA in overweight/obese and normal-weight BD patients and healthy subjects ...................................................................................................................................................... 116 Figure 6.3: Correlation between BMI and mean hippocampal NAA in BD patients and healthy       subjects ...................................................................................................................................................... 117 xiv  List of abbreviations  1H-MRS  Proton Magnetic Resonance Spectroscopy 3T   3 Tesla ADHD   Attention Deficit-Hyperactivity Disorder AGRP   Agouti-Related Peptide ANCOVA  Analysis of Covariance ANOVA  Analysis of Variance BBB   Blood-Brain Barrier BD   Bipolar I Disorder BDNF   Brain-Derived Neurotrophic Factor BET   Brain Extraction Tool BMI   Body Mass Index BPRS   Brief Psychiatric Rating Scale CANMAT  Canadian Network for Mood and Anxiety Treatments CART   Cocaine and Amphetamine Regulated Transcript CGI-BP  Clinical Global Impression Scale for Bipolar Disorder CSF   Cerebrospinal Fluid CSWG   Clinically Significant Weight Gain DARTEL Diffeomorphic Anatomical Registration Through Exponential Lie Algebra df Degrees of Freedom   xv  DSM-IV-TR  Diagnostic and Statistical Manual of Mental Disorders, 4th Edition - Text     Revision DTI   Diffusion Tensor Imaging FAST 4  FMRIB's Automated Segmentation Tool FEM   First-Episode Mania FEP   First-Episode Psychosis FLIRT   FMRIB's Linear Image Registration Tool fMRI   Functional Magnetic Resonance Imaging FMRIB  Functional Magnetic Resonance Imaging of the Brain FSL   FMRIB Software Library FTO   Fat Mass and Obesity-Associated Gene FWHM  Full-Width at Half-Maximum FWE   Family-Wise Error  GAF   Global Assessment of Functioning GM   Grey Matter GMV   Grey Matter Volume HAM-D-29  Hamilton Rating Scale for Depression, 29-Item Version HDL   High-Density Lipoprotein HR   Hazard Ratio HS   Healthy Subjects  IBM   International Business Machines xvi  IL   Interleukin IL-6   Interleukin-6 IL-10   Interleukin-10 IU   Institutional Units JHU   Johns Hopkins University  LCModel  Linear Combination of Models LDL   Low-Density Lipoprotein LOCF   Last Observation Carried Forward MADRS  Montgomery-Asberg Rating Scale for Depression MANCOVA  Multiple Analysis of Covariance  MDD   Major Depressive Disorder MetS   Metabolic Syndrome MINI   Mini International Neuropsychiatric Interview MNI   Montreal Neurologic Institute MR   Magnetic Resonance MRI   Magnetic Resonance Imaging MRS   Magnetic Resonance Spectroscopy MSIF   Multidimensional Scale of Independent Functioning NAA   N-acetylaspartate NCEP   National Cholesterol Education Program nGMV   Normalized Grey Matter Volume xvii  NIMH   National Institutes of Mental Health NPY   Neuropeptide Y NS   Not Significant nWMV  Normalized White Matter Volume PANSS  Positive and Negative Syndrome Scale PASW   Predictive Analytics Software PET   Positron Emission Tomography POMC   Pro-Opiomelanocortin PRESS  Point Resolved Spectroscopy SD   Standard Deviation SES   Socioeconomic Status SGA   Second Generation Antipsychotic SMR   Standardized Mortality Rate SPGR   Spoiled Gradient Recalled  SPSS   Statistical Package for the Social Sciences SPM8   Statistical Parametric Mapping, Version 8 STOP-EM  Systematic Treatment Optimization Program for Early Mania T1   Longitudinal Relaxation Time T2   Transverse Relaxation Time TBV   Total Brain Volume TE   Time to Echo xviii  TNF-?   Tumour Necrosis Factor-Alpha TR   Time to Repetition T/R   Transmit/Receive TSH   Thyroid Stimulating Hormone UBC   University of British Columbia VBM   Voxel-Based Morphometry WFU    Wake Forest University WM   White Matter WMV   White Matter Volume YMRS   Young Mania Rating Scale xix  Acknowledgements  I wish to thank my PhD supervisor, Dr. L. N. Yatham. His guidance and support early in my tenure at the UBC Mood Disorders Centre inspired me to pursue a research career in the field of bipolar disorder, and to investigate the neurobiological basis of the illness. His generosity in sharing his data, his breadth of knowledge, and his extensive contacts with other investigators in the field helped make my research possible.  I would also like to acknowledge the other members of my PhD Committee ? Dr. W. G. Honer (Psychiatry), Dr. A. J. Stoessl (Neurology), and Dr. A. L. MacKay (Radiology) - for their time, patience, and thoughtful advice during my doctoral studies.  I would additionally like to recognize the contributions of my fellow students, colleagues, and co-authors at the UBC Mood Disorders Centre. My conversations with them have broadened my understanding of bipolar illness, and their suggestions improved the veracity, clarity, and brevity of my manuscripts.  Finally, and most importantly, I would like to thank everyone who volunteered to take part in psychiatric research though the UBC Systematic Treatment Optimization Program for Early Mania (STOP-EM).  They frequently did so to their own detriment, taking time away from their jobs, education, and families, with no reward other than aiding in the scientific understanding of bipolar disorder and to improving treatments for future patients. I hope that my work has helped to further both of those goals.    xx  Dedication  For my wife, Melissa Maria Noronha Her unconditional love, unfailing support, and unflinching critiques made this work far better than it would have otherwise been  and   For my parents, Barbara Ann Gibbons and the late Aubrey Courtney Bond  Who started me on the road to lifelong learning,  and supported me every step of the way1  1.  Introduction and literature review  1.1 Prevalence and burden of bipolar disorder  Bipolar I disorder (BD) is a chronic medical illness in which patients experience recurrent episodes of depression and mania. The lifetime prevalence of BD in Canada is 2.4%, meaning that over 750,000 Canadians are affected (1). Depressions often last for months or years, and are characterized by unrelenting sadness; reduced desire to take part in, and diminished ability to enjoy, normally rewarding activities; perturbations in biological functions such as sleep and appetite; impaired cognition, including reduced memory, concentration, and decision-making abilities; profound feelings of hopelessness; and, in severe cases, even suicidal thoughts and acts. Manias are typically shorter, usually lasting 2-4 months, but are no less impairing, as the abnormally euphoric mood, dramatically increased energy and activity levels, impulsivity, and unrealistically grandiose thought processes that typify mania frequently lead to erratic and even dangerous behaviours.   BD has an early age of onset, most commonly in the late teens or early 20s, and even in treated samples, patients spend approximately half of their lives with mood symptoms of fluctuating severity (2,3). Not surprisingly, then, people living with BD suffer substantial long-term functional impairment (4), and in fact in a ranking of all psychiatric and medical conditions, the World Health Organization reported that BD is the 6th leading cause of disability worldwide in the 15-44 year age group (5). The direct and indirect costs from BD in the United States have been estimated at $71.9 billion annually (6), and the per capita Canadian costs are likely to be similar, making BD a significant Canadian public health problem.  1.2  Prevalence of obesity in BD  While manic and depressive episodes are the diagnostic hallmarks of BD, a great deal of the symptom burden, disability, and medical costs attributable to the illness are actually caused or exacerbated by high rates of obesity in BD patients. Obesity is endemic throughout the entire western world: between 1980 and 2010, the proportion of the North American population with 2  obesity doubled, and 60% of Canadians and 68% of Americans are now overweight or obese (7,8). Moreover, it is rapidly becoming a major problem in the developing world, particularly in countries with growing middle class populations such as China, India, and Brazil (9). In keeping with this, the World Health Organization estimated that in 2008, 1 billion adults were overweight and 500 million were obese, and predicted that these numbers would increase to 1.6 billion and 700 million, respectively, by 2015 (10). Obesity causes a chronic low-grade systemic inflammatory response which damages many body organs, and it is unambiguously associated with increased morbidity and mortality from numerous medical conditions, including hypertension, type II diabetes, ischemic heart disease, stroke, and several cancers (11). The enormous health consequences of obesity are starkly illustrated by the fact that it is the fifth leading cause of premature death globally, and that for the first time in human history, mortality due to obesity is now greater than that resulting from malnutrition (10).  As profound a problem as obesity is in the general population, it is an even greater problem for people with psychiatric disorders. Obesity disproportionately affects people with serious mental illnesses, including major depressive disorder, schizophrenia, and BD (12-14), and BD, in fact, has among the highest obesity rates of all psychiatric illnesses (15). Numerous studies in North America, Europe, and Australasia have reported that people with BD have higher obesity rates than unaffected individuals (Figure 1.1). Consistent with general population trends, North American patients were especially likely to be obese, with prevalence rates ranging from 22%-53%, compared to 9%-32% in people without BD (14,16-21). European and Australasian patients had modestly lower obesity rates of 19%-41%, but these were nonetheless elevated compared to non-psychiatric comparator groups from the same countries (12%-32%) (22-26). Across studies, obesity rates in North American patients were 43%-465% greater, and those in European/Australasian patients 14%-66% greater, than in the respective general populations. To date, only one negative study was reported, which found that BD patients and comparators were approximately equally likely to be obese (27% vs 30%) (27). Finally, it is especially noteworthy that the single study to examine obesity in pediatric BD patients (mean age=13 years) found them to have an obesity rate 29% greater than children without the illness (28).   3  Additional reports examining the related phenomenon of metabolic syndrome (MetS) produced very similar results. MetS is diagnosed in people who have central (abdominal) obesity and metabolic complications thereof, including hypertension, hypertriglyceridemia, hyperglycemia, and low high-density lipoprotein (HDL) cholesterol (29). Like obesity, it has been strongly linked with increased rates of cardiovascular and other medical illnesses (30-32). Studies on four continents have reported that people with BD were more likely than those in the general population to have MetS (18,20,22,25,33-41) (Figure 1.2), with the rates increased by 10%-108% in North America and 49%-1206% in other countries. The single Canadian study reported a greater-than two-fold increase in MetS in BD patients (34). The only study that did not find this relationship reported that BD patients and comparators were approximately equally likely to have MetS (41% vs 45%) (42). It is worth noting that the rate of MetS in healthy subjects in the negative study was unusually high, approximately double that of most of the other reports cited here.  The pervasive comorbidity of obesity with BD suggests that a propensity for weight gain is inherent to the illness. This argument is supported by an Italian study of 76 never-medicated BD patients, who nonetheless had a rate of overweight/obesity which was 4 times greater than a group of 65 untreated obsessive-compulsive disorder patients, and 3 times greater than the age-matched Italian general population (43). Moreover, even first-degree relatives of BD patients have been shown to have a mean body mass index (BMI, defined as weight [kg] / height [m]2) that is approximately one point greater than healthy controls, and to have significantly elevated rates of obesity-related metabolic complications, such as low HDL cholesterol (44). In addition to illness factors, pharmacologic treatment also clearly plays an important etiologic role in weight gain. Clinical trials have demonstrated the obesigenic effects of almost all medications routinely used to treat BD, including the mood stabilizers lithium and divalproex, the second-generation antipsychotics risperidone, olanzapine, quetiapine, and aripiprazole, and many antidepressants (45). In fact, the necessity of treating and preventing both mania and depression, as well as managing psychotic symptoms and comorbid psychiatric conditions such as anxiety disorders, means that most BD patients in clinical practice are treated with multiple medications. Unfortunately, the degree of weight gain increases with the number of obesigenic medications patients are exposed to (21,26). Thus, excessive weight gain is the rule rather than the exception 4  in BD, and both illness and treatment factors make preventing it a daunting challenge. This creates a major dilemma for patients and clinicians. BD is a highly recurrent illness, and relapse rates in untreated patients are unacceptably high, approaching 90% (46). However, as the sections below will demonstrate, treating patients who are prone to weight gain with obesigenic medications increases their long-term morbidity and mortality from serious medical illnesses. Furthermore, converging evidence suggests that the substantial fraction of patients who experience weight gain also have a more severe psychiatric illness than normal-weight patients, with more time spent in mood episodes, poorer treatment responses, and greater disability.  1.3  Health consequences of obesity in BD  Until recently, adipose tissue was believed to be an inert energy storage depot. This view changed irrevocably with the discovery in the 1990s that adipocytes produce the satiety hormone leptin (47), and adipose tissue is now understood to be key endocrine organ which plays a critical role in regulating metabolism and immune functioning. It does so by producing a plethora of biologically active compounds - over 50 have been identified - including adipokines such as leptin and adiponectin, and cytokines such as tumor necrosis factor-alpha (TNF-?) and various interleukins (ILs) (48,49). Most adipokines and cytokines have inflammatory or anti-inflammatory properties, and the ratio of inflammatory to anti-inflammatory molecules is normally kept finely balanced. Obesity, however, is a pathological state which causes profound alterations in the composition and physiology of adipose tissue. As adipose tissue mass expands, it becomes infiltrated with activated macrophages, leading to increased production of pro-inflammatory molecules such as leptin, resistin, TNF-?, and IL-6, and reduced synthesis of anti-inflammatory molecules, such as adiponectin and IL-10 (50-54). This in turn causes further downstream effects, including an increased production of prothrombotic factors such as plasminogen activator inhibitor I, and reactive oxygen species such as peroxynitrite (55,56). The chronic pro-inflammatory, hypercoagulable state created by obesity culminates in endothelial damage and organ malfunction (57-60), and has been causally linked to the development of obesity-related medical complications such as type II diabetes (61,62).  5  The damage caused by obesity to multiple body systems is dramatically illustrated by the high rates of serious medical conditions that accompany it. Numerous studies have examined this relationship, and several meta-analyses provide the most definitive assessments of the increased risks. One pooled analysis of 89 studies found a ?dose-response? relationship between BMI and hypertension, type II diabetes, ischemic heart disease, stroke, and various cancers (especially colon, kidney, and pancreas), such that the prevalence of these conditions was increased in overweight people (hazard ratios [HRs]=1.15-3.92) and further increased in obese people (HRs=1.49-12.42) (63). Two large meta-analyses, including a combined total of 2.4 million participants, reported similar dose-response relationships between BMI and all-cause mortality over 8-10 years of follow-up (64,65). The HRs for early mortality were 1.03 for mild overweight (BMI=25-27.4), 1.11 for moderate overweight (BMI=27.9-29.9), 1.25 for class I obesity (BMI=30-34.9), 1.59 for class II obesity (BMI=35-39.9), and 1.99 for class III obesity (BMI=40-49.9) (65).   As expected from their high obesity rates, people with BD have an even greater incidence of obesity-related health consequences than the general population. BD patients are approximately 25% more likely than people without BD to have hypertension (22,66,67), 60%-100% more likely to have elevated serum triglycerides (22,40,66), and 200%-500% more likely to have type II diabetes (68-71). Much of this increased medical burden directly attributable to obesity, as illustrated by a US population-based study (n=43,093) which compared the health status of obese and normal-weight BD patients, and reported that obese patients had 35%-96% greater rates of hypertension, arteriosclerosis, and myocardial infarction (21). A seminal European study, which prospectively followed 158 BD patients for over 30 years, reported that compared to people without BD, their standardized mortality rates (SMR, defined as observed/expected deaths) were significantly increased for obesity-related medical conditions, including cardiovascular disease (SMR=1.84), cerebrovascular disease (SMR=1.37), and, in untreated patients, cancer (SMR=1.39). Subsequent reports have confirmed that age-adjusted mortality rates from cardiovascular and cerebrovascular disease are increased by 50%-100% in people with BD (72,73). In fact, the average lifespan of a person with BD is 10-12 years shorter than a person without the illness, and the greatest cause of excess mortality is the increased incidence and earlier onset of cardiovascular disease (73-75). 6  1.4  Obesity and the brain  The brain is one of the most extensively-vascularized organs, and one of the most susceptible to inflammatory, hypoxic, and other insults. Thus, it is puzzling that the impact of obesity on the brain has until recently received little attention. In the past decade, however, converging evidence from numerous studies has demonstrated that the brain is a target organ for obesity-related damage, and that obesity has serious mental health consequences. A number of lines of evidence support this:   1) In general population samples, neuroimaging and neurocognitive studies show clear obesity-related decrements in brain structure and function. Consistently, otherwise-healthy overweight/obese people have smaller total brain volumes, and particularly smaller grey matter volumes (GMV), than people with normal weight, a finding which was recently demonstrated in children as young as five (76-80). Prospective studies further demonstrate that obese individuals experience greater brain volume loss over time than those with normal weight (81). Obesity-related impairments in executive functioning (82-84), processing speed (85,86), and verbal memory (85,87) are also detectable throughout the lifespan. These cognitive deficits are at least partly reversible with voluntary weight loss, suggesting a causal relationship between weight and brain functioning (88).   2) Epidemiologic studies show that obesity is a risk factor for developing serious neuropsychiatric illnesses. A recent meta-analysis of 8 prospective studies (n=55,387) reported that overweight individuals were 27% more likely to develop depression than people with normal weight, while obese people were 55% more likely (89). A second meta-analysis (n=25,624) demonstrated that people who were overweight or obese in mid-life had 35% and 104% increased risks, respectively, for subsequently developing Alzheimer?s disease (90). Similarly increased risks have been found for multiple sclerosis (91,92) and Parkinson`s disease (93,94).   3) Most compellingly, animal studies, which permit random assignment to obesigenic conditions, have shown that experimentally-induced weight gain causes brain volume reductions (95,96) and 7  cognitive impairments (97,98) in mice and primates, confirming a causal relationship between weight gain and brain structure and function.  1.5  Psychiatric consequences of obesity in BD  In keeping with the deleterious brain effects of obesity, a substantial body of evidence shows that obese BD patients experience more severe psychiatric symptoms than normal-weight patients. This is particularly true for the depressive phase of the illness. On average, BD patients spend half of their lives with mood symptoms, and two-thirds of that time is with depression (3,99). Compared to normal-weight patients, obese patients experience a greater number of depressive episodes over the course of their illnesses (16,17), are more likely to receive medication for depression (21), and are more likely to be hospitalized for depression (21,34). Serious consequences of depression, including suicidal ideas and attempts (14,18,21,27) are also more common in obese patients. While the bulk of these data come from studies that retrospectively ascertained mood episodes and suicidality, one 2-year prospective study (n=175) also reported that obese patients had shorter periods of wellness and more frequent depressive relapses (17). Obesity may also negatively impact the manic phase of BD. A number of retrospective studies report that obese patients experience greater numbers of manic episodes over their illnesses (14,17), although not all studies have replicated this finding (16,21), and it was not confirmed in the 2-year prospective study (17). Nonetheless, medication and hospitalization for mania are more common in obese patients (21), suggesting that the severity of manic episodes may be increased, even if the frequency is not. Finally, obese patients have lower response rates to pharmacotherapy for BD, with one clinical trial reporting that the response rate to lithium decreased by 7% for every 1-point increase in BMI above a BMI of 22 (100), and a second showing that obese patients were only half as likely as non-obese to respond to ziprasidone (101).   Even during periods of relative wellness, obese BD patients fare less well than their normal-weight counterparts. They are more likely to have persisting cognitive impairment between mood episodes, including reduced processing speed and verbal fluency, than normal weight patients (102). They are less also likely to achieve full functional recoveries between mood 8  episodes. People with BD frequently suffer prolonged social, occupational, and relationship impairment (4), and functional recovery is often delayed or absent even when patients are in full remission, with only 25%-39% of remitted patients returning to their premorbid functioning in the year following the resolution of a mood episode (103-106). In the general population, obesity has been strongly associated with impaired functioning (107,108), and recent studies comparing functional outcomes in overweight/obese and normal-weight BD patients also consistently show that obese patients have significantly greater impairment across numerous functional domains (21,109,110). Patients themselves identify weight gain as the most problematic aspect of their treatment, and perceive it to have a highly detrimental impact on their quality of life (111).  Thus, in addition to causing increased morbidity and mortality from serious medical illnesses, overweight/obesity puts BD patients at risk for more frequent and severe depressions, serious suicidal thoughts and acts, refractoriness to pharmacologic treatments, persisting cognitive impairment, and greater disability. While psychological and social consequences of obesity, such as poor self-esteem and the well-known stigma directed toward obese individuals (reviewed in chapter 3) may partly account for these findings, BD is fundamentally a brain illness. This, together with the converging evidence from human and animal studies that obesity is harmful to the brain, suggests the possibility that elevated BMI negatively impacts the neurobiology of BD. It further provides a compelling rationale for examining neurobiological correlates of obesity in BD. If obesity were shown to be associated with structural and functional brain changes in BD, particularly in brain circuits known to be relevant to the illness, this would suggest a biological explanation for why obese patients have more severe illness courses. However, to date, no neurobiological studies have examined BMI-related brain changes in patients with BD, or for that matter any psychiatric illness.  1.6  The Systematic Treatment Optimization Program for Early Mania (STOP-EM)  Thus, I chose to direct my doctoral research toward investigating clinical and neurobiological aspects of elevated BMI in people with BD. Specifically, I was interested in: 1) quantifying weight, BMI, and rates of overweight and obesity in patients at their first diagnosis of BD, and determining how these change during the early phase of maintenance treatment; 2) examining 9  the association of elevated BMI with the amount of time spent with mood symptoms in patients, and their overall level of psychosocial functioning, during early maintenance treatment; and 3) investigating whether elevated BMI is associated with structural and neurochemical brain changes in patients, particularly in brain areas known to be vulnerable in BD.   Thus, I used as my research subjects BD patients and healthy individuals enrolled in the University of British Columbia (UBC) Systematic Treatment Optimization Program for Early Mania (STOP-EM). STOP-EM, described in detail in Chapter 2, is a comprehensive study of clinical and functional outcomes, brain morphology, and neurochemistry in patients with BD who recovered from their first manic episode within the three months preceding enrolment (112). Age- (? 2 years) and gender-matched healthy subjects with no personal or family history of psychiatric illness were also recruited and followed for comparison purposes. Patients in the program received naturalistic treatment for BD according to current Canadian clinical practice guidelines (113). At enrolment and every 6 months, patients? psychiatric status and current pharmacologic treatments were assessed using standard clinical rating scales, and patients and healthy subjects underwent a physical examination, including height, weight, and calculation of BMI. At baseline, year 1, and then every two years, they received a cerebral magnetic resonance imaging scan (MRI) and underwent MR spectroscopy (MRS) to measure brain chemistry.   STOP-EM thus provided an excellent platform to examine clinical and neurobiological changes associated with overweight/obesity early in the course of BD. Among its advantages were: 1) Investigating neurobiological markers of BMI in euthymic first-episode mania patients allowed me to minimize the confounding effects of current mood state, illness duration, predominant episode polarity, and multiple medication trials, and also to assess whether BMI affected brain structure and function early in the course of BD. 2) Comparing patients to age- and gender-matched healthy subjects allowed me to assess the degree to which rates of overweight/obesity are truly elevated early in BD when other predictors of weight, such as age, gender, and location of residence, are taken into account, and the specificity of any neurobiological findings to BD.    10  1.7  Thesis overview  This thesis summarizes and synthesizes the results of investigations I conducted in STOP-EM patients, examining the prevalence of overweight and obesity in first-episode mania patients, their correlation with symptomatic and functional outcomes, and their relationship to structural and neurochemical changes in emotion-generating and regulating brain areas. In addition to this introduction, I have divided the thesis into five research chapters and a concluding discussion.  Chapters 2 and 3 describe clinical aspects of weight gain in BD. Chapter 2 reports the weight status of STOP-EM patients and healthy subjects at enrolment into the study, and quantifies the weight change each group experienced during the first year of maintenance treatment. Chapter 3 describes the relationships between one-year weight gain and clinically important outcomes such as the amount of time spent with mood symptoms and overall psychosocial functioning in the first year of treatment.  Chapters 4-6 report the association between baseline overweight/obesity and structural and neurochemical brain changes in STOP-EM patients and healthy subjects. Chapter 4 is an exploratory study investigating how weight is correlated with large-scale brain volumes, including total grey matter, total white matter, and total lobe volumes in each group. Based on the results of this initial work, I formulated a testable hypothesis: that elevated BMI is associated with unique brain changes in BD patients, such that the pathophysiologic alterations characteristic of the illness are exacerbated in overweight/obese patients. Chapters 5 and 6 report the results of additional investigations designed to test of this hypothesis, by examining the impact of elevated BMI on regional brain volumes in patients and healthy subjects (Chapter 5), and on hippocampal neurochemical changes (Chapter 6).  Finally, Chapter 7 synthesizes these findings in the context of current neurobiological models of BD, highlighting their clinical significance, and their importance in understanding the pathophysiology of the illness and interpreting the results of structural and functional brain imaging studies in BD patients. Novel research directions suggested by my results are also discussed. 11  Figure 1.1: Prevalence of obesity in studies comparing BD patients and healthy subjects      All studies except the one indicated reported higher obesity rates in people with BD than in healthy comparator subjects. Rate of Obesity (%)  United States  Europe  Australasia 12  Figure 1.2: Prevalence of MetS in studies comparing BD patients and healthy subjects       All studies except the one indicated reported higher MetS rates in people with BD than in healthy comparator subjects. Rate of MetS (%)  United States  Europe  Asia  Africa 13  2.  BMI at recovery from the first manic episode in patients with BD, and weight gain during the initial 12 months of maintenance treatment  2.1  Introduction  Studies in North American samples report that 55%-75% of patients with BD are either overweight (BMI = 25.00?29.99) or obese (BMI ? 30.00) (14,16-27), that the prevalence of overweight and obesity in BD are increasing with time (18), and that patients are more likely to be obese than people without BD (14,17,18). However, determining the degree to which the obesity rates in BD are truly elevated when other mediators of obesity are taken into account is difficult based on the currently available evidence. This is because studies examining obesity in BD share a number of limitations, including assessing patients with lengthy illnesses using cross-sectional or retrospective designs, and either not utilizing comparison groups (16,20,27) or comparing patients to population norms (14,17,18) or control subjects poorly matched for important sociodemographic variables (19).  These are important considerations, since obesity rates in the general population are also high, and have risen alarmingly in recent years. In 2004, the year for which the most recent data are available, 8.6 million Canadians, or 36.1% of the population, were overweight, and 5.5 million, or 23.1%, were obese (114). The rate of obesity in men almost doubled, from 13% to 23%, between 1978-79 and 2004, and the percentage of women with obesity increased substantially, from 16% to 23%, during the same period (115). Furthermore, many sociodemographic variables have been shown to impact weight and BMI, including age, gender, socioeconomic status, and place of residence (116). Thus, a control group that is well matched for sociodemographic predictors of obesity is essential to accurately gauge the degree of adiposity in BD relative to people without the illness. However, to our knowledge, the only study that made a direct comparison between BD patients and a control group, as opposed to a general population estimate, in fact utilized a historical control and reported only a moderately increased rate of overweight/obesity in the patients (55% vs 46%) (26).  14  Additional limitations of studies examining obesity in BD include their largely cross-sectional designs, which provide only a snapshot  in time, and generate little information regarding the course of weight gain relative to the affective component of the illness. Thus, whether weight gain occurs rapidly or gradually, and whether it is an early or late complication of BD, has not been well documented. Data from early in the illness are especially lacking. This is important clinically, since the point at which clinicians should intervene to prevent weight gain is unclear, and neurobiologically, because if obesity does in fact impact on brain structure and function in BD, the timing of this is also unknown. However, only two studies have assessed weight early in the course of bipolar disorder, and they produced contradictory results. One US study found no difference in obesity rates between BD patients and healthy controls at age 18 years, but it ascertained illness-onset weight retrospectively, and so was limited by recall bias (19). In contrast, an Italian study which enrolled never-medicated BD patients reported their rate of overweight/obesity was 3 times greater than the age-matched Italian general population (43).   A final limitation of the available data is that it is not known whether the same sociodemographic variables that predict weight gain in the general population also do so in BD, or which clinical and treatment factors play roles. For example, many medications used to treat bipolar disorder, including lithium (117-120), valproic acid (118,119), and second-generation antipsychotics (SGAs) (38,121) have been associated with weight gain in clinical samples, but their relative propensities for this have not been systematically studied. Knowledge regarding risk factors for weight gain would allow psychiatrists and researchers to identify high-risk populations for clinical intervention and for neurobiological and clinical research.   The purposes of the current analysis, then, are 1) to quantify baseline data on BMI and rates of overweight/obesity in a sample of BD patients who recently recovered from their first manic episode, compared to an age- and gender-matched sample of healthy subjects; 2) to prospectively measure weight gain and changes in BMI and rates of overweight/obesity in patients and healthy subjects during the initial 12 months of maintenance treatment for BD, and 3) to determine which sociodemographic, clinical, and treatment factors are associated with weight gain in BD. We hypothesized that mean BMI and rates of overweight/obesity would be similar at enrolment 15  into STOP-EM, but that patients would experience significantly greater weight gain than healthy subjects over the initial 12 months of maintenance treatment.  2.2  Methods  2.2.1  STOP-EM  STOP-EM is a prospective study of clinical and functional outcomes, health status, brain morphology, neurochemistry, and quality of life in BD patients who recently recovered from their first episode of DSM-IV-TR-defined mania (112). The program was initiated at the UBC Mood Disorders Centre in July 2004, and has been continuously active since then. STOP-EM and the procedures described here were in accordance with the Declaration of Helsinki of 1975 regarding research in human subjects, and were approved by the UBC Clinical Research Ethics Board. Written informed consent was obtained from all patients and healthy subjects prior to any study procedures taking place.  Eligible patients were recruited from Vancouver General and UBC Hospitals and affiliated sites, as well as through referrals from community physicians and psychiatrists. Inclusion criteria were deliberately broad, so as to capture the full range of BD patients routinely encountered in clinical practice: 1) age 14 to 35, 2) recovered from the first manic episode within the 3 months preceding enrolment, 3) fluent in English, and 4) capable of understanding, consenting to, and complying with the study requirements. Patients could be enrolled as inpatients or outpatients, and could be diagnosed with a pure or mixed manic episode, with or without psychotic features, and with or without comorbidity, including substance abuse or dependence. Exclusion criteria were few, including only: 1) manic episode due to a medical condition, such as multiple sclerosis, or substance misuse, such as cocaine intoxication (which would negate the diagnosis of BD and mandate instead the diagnoses of manic episode secondary to a general medical condition or substance-induced manic episode, respectively), 2) history of a previous manic episode diagnosed retrospectively on structured interview or via collateral information, and 3) inability to take part in MRI scans for medical or other reasons.  16  Healthy subjects matched for age (? 2 years) and gender were also recruited into STOP-EM and assessed for comparison purposes. Inclusion criteria for healthy subjects were: 1) age 14 to 35, 2) fluent in English, and 3) capable of understanding, consenting to, and complying with the requirements of the study. Exclusion criteria included: 1) history of a psychiatric illness in the subject or his/her first-degree relatives, and 2) inability to take part in magnetic resonance imaging for medical or other reasons.  At baseline, all patients underwent a thorough psychiatric assessment by a research psychiatrist with expertise in the diagnosis of mood disorders. With patients? consent, collateral information was also obtained from family members, community physicians and psychiatrists, and/or health records. The diagnoses of BD and first manic episode were confirmed based on all available clinical information, supplemented by the semi-structured Mini International Neuropsychiatric Interview (MINI) (122). Healthy subjects were assessed with a structured interview and the MINI, and were enrolled if they had no personal history of psychiatric illness and no family history in their first- or second-degree relatives.   Sociodemographic data and information about prior depressive and hypomanic episodes and comorbid conditions were collected using a standardized protocol, and historical timelines of patients? illnesses were constructed using National Institutes of Mental Health (NIMH) Life Charts (123). Current psychiatric symptoms were quantified using a comprehensive battery of standard clinical rating scales, including the Young Mania Rating Scale (YMRS) (124), Hamilton Depression Rating Scale, 29-item version (HAM-D-29) (125), Montgomery-Asberg Depression Rating Scale (MADRS) (126) , Brief Psychiatric Rating Scale (BPRS) (127), Positive and Negative Syndrome Scale (PANSS) (128), and Clinical Global Impression Scale for BD (CGI-BP) (129). Overall psychosocial functioning was assessed using the Multidimensional Scale of Independent Functioning (MSIF) (130).  Patients enrolled in the program received open-label maintenance treatment for BD from clinicians with expertise in managing mood disorders. Symptomatic remission was defined as a YMRS score of ? 8 and HAM-D score of ? 8, a stringent definition which excludes patients with even mild subsyndromal mood symptoms (131,132). Recurrence was defined as the re-17  emergence of clinically significant mood symptoms, in the opinion of the psychiatrist, occurring after a period of remission. They were assessed as clinically indicated, and at a minimum of every 6 months. At each visit, they underwent a psychiatric assessment, and clinical rating scales, as above, were re-administered. Information about any mood episodes between study visits was ascertained using Life Charts. Recurrences were treated according to Canadian Network for Mood and Anxiety Treatments (CANMAT) clinical practice guidelines (113). Within the framework of these guidelines, clinicians had the latitude to tailor treatments individually to meet each patient?s specific needs. In addition to regular psychiatric assessment and medication treatment, patients were provided with supportive therapy, and were offered the opportunity to participate in a standardized 8-week group psychoeducation program geared toward promoting optimal self-management strategies for BD.  2.2.2  Magnetic resonance imaging (MRI) and proton MR spectroscopy (1H-MRS)  At baseline, year 1, and then every two years, all patients and healthy subjects underwent a 3-Tesla (3T) MRI scan to assess brain structure, and single-voxel hippocampal 1H-MRS to assess hippocampal neurochemistry. These procedures are outlined in detail in chapters 4, 5, and 6.   2.2.3  Assessment of weight, BMI, and metabolic indices  A physical examination, including blood pressure, height, and weight was performed on all patients and healthy subjects at baseline and at each 6 month visit. Blood pressure was measured with the participant in the sitting position for 5 minutes. Patients and healthy subjects were weighed in a non-fasting state in light clothing with footwear removed. Weights were collected by a single research coordinator using the same scale. Laboratory metabolic indices, including total cholesterol, HDL, low-density lipoprotein (LDL), triglycerides, blood glucose, serum insulin, and thyroid stimulating hormone (TSH) were measured in patients after a 12 hour fast at baseline and every 6 months.   BMI was calculated using the formula BMI = weight (kg) / height (metres)2. Definitions of underweight, normal weight, overweight, and obesity were based on criteria devised by the 18  World Health Organization (133). Specifically, underweight was defined as a BMI < 18.50; normal weight as a BMI of 18.50-24.99; overweight as a BMI of 25.00-29.99; and obesity as a BMI ? 30.00. Normal ranges for metabolic indices were taken, when available, from the definitions of the National Cholesterol Education Program (NCEP) (29), including: 1) elevated serum glucose: serum glucose ? 6.1 mmol/L; 2) hypertriglyceridemia: serum triglycerides ? 1.69 mmol/L; and 3) low HDL: serum HDL ? 1.04 mmol/L in men and ? 1.29 mmol/L in women. Normal values for the remainder of the metabolic indices were based on reference ranges from the study laboratory and include: 1) elevated total cholesterol: serum cholesterol ? 4.6 mmol/L; 2) elevated LDL: serum LDL > 3.0 mmol/L; 3) elevated insulin: fasting serum insulin > 70 mU/L; and 4) elevated TSH: TSH > 5.50 mU/L. The NCEP definition of elevated blood pressure, ie a pressure of ? 130/85, was utilized.  2.2.4  Data analysis and statistics  Our primary objective was to measure the percentage of patients and healthy subjects who experienced clinically significant weight gain (CSWG), ie. who gained ? 7% of their baseline body weight, after 6 months and 12 months of follow-up. We included in our analysis all patients and healthy subjects who had baseline data for height and weight, and data for weight from at least one follow-up visit. Missing weight data at the 6 month and 12 month follow-up visits were accounted for using the last observation carried forward (LOCF) method. Secondary outcome measures included mean BMI and rates of overweight and obesity in patients and healthy subjects at each time point, mean values for patients? laboratory metabolic indices, and the percentage of patients with laboratory values outside the normal ranges.   We used logistic regression to examine predictors of clinically significant weight gain at 6 months and 12 months. Finally, as second generation antipsychotics (SGAs) (134), mood stabilizers (134), the number of medications prescribed (28), and male gender (14,16,27) have been associated with overweight and obesity in previous studies, we performed subgroup analyses comparing weight gain, overweight and obesity in patients taking the SGAs olanzapine versus risperidone versus quetiapine; in those taking the mood stabilizers lithium versus divalproex; in patients prescribed zero, one, two, three or four medications at any point over the 19  follow-up period; and in males versus females. For clarity of analysis, in the olanzapine versus risperidone versus quetiapine analysis and the lithium versus divalproex analysis, we did not include patients who were prescribed more than one SGA or more than one mood stabilizer; this resulted in the exclusion of 5 patients from the SGA analysis and 6 patients from the mood stabilizer analysis. For the analysis in patients taking different numbers of medications, we counted all mood stabilizers and antipsychotics prescribed at any point during the 12-month follow-up period.  We carried out all statistical analyses using SPSS for Windows, version 16.0 (SPSS Inc, Chicago, IL). Baseline sociodemographic variables were analyzed using t-tests, analysis of variance (ANOVA), and ?2 tests as appropriate. Comparisons of weight gain, BMI, mean values for laboratory metabolic indices, and other continuous variables between patients and healthy subjects were performed using analysis of covariance (ANCOVA), with age and gender as covariates. In the subgroup analyses of patients taking specific medications, the patients were not well-matched for initial BMI, and this was entered as a covariate in these analyses. As well, in the analyses of patients taking olanzapine versus risperidone versus quetiapine, mood stabilizer use was entered as a covariate, while in the analyses of patients taking lithium versus divalproex, SGA use was entered as a covariate. Comparisons between male and female patients used age as a covariate. Categorical variables, including rates of CSWG, overweight, obesity, and abnormal metabolic indices were analyzed using ?2 tests. In any instance where expected cell counts were < 5, Fisher?s exact test was employed. All statistical tests were two-tailed and were carried out using a significance level of ?=0.05. Aside from our primary objective, our analyses were exploratory and we did not correct for multiple comparisons.  We performed logistic regression using the forced-entry method. We selected variables for inclusion in the regression models based on research demonstrating an association with obesity in BD (14,16,17) and/or the general population (116). These included age, gender, ethnicity, socioeconomic status (SES), baseline weight, age at onset of mood disorder, number of previous depressive and hypomanic episodes, history of substance abuse, treatment with mood stabilizers and SGAs, level of functioning, and recurrence of depression during the follow-up period.   20  2.3  Results  2.3.1  Demographic and clinical data   Of the initial 62 patients and 25 healthy subjects enrolled in STOP-EM, 47 patients and 24 healthy subjects had baseline data for height and weight, and data for weight from at least one post-baseline assessment, and were included in our analyses. Weight data were available for 44 patients and 21 healthy subjects at the 6 month follow-up, and for 39 patients and 21 healthy subjects at the 12 month follow-up.   Sociodemographic characteristics of patients and healthy subjects are reported in Table 2.1, and patients? clinical characteristics are enumerated in Table 2.2. Patients and healthy subjects were well-matched with respect to age, gender, ethnicity, years of education, and marital status. The average duration of the first manic episode was 65.0 days. Sixty-four percent of patients had experienced a previous depression or hypomania, with the first onset of mood symptoms a mean of 2.9 years prior to the index mania. Comorbidity, particularly with substance use, was common. However, patients had low levels of symptomatology at enrolment, with 66.0% in full remission.   Psychiatric medications at baseline and during the 1-year follow-up period are listed in table 2.3. At enrolment, 87.2% of patients were treated with a mood stabilizing medication, and 80.9% were treated with an SGA. Seventy-five percent received combination therapy with a mood stabilizer and an SGA. Mood stabilizer use remained relatively constant throughout the follow-up period, but the percentage of patients treated with antipsychotics dropped to approximately 50% at 6 months and 12 months.   2.3.2  Weight gain, overweight, obesity, and metabolic indices  Table 2.4 presents baseline, 6 month, and 12 month data for BMI, overweight, and obesity in patients compared to healthy subjects, patients taking risperidone versus olanzapine versus quetiapine, patients taking lithium versus divalproex, and males versus females. Figure 2.1 21  illustrates mean weight gain and rates of CSWG in patients and healthy subjects at each time point. Figure 2.2 shows the proportions of patients and healthy subjects experiencing various degrees of weight gain or loss by 12 months, expressed as a percentage change from their baseline weight.  Baseline   At baseline, patients and healthy subjects were similar in mean BMI and rates of overweight and obesity (Table 2.4). There were no significant differences between obese and non-obese patients in mean values for any laboratory metabolic indices, or in the proportion with elevated blood pressure or values for laboratory metabolic indices outside the normal ranges.   6 Month Follow-up  During the first 6 months of maintenance treatment, 46.8% of patients and 4.2% of healthy subjects experienced CSWG (?2=13.191, df=1, p=0.000) (Figure 2.1a). Patients gained a mean of 4.57 kg, compared to 0.51 kg in healthy subjects (F=5.375, df=1, p=0.023) (Figure 2.1b). Mean 6-month BMI increased to the overweight range in patients, though rates of overweight and obesity did not differ significantly between patients and healthy subjects. Logistic regression demonstrated that CSWG was associated with lower initial weight (B=-0.276, df=1, p=0.037), male gender (B=7.271, df=1, p=0.047), and treatment with olanzapine (B=8.761, df=1, p=0.013) or risperidone (B=9.440, df=1, p=0.029).   Patients who were obese at 6 months had significantly greater mean serum triglyceride levels (2.15 vs 1.13 mmol/L, F=5.187, df=1, p=0.037) and fasting glucose levels (5.80 vs 4.89 mmol/L, F=6.480, df=1, p=0.023) than non-obese patients. No differences were detected in the proportions of values outside the normal ranges for any of the laboratory metabolic indices we measured.     22  12 Month Follow-up  At 12 months, a non-significant trend suggested that more patients than healthy subjects experienced CSWG from baseline (48.9% vs 25.0%; ?2=3.767, df=1, p=0.052) (Figure 2.1a). Nineteen percent of patients and 4% of healthy subjects gained ? 15% of their baseline weight, and 4% of patients and 0% of healthy subjects gained ? 25% of their baseline weight (Figure 2.2). Mean weight gain was 4.76 kg in patients and 1.50 kg in healthy subjects (F=4.097, df=1, p=0.047) (Figure 2.1b). Mean 12-month BMI remained in the overweight range for patients, and a non-significant trend suggested a greater combined rate of overweight and obesity in patients compared to healthy subjects (?2=3.704, df=1, p=0.054).  The differences we observed between patients and healthy subjects at 12 months were due primarily to differential weight gain during the first 6 months of follow-up, as mean weight change during the second 6 months was minimal in patients and did not differ significantly from healthy subjects (0.19 kg vs 0.98 kg, F=0.439, df=1, p=NS). From 6 to 12 months, 13% of patients and 8.3% of healthy subjects gained ? 7% of their 6-month body weight, while 8.5% of patients and 0% of controls lost ? 7% of their 6-month weight. Logistic regression did not demonstrate an association between CSWG from 6 to 12 months and any of the variables we assessed.   Patients who were obese at 12 months had greater mean serum triglyceride levels (4.29 vs 1.19 mmol/L, F=10.609, df=1, p=0.006) and fasting glucose levels (11.67 vs 4.92 mmol/L, F=4.834, df=1, p=0.043) compared to non-obese patients. No differences were detected between obese and non-obese patients in the percentages with values outside the normal ranges for any laboratory metabolic indices.  Subgroup Analyses  Patients taking SGAs.  At baseline, numerically more quetiapine-treated patients than olanzapine- or risperidone-treated patients were overweight or obese (50.0%, compared to 30.0% and 25.1%, respectively) (Table 2.4), but this difference did not reach statistical significance. 23  When initial BMI was controlled for, there was a significant difference in weight gain between SGAs, with patients who were prescribed olanzapine at any time point (N=10) gaining a mean of 11.38 kg by 12 months, patients treated with risperidone (N=16) gaining 4.12 kg, and those treated with quetiapine (N=10) losing 0.35 kg, (F=3.378, df=2, p=0.048). Weight gain in olanzapine-treated patients was significantly greater than in patients prescribed risperidone (p=0.010) or quetiapine (p=0.000). The proportion of patients with CSWG did not differ significantly by treatment (70.0% for olanzapine, 43.8% for risperidone, and 30.0% for quetiapine; ?2=3.349, df=2, p=NS). Rates of overweight and obesity at 12 months were also similar for the three treatments.   Patients taking lithium versus divalproex. At baseline, numerically more divalproex treated patients than lithium-treated patients were overweight or obese (52.6% vs 31.6%; ?2=1.610, df=1, p=NS) (Table 2.4). When initial BMI was controlled for, mean weight gain did not differ significantly between patients who were prescribed lithium at any time point (N=19) and those prescribed divalproex at any time point (N=19) (6.57 kg for lithium and 2.60 kg for divalproex; F=0.821, df=1, p=NS), although more lithium-treated patients than divalproex-treated patients experienced CSWG (63.2% vs 26.3%; ?2=5.216, df=1, p=0.022). Rates of overweight and obesity at 12 months were similar in lithium- and divalproex-treated patients.    Number of Medications. At baseline, rates of overweight or obesity were similar in patients who would go on to be treated with zero, one, two, three, or four medications. When initial BMI was controlled for, weight gain over 12 months did not differ significantly depending on the number of medications prescribed (Table 2.4). Only one patient was prescribed no medication, and only one was prescribed 4 medications; the mean weight gain in these patients was 5.40 kg and 3.72 kg, respectively. Mean weight gain did not differ significantly between patients prescribed one medication (3.90 kg; N=6), those prescribed two medications (5.07 kg; N=31), and those prescribed three medications (5.08 kg; N=8) (F=0.283, df=4, p=NS). The two patients prescribed either one medication or 4 medications each experienced CSWG; rates of CSWG ranged from 37.5%-50% in patients prescribed one, two or three medications, with no significant difference between treatments (?2=2.512, df=4, p=NS).  24  Males versus females.  At baseline, male patients and male healthy subjects had similar mean BMIs and did not differ in rates of overweight or obesity (Table 2.4). Over 12 months of treatment, CSWG occurred significantly more frequently in male patients (58.3% vs 16.7%; ?2=5.625, df=1, p=0.018), and a non-significant trend suggested greater weight gain in male patients than male healthy subjects (7.20 kg vs 2.03 kg; F=3.969, df=1, p=0.055). Combined rates of overweight and obesity at 12 months were greater in patients than healthy subjects (?2=5.573, df=1, p=0.018).   Like males, female patients and healthy subjects were similar with respect to mean baseline BMI and rates of overweight and obesity. In contrast to males, comparable numbers of female patients and healthy subjects experienced CSWG over 12 months (39.1% vs 33.3%; ?2=0.114; df=1; p=NS), with mean weight changes of 2.51 kg and 1.04 kg, respectively (F=0.506, df=1, p=NS). Rates of overweight and obesity were similar in female patients and healthy subjects at 12 months.  2.4  Discussion  Our results support our hypotheses by demonstrating that at recovery from a first manic episode, there was no difference between BD patients and healthy subjects in mean BMI or rates of overweight or obesity. However, during just 6 months of maintenance treatment, patients gained a mean of 4.5 kg, and approximately half gained ? 7% of their initial weight, the generally-accepted definition of CSWG. Weight gained during the first 6 months was maintained during the subsequent 6 months, and at both 6 months and 12 months the combined rate of overweight and obesity in patients exceeded 50% and was approximately twice that of healthy subjects. Male patients and those treated with olanzapine and risperidone were particularly prone to weight gain. We also detected early evidence of metabolic consequences of obesity, with significant differences in serum triglyceride and glucose levels in obese compared to non-obese patients. In contrast, the healthy subjects in our sample experienced modest weight change, and their rates of overweight and obesity remained relatively constant during the follow-up period.   25  Overweight and obesity are associated with increased risks for numerous medical conditions, including hypertension, type II diabetes, ischemic heart disease, stroke, and several types of cancer (11). The direct medical costs of obesity in the US have been estimated at $92.6 billion annually (135), and are likely to be similar in Canada on a per capita basis. Similarly, obesity adds to the burden of illness experienced by patients with in BD and the health care costs borne by society. Obesity has been associated with poor psychiatric outcomes in BD, including shorter time to recurrence of mood episodes (17) and more frequent suicide attempts (136). The prevalence of medical conditions associated with obesity has also been reported to be increased in patients with BD (66), and SMRs due to cardiovascular disease are approximately twice the age-adjusted population rates (75). Health care costs for medical conditions related to obesity, such as hyperlipidemia, atherosclerosis, and myocardial infarction, are over twice as high in patients with BD compared to people without BD (137).   The substantial weight gain and changes in metabolic indices we observed early in the course of BD underscores the importance of frequently monitoring the physical health of patients from the outset of treatment. A joint committee of the American Psychiatric Association, the American Diabetes Association, the American Association of Clinical Endocrinologists, and the North American Association for the Study of Obesity recently published guidelines for monitoring the health status of patients taking antipsychotic medications, and these are easily adaptable for monitoring patients with BD (138). They include taking a personal and family history of obesity and related medical conditions during the initial patient interview; measuring BMI, blood pressure, and laboratory metabolic indices at baseline and regularly during treatment; counselling patients regarding healthy lifestyles; and intervening if necessary to minimize and reverse weight gain. We recommend that clinicians treating patients with BD use these guidelines to monitor, prevent, and manage weight gain and related medical conditions.   Although the open-label design of our study precludes making inferences regarding causal factors for weight gain, it is noteworthy that at enrolment in STOP-EM, patients with BD had similar BMIs and rates of overweight and obesity as healthy subjects. This is despite the fact that almost two-thirds had experienced previous hypomanic and/or depressive episodes. This suggests that maintenance treatment with mood stabilizing and antipsychotic medications, rather 26  than the presence of a mood disorder, is primarily responsible for the weight gain we observed. In keeping with this, we, like others, found that patients taking olanzapine experienced greater weight gain than those taking risperidone or quetiapine, and that olanzapine and risperidone were associated with clinically significant weight gain at 6 months in our regression analysis. Clinicians should consider both efficacy and safety when choosing medications for BD patients, and alternative pharmacotherapeutic options should be considered if substantial weight gain occurs.  Our finding that male patients experience greater weight gain than females is also in keeping with previous studies that reported higher rates of overweight and obesity in males with BD (14,16,27). The magnitude of the difference that we observed between males and females was nonetheless surprising. Male patients gained over 2-1/2 times as much weight as females, and were almost twice as likely to be overweight or obese at 12 months. In fact, essentially the entire difference between BD patients and healthy subjects in CSWG, overweight, and obesity was due to greater rates in male patients. These results suggest that males and females with BD have different metabolic sequelae, and that considering them together may be misleading, confounding the overall results of studies of BMI and obesity.   The results presented here, and the conclusions we have drawn, must be evaluated in light of the limitations of our report. As noted, this was a naturalistic study, and so we are not able to make causal inferences regarding medications or other factors that we identified as being associated with weight gain. As well, the STOP-EM program enrolled patients following treatment of their first manic episode, and we do not have information on weight changes during treatment of the acute mania or related to prior depressive or hypomanic episodes. Much of the data that we collected at the time of enrolment, for example regarding previous course of mood illness, was ascertained retrospectively. We did not have complete follow-up weight data for all of our patients, and we cannot exclude the possibility that the missing values might have affected our findings. We did not gather information on non-pharmacological interventions to prevent weight gain, such as modifying diet and exercise patterns. Finally, we did not obtain blood samples from healthy subjects, and we are unable to compare them to patients regarding laboratory metabolic indices.  27   Nonetheless, the current investigation has several strengths, and this report contains several notable findings. STOP-EM is the first program to prospectively measure weight gain and rates of overweight and obesity in BD starting at recovery from the first manic episode, and the first to compare weight gain in patients with BD to a well-matched comparison group of healthy subjects. It has broad inclusion criteria and few exclusion criteria, and the patients in our sample are thus highly reflective of those seen in routine clinical practice. We provided treatment according to current clinical practice guidelines, and within this framework clinicians were given the latitude to tailor treatment to the needs of individual patients, making the results of the current report highly generalizable to real-world clinical settings. To the best of our knowledge, this is the first prospective study to report that although BMI and rates of overweight and obesity do not differ between patients with BD and healthy subjects at recovery from a first manic episode, rapid and CSWG occurs early in the course of maintenance treatment. Some patient subgroups, particularly males and those treated with olanzapine and risperidone, are at particular risk of weight gain. We also detected early evidence of metabolic abnormalities in obese patients early in the course of maintenance treatment. Given the long-term health consequences of overweight and obesity, these findings underscore the importance of considering weight and metabolic factors when making even the earliest treatment decisions for patients with BD, and of frequently monitoring for and addressing weight gain.  28  Table 2.1. Baseline sociodemographic characteristics of 47 BD patients and 24 healthy subjects enrolled in STOP-EM      Age Patients (N=47) Healthy Subjects (N=24) P value Mean SD Mean SD  22.8 3.7 22.5 4.8 0.719 Education (years) 14.1 2.2 14.7 2.8 0.330    N  %  N  %  Gender      0.932 male 24 51.1% 12 50.0%  female 23 48.9% 12 50.0%  Ethnicity (*)     0.467 Caucasian 36 78.3% 18 75.0%  Asian 8 17.4% 6 25.0%  Other 2 5.6% 0 0%  Marital status     0.253 Single 42 89.4% 24 100%  Married 4 8.5% 0 0%  Separated or divorced 1 2.1% 0 0%  Employment     0.639 Full time or part-time employment 19 40.4% 7 29.1%  Student 26 55.3% 16 66.7%  Unemployed 2 4.3% 1 4.2%    * - 1 value missing for patients; N=4629  Table 2.2: Baseline clinical characteristics of BD patients    Mean SD Duration of index mania (days) (*) 65.0 55.0 Baseline YMRS score 3.7 6.1 Baseline MADRS score 5.8 8.1 Baseline HAM-D-29 score 6.4 9.1 Baseline BPRS score 22.4 5.8 Baseline PANSS score 8.0 1.7    Age at onset of illness (years) (*) 20.0 4.7 Duration of illness (years) (*) 2.9 3.5     N % Psychotic features with index mania   Yes 35 74.5% No 12 25.5%    In remission at enrolment  31 66.0%    Previous depressive episode 28 60.0% Previous hypomanic episode 10 21.3% Previous mood episode (depressive or hypomanic) 30 63.8% Previous suicide attempt 6 12.8%    Lifetime comorbidity   Anxiety 5 10.6% Attention deficit-hyperactivity disorder (ADHD) 4 8.5% Alcohol abuse or dependence 10 21.3% Cannabis abuse or dependence 19 40.4% Other drug abuse or dependence 5 10.6% Medical (+) 11 23.4%   * - N=46; 1 value missing + - N=43; 4 values missing30  Table 2.3: Medication use by BD patients at baseline, 6 months, and 12 months        Baseline (N=47) 6 Months (N=46) 12 Months (N=39)  N % N % N %        Mood Stabilizers       Lithium 19 40.4% 20 43.5% 17 43.6% Divalproex 23 48.9% 21 45.7% 19 48.7% Carbamazepine 0 0% 0 0% 0 0% Lamotrigine 1 2.1% 0 0% 0 0% Topiramate 0 0% 0 0% 0 0% Any mood stabilizer 41 87.2% 41 89.1% 34 87.2% > 1 mood stabilizer 2 4.3% 0 0% 2 5.1%        Antipsychotics (*)       Risperidone 17 36.2% 11 23.9% 8 20.5% Olanzapine 10 21.3% 6 13.0% 8 20.5% Quetiapine 11 23.4% 5 10.9% 6 15.4% Loxapine 2 4.3% 0 0% 0 0% Any antipsychotic 38 80.9% 22 47.8% 21 53.8% > 1 antipsychotic 2 4.3% 0 0% 1 2.6%        Monotherapy and Combination Therapy       Mood stabilizer only 6 12.8% 22 47.8% 15 38.5% Antipsychotic only 3 6.4% 3 6.5% 2 5.1% Mood stabilizer + antipsychotic 35 74.5% 19 41.3% 19 48.7% No treatment 3 6.4% 2 4.3% 3 7.7%        Other       Antidepressant 3 6.4% 6 13.6% 5 12.8% Benzodiazepine 4 8.5% 0 0% 3 7.7% Calcitonin 2 4.3% 0 0% 0 0% Tapazole 1 2.1% 0 0% 0 0% L-thyroxine 0 0% 2 4.5% 2 5.1% Orlistat 0 0% 1 2.3% 0 0% Warfarin 0 0% 0 0% 1 2.6% 31  * - A number of patients were enrolled in a placebo-controlled clinical trial and were randomized to receive either a SGA or placebo:  ? baseline: 2 patients received risperidone or placebo ? 6 months: 2 patients received risperidone or placebo and 1 patient received olanzapine or placebo ? 12 months: 1 patient received olanzapine or placebo 32  Table 2.4: Mean BMI and rates of overweight and obesity at baseline, 6 months, and 12 months in BD patients and healthy subjects     Baseline 6 months 12 Months  BMI (kg/m2) Over-weight (%) Obese (%) BMI (kg/m2) Over-weight (%) Obese (%) BMI (kg/m2) Over-weight (%) Obese (%)           Patients (N=47) 23.97 25.5% 8.5% 25.48  46.8% 8.5% 25.57 42.6% 10.6% Healthy Subjects (N=24) 23.61 20.8% 4.2% 23.76 29.2% 4.2% 24.07 20.8% 8.3%           Patients Prescribed Risperidone (N=16)* 23.83 18.8% 6.3% 25.70 50.0% 12.5% 25.23 31.3% 12.5% Patients Prescribed Olanzapine (N=10)* 23.65 30.0% 0% 26.63 60.0% 10.0% 27.35 60.0% 20.0% Patients Prescribed Quetiapine (N=10)* 25.11 20.0% 30.0% 24.32 30.0% 10.0% 24.96 40.0% 10.0%           Patients Prescribed Lithium (N=19) * 23.92 26.3% 5.3% 25.68 57.9% 5.3% 26.06 52.6% 10.5% Patients Prescribed Divalproex (N=19) * 25.26 36.8% 15.8% 26.31 42.1% 15.8% 26.14 36.8% 15.8%           Male Patients (N=24) 24.36 33.3% 4.2% 26.62 54.2% ** 12.5% 26.64 50.0% ** 16.7% Male Healthy Subjects (N=12) 23.86 16.7% 8.3% 24.11 16.7% 8.3% 24.46 8.3% 16.7% Female Patients (N=23) 23.56 17.4% 13.0% 24.28 39.1% 4.3% 24.52 34.8% 4.3% Female Healthy Subjects (N=12) 23.35 25.0% 0% 23.41 41.7% 0% 23.71 33.3% 0%  * patients prescribed both lithium and divalproex during the 1-year follow-up period, or prescribed more than one SGA, were not included in the analysis ** p < 0.05 vs male healthy subjects      33  Figure 2.1: a) Mean weight gain (kg) and b) proportions with CSWG over 12 months in 47 BD patients and 24 healthy subjects            a)  * p < .05          b)  ** p < 0.001   *** p = 0.052   *   * Weight Gain (Kg) Per Cent ** *** 0123456Baseline 6 Months 12 MonthsFEM Patients(observed)FEM Patients(LOCF)Healthy Subjects(observed)Healthy Subjects(LOCF)*  002030405060Proportion with Clinically Significant Weight Gain  from Baseline to 6 MosProportion wi th Clinically Significant Weight Gain form Baseline to 12 MosFEM PatientsHealthy Subjects* 34  Figure 2.2: Weight changes over 12 months a) BD patients and b) healthy subjects (as a percentage of baseline weight)                     a)                                b) 051015202530-15% --10%-9.9% - -5%-4.9% -0%0.1% -5%5.1% -10%10.1%- 15%15.1%- 20%20.1%- 25%25.1%- 30%30.1%- 35%051015202530-15% --10%-9.9%- 5%-4.9%- 0%0.1% -5%5.1% -10%10.1%- 15%15.1%- 20%20.1%- 25%25.1%- 30%30.1%- 35%Percent of Patients Percent of Healthy Subjects Weight Change (as a Percentage of Baseline Weight) Weight Change (as a Percentage of Baseline Weight)  35  3.  The association of weight gain with mood symptoms and impaired psychosocial functioning in BD patients during the initial 12 months of maintenance treatment after a first manic episode  3.1  Introduction  As described in the previous chapter, BD patients gained, on average, approximately 5 kg in body weight in the 12 months following their first manic episode, and almost half met the criteria for CSWG. This weight gain was substantially greater than that experienced by age-and gender matched healthy subjects free of psychiatric illnesses, and it was associated with demonstrable adverse medical consequences, including elevated serum triglyceride and glucose levels.   However, it is not only physical health that is impacted by obesity in people with BD. A growing body of evidence suggests that obesity also has deleterious effects on the affective component of the illness. Patients with BD spend approximately half of their lives with mood symptoms, with the depressive phase particularly difficult to stabilize (3,99). Compared to normal-weight patients, obese patients have histories characterized by greater numbers of depressive and possibly manic episodes (16,17), more hospitalizations for depression (21,34), more frequent suicidal ideation (27), and more suicide attempts (18,136). While these data are limited by the retrospective ascertainment of mood symptoms and suicidality, one prospective study also reported that obese BD patients had shorter periods of wellness and experienced more frequent depressive relapses (17). However, all of the studies to date were carried out in patients with lengthy histories of BD, and data from early in the course of the illness are almost completely lacking.  Suggestive preliminary studies also point to the possibility that excessive weight gain adversely affects functional outcomes in BD. Full recovery of pre-illness psychosocial functioning, and not just symptomatic recovery, is the ultimate goal in treating BD. Nonetheless, BD patients often suffer prolonged functional impairment in numerous domains, including the ability to live independently, work productively, and maintain satisfying relationships (4). Furthermore, functional recovery is frequently delayed or absent even when patients are in full clinical 36  remission, with only 25%-39% of remitted patients returning to their premorbid level of functioning in the year following the resolution of a mood episode (103-106). This suggests that factors other than mood symptoms contribute to the disability associated with BD. In the general population, obesity has consistently been associated with impaired occupational, social, and relationship functioning, which improves significantly following even modest weight loss, suggesting a causal relationship between obesity and functional impairment (107,108). To date, however, only two studies have assessed the impact of weight on functional outcomes in clinical samples of BD patients. The first reported that obesity was associated with diminished self-esteem, impaired social and occupational functioning, and reduced physical abilities in a sample of 100 US patients (139). In the second, obese patients had lower Global Assessment of Functioning (GAF) scale scores than non-obese patients in a Canadian sample (109). However, these reports were limited by their cross-sectional designs and assessment of patients with lengthy illnesses. Prospective data, and data from early in the course of BD, are lacking.  Thus, there is a clear need for prospective studies early in the course of BD to determine the impact of weight gain, overweight, and obesity on the amount of time BD patients spent with mood symptoms, and on their overall level of psychosocial functioning. The analyses in the current chapter, then, examine the relationship between weight gain in the first 12 months of maintenance treatment and psychiatric outcomes in our first-episode mania sample. Specifically, we hypothesized that patients who experienced CSWG in the 12 months following their first manic episode would spend significantly more time depressed, and have poorer overall psychosocial functioning at 12 months.  3.2   Methods  3.2.1  STOP-EM  The analyses reported here were carried out in BD patients enrolled in the UBC STOP-EM program. The inclusion and exclusion criteria and assessment procedures for STOP-EM were as described in detail in Chapter 2. Because time spent with mood symptoms and psychosocial 37  functioning were not measured in STOP-EM healthy subjects, the analyses reported here are limited to BD patients.  3.2.2  Assessment of time spent with mood symptoms and relapse rates   Mood symptoms at enrolment into STOP-EM and at the 6-month and 12-month visits were prospectively assessed with clinical rating scales, including the YMRS, HAM-D-29, and MADRS. The number of days patients spent depressed, manic, hypomanic, and euthymic (free of mood symptoms) in the previous 6 months was quantified using NIMH Life Charts, which were completed at the 6-month and 12-month visits (123). Relapse was defined as the emergence of clinically significant mood symptoms, in the opinion of the treating psychiatrist, following a period of remission. We used data from the NIMH Life Chart, supplemented by clinical interview and collateral information, to determine if the patient experienced a relapse during the preceding 6 months. The number of previous (ie. pre-manic) mood episodes, and the presence or absence of substance use and other comorbid conditions, was obtained by self-report using a standardized protocol.  Functional outcomes were measured using the MSIF, which was administered to patients at baseline and 6-month and 12-month visits (130). This instrument uses seven-point Likert scales to independently assess functioning in three domains (role position, required level of support, and role performance) in each of three environments (work, educational, and residential), and to provide a summary measure of global functioning. The anchor points for the scales are: 1 ? essentially normal functioning; 2 ?very mild disability, possibly at the low end of the normal range; 3 ? somewhat disabled; 4 ? moderately disabled; 5 ? significantly disabled; 6 ? extremely disabled; 7 ? totally disabled. Unlike some measures of functioning, such as the GAF scale, the MSIF does not include symptoms in its anchor points, and it thus facilitates an assessment of functional outcomes independent of symptomatic status. In addition, since it collects data on functioning across various domains and environments, it is sensitive to contextual factors that influence functioning. The MSIF was designed to assess functioning in the month preceding its administration, and it has been validated for this purpose in 143 patients with BD (140). In 38  addition to assessing functioning during the preceding month, we also administered a modified version of the scale to measure functioning over the preceding 6 months.   3.2.3  Assessment of weight and BMI  As described in Chapter 2, each patient was weighed and had their height measured at enrolment and every 6 months. BMI was calculated using the formula BMI = weight (kg) / height (metres)2. Underweight was defined as BMI < 18.50; normal weight as BMI = 18.50-24.99; overweight as a BMI = 25.00-29.99; and obesity as a BMI ? 30.00. CSWG was defined as gaining ? 7% of baseline weight.  3.2.4  Data analysis and statistics  We carried out statistical analyses using SPSS for Windows, version 16.0 (SPSS Inc, Chicago, IL). Baseline sociodemographic and clinical variables were analyzed using t-tests, ANOVA, and ?2 tests as appropriate. For our primary analyses, we compared patients with CSWG to those without CSWG on the number of days with mood symptoms over 12 months, and 12-month MSIF global functioning scores. For exploratory purposes, we also compared the same outcomes in patients who were overweight or obese at baseline to patients with baseline normal weight, and in patients who were overweight or obese at 12 months to patients with 12-month normal weight. As we previously reported greater weight gain in male than female STOP-EM patients (141), we repeated the analyses for the number of days with mood symptoms and previous-30-day MSIF global functioning scores for males and females with CSWG separately.   Given the paucity of data on the psychiatric impact of weight gain early in BD, we examined a number of secondary outcomes. These included comparing patients with and without CSWG, and with and without baseline and 12-month overweight/obesity, on the number of days with each of depressive symptoms, manic symptoms, and hypomanic symptoms over 12 months; rates of recurrence into any mood episode, depression, mania, and hypomania over 12 months; and mean 12 month YMRS, HAM-D-29, and MADRS scores. We also compared patients with baseline overweight/obesity to those with baseline normal weight on rates of pre-manic 39  depressive and hypomanic episodes. Secondary functional outcomes included comparing patients with and without CSWG, and with and without baseline and 12-month overweight/obesity, on 12 month MSIF subscale scores for each of role position, required support, and role performance. As well, we considered patients with MSIF global scores of 1-3 to have no-to-minimal impairment, and those with scores of 4-7 to have moderate-to-severe impairment, and we compared the proportions of patients with and without CSWG, and with and without baseline and 12-month overweight/obesity, who exhibited at least moderate functional impairment. Finally, we used linear regression to examine predictors of time spent with mood symptoms over 12 months, and of MSIF global functioning scores at 12 months.   Comparisons of time spent with mood symptoms, clinical rating scale scores, MSIF scores, and other continuous variables were performed using ANCOVA. Based on previous research, covariates for time spent with mood symptoms included the age at onset of the mood disorder (including prior depressions and hypomanias), the number of previous depressions and hypomanias, the duration of illness at intake, and the duration of the initial manic episode (106,142-145). For analyses involving MSIF scores, covariates included premorbid socioeconomic status, age at the first mania, duration of the first mania, and the number of days with manic or depressive symptoms in the 6 months preceding assessment (103,104,146). Categorical variables, such as recurrence rates and the proportions of patients with moderate-to-severe impairment, were analyzed using ?2 tests. In any instance where expected cell counts were < 5, we employed Fisher?s exact test. All statistical tests were two-tailed and were carried out using a significance level of ?=0.05. As our analyses were exploratory, we did not correct for multiple comparisons.  We performed linear regression analyses using the forced-entry method. We selected variables for inclusion in the regression models based on previous research. The variables for the number of days spent with mood symptoms included CSWG at 12 months, gender, socioeconomic status, age at onset of mood disorder, polarity of the index episode, number of previous depressions and hypomanias, psychosis with any mood episode, substance dependence or anxiety disorder comorbidity, duration of illness at intake, duration of first mania, baseline YMRS and MADRS scores, and treatment with lithium, divalproex, risperidone, olanzapine, and quetiapine. The 40  predictors we assessed for 12-month MSIF global functioning included CSWG at 12 months, premorbid socioeconomic status, number of previous depressions and hypomanias, number of previous suicide attempts, substance dependence or anxiety disorder comorbidity, duration of illness at intake, age at first mania, duration of first mania, and the number of days with depression, mania, and hypomania during the follow-up period.  We included in our analyses all patients enrolled in STOP-EM who had baseline and 12 month data for weight and MSIF scores, and 12-month data for the number of days with mood symptoms. For the assessment of pre-manic illness characteristics only, we used a slightly larger sample by not excluding those who did not have 12-month weight data.  3.3  Results  3.3.1  Sociodemographic and clinical characteristics  Of the first 66 patients enrolled in STOP-EM, 46 had baseline and 12 month data for weight, the number of days with mood symptoms, and MSIF scores, and were included in our primary analyses. Of the 20 patients with missing data, 7 were lost to follow-up, 6 had not yet reached the 12 month follow-up visit, 4 withdrew consent, and 3 moved out of the study area.   Sociodemographic characteristics of patients with and without CSWG by 12 months are enumerated in table 3.1, and their baseline clinical characteristics are listed in table 3.2. Patients with and without CSWG were well-matched for age, ethnicity, and employment status, and were similar with respect to baseline MSIF global functioning scores, illness course, and comorbid diagnoses. Medication use over 12 months is reported in Table 3.3. Patients with and without CSWG were similar in their use of mood stabilizers, SGAs, combination treatment, and the total number of medications prescribed.   41  3.3.2  Baseline and 12-month weight and CSWG   At baseline, the 46 patients in our sample had a mean weight of 70.6 kg and a mean BMI of 23.8. Twenty-eight percent (N=13) were overweight and 6.5% (N=3) were obese. Over the 12 month follow-up period, 41.3% of patients (N=19) experienced CSWG, with a mean weight gain of 11.2 kg. In contrast, the 58.7% of patients (N=27) who did not experience CSWG lost a mean of 0.5 kg. By 12 months, the mean weight in our sample increased to 74.9 kg, the mean BMI increased to 25.3, and the proportions of patients with overweight and obesity increased to 39.1% (N=18) and 10.9% (N=5), respectively. Although males and females did not differ significantly in the proportions who experienced CSWG (47.6% vs 36.0%; ?2=0.636, df=1, p=NS), the magnitude of the weight gain was greater in males (7.0 kg (SD=8.2) vs 2.0 kg (5.9); F=5.622, df=1, p=0.020), and significantly more males than females were overweight or obese at 12 months (66.7% vs 34.6%; ?2=4.776, df=1, p=0.029).  3.3.3  Pre-manic clinical outcomes  When patients with baseline overweight/obesity (n=18 in the larger sample, not excluding those without 12-month data) were compared to those with baseline normal weight (n=34) on pre-manic illness characteristics, there were no significant differences between them in the duration of the first manic episode, numbers of previous depressive or hypomanic episodes, rates of premorbid ADHD or anxiety disorders, or baseline scores on any clinical rating scales (Table 3.4). However, significantly more overweight/obese patients had a lifetime history of cannabis abuse or dependence (66.7% vs 37.5%, chi-square=3.926, p=0.048), and there was a statistical trend for them to have a greater family history of hypercholesterolemia (22.2% vs 3.2%, ?2=4.472, p=0.057). Similar to previous reports, almost 3 times as many overweight/obese patients made a previous suicide attempt, but this was not a statistically significant difference (16.7% vs 5.9%, ?2=1.575, p=0.327).   When only male patients were included in the analyses, a lifetime history of cannabis abuse or dependence was much more frequent in overweight/obese patients (90.9% vs 28.6%, ?2=9.715, p=0.004), and a history of a previous suicide attempt was over 4 times greater (27.3% vs 6.3%, 42  ?2=2.283, p=0.273). When only female patients were considered, lifetime rates of cannabis abuse or dependence were numerically lower in overweight/obese patients (28.6% vs 44.4%, ?2=0.529, p=0.659), and rates of previous suicide attempts were more similar between overweight/obese patients normal-weight patients (0% vs 5.9%, ?2=0.405, p=1.000).  3.3.4  Prospective 12-month clinical outcomes   Figure 3.1 illustrates the number of days with mood symptoms over 12 months in patients with baseline and 12-month overweight or obesity, and in patients with CSWG. Patients with baseline overweight or obesity had numerically fewer days with depression, mania, or any mood symptoms than those with normal weight, but these differences did not reach statistical significance. Results were similar for patients with 12-month overweight or obesity. Patients who experienced CSWG at 12 months spent 23% more days with depression, and numerically more days with any mood symptoms, than patients without CSWG, but these differences also were not significant.   Fifty percent (N=8) of patients with baseline overweight or obesity experienced a relapse into any mood episode over 12 months, compared to 53.3% (N=16) with normal weight (?2=0.046, df=1, p=NS). The proportions for patients with 12-month overweight or obesity and 12-month normal weight were 52.2% and 54.2% (?2=0.019, df=1, p=NS). Among patients with CSWG, 57.9% experienced a relapse into any mood episode, compared to 48.1% without CSWG (?2=0.425, df=1, p=NS). Relapse rates specifically into depression, mania, and hypomania were similar between all groups. Mean 12-month YMRS, MADRS, and HAM-D-29 scores were low in patients with and without baseline and 12-month overweight or obesity, and in patients with and without CSWG, and no significant differences were detected between groups.   When males and females were analyzed separately, CSWG was not associated with the number of days with depression, mania, or any mood symptoms in either gender (data not shown). In the entire sample, linear regression did not demonstrate an association between the number of days with mood symptoms and CSWG or any of the other variables we assessed.    43  3.3.5  Functional outcomes   At baseline, the mean previous-30-day MSIF global functioning score in our sample was 4.15 (2.37), consistent with moderate disability. Fifty-two percent of patients (N=24) had global functioning scores ? 4, and thus met our definition of moderate-to-severe impairment, while 47.8% (N=22) had no-to-minimal impairment. By 12 months, the mean global functioning score in the entire sample improved to 1.96 (1.12) (t=6.204, df=45, p=0.000), consistent with very mild disability, and the percentage of patients with moderate-to-severe impairment decreased to 8.7% (N=5). Males and females did not differ significantly in 12-month global functioning scores (2.19 (1.37) vs. 1.76 (0.83); F=1.904, df=1, p=NS).   Figure 3.2 illustrates mean 12-month MSIF scores in patients with baseline overweight or obesity vs normal weight. Figure 3.3 shows these results for patients with 12-month overweight or obesity vs normal weight, and Figure 3.4 for patients with and without CSWG. Baseline and 12-month overweight or obesity were not associated with impaired functioning in any of the MSIF domains, whether functioning was assessed over the previous 30 days or the previous 6 months. However, patients with CSWG had significantly greater impairment in global functioning over the previous 30 days than those without CSWG (MSIF score = 2.26 (1.24) vs 1.74 (0.98); F=7.089, df=1, p=0.011), and a non-significant trend suggested impaired role functioning (2.63 (1.42) vs 2.33 (1.36); F=3.396, df=1, p=0.073). Over the previous 6 months, patients with CSWG had significant impairment in both global functioning (MSIF score = 2.42 (1.35) vs 1.93 (1.04); F=6.491, df=1, p=0.015) and role functioning (3.05 (1.51) vs 2.26 (1.06); F=9.627, df=1, p=0.004). Impaired global functioning was independent of the number of days with depressive and manic symptoms over the previous 6 months, as these were entered as covariates in our analyses. Our results did not change when 12-month YMRS and MADRS scores were entered as covariates (p=0.010 for previous-30-day functioning and p=0.036 for previous-6-month functioning), indicating that BMI-related impairment was  independent of mood symptoms at the time of assessment. Finally, the findings were unchanged when the analyses were repeated using baseline MSIF global functioning scores as a covariate (p=0.014 for previous-30-day global functioning and p=0.019 for previous-6-month functioning), 44  suggesting that they were not due to pre-existing differences in functioning between the two groups.   The proportion of patients with baseline overweight or obesity who experienced moderate-to-severe impairment in 12-month global functioning was very similar to that of patients with normal weight (12.5% vs 10.0%; ?2=0.067, df=1, p=NS). Similar results were seen for 12-month overweight/obesity (8.7% vs 13.0%; ?2=0.224, df=1, p=NS). Numerically more patients with CSWG experienced moderate-to-severe impairment than patients without CSWG, but this difference did not reach statistical significance (21.1% vs 3.7%; ?2=3.465, df=1, p=NS).   When males and females were analyzed separately, CSWG was associated with impaired 12-month global functioning in females (MSIF score = 2.11 [0.93] in females with CSWG vs 1.56 [0.73] in females without; F=3.273, df=1, p=0.020) but not in males (2.40 [1.51] vs 2.00 [1.27]; F=1.343, df=1, p=NS). In the entire sample, linear regression demonstrated that CSWG (t=2.079, p=0.049), the number of days with depression in the previous 6 months (t=3.114, p=0.005), and the number of days with mania in the previous 6 months (t=2.900, p=0.008) were significant predictors of 12-month MSIF global functioning scores.   3.4  Discussion  Our results support our hypothesis that CSWG is associated with functional impairment very early in the course of BD. In the 12 months following their first manic episode, patients with CSWG experienced 30% greater global functional impairment than those without CSWG, and particularly had a reduced ability to attain satisfying and productive roles. Although psychosocial functioning is influenced by a multitude of sociodemographic and clinical factors, the relationship between CSWG and functioning was independent of mood symptoms, which have been strongly associated with functional impairment in previous studies (147). It was also unlikely to be related to differences in sociodemographic or clinical variables, illness severity, or differential medication use, as patients with and without CSWG were well-matched on these measures. Finally, linear regression confirmed that CSWG, along with recent depressive and 45  manic symptoms, was significantly associated with functional impairment at 12 months when other predictors of functioning were controlled for.   In contrast to the association between CSWG and functional outcomes, we did not detect a significant difference between patients with and without CSWG in the number of days with mood symptoms or the rates of recurrence into mood episodes. This would appear to contradict the findings of previous reports (16,17), which demonstrated a relationship between obesity and more frequent mood episodes. However, it is important to note that our study had a smaller sample size than the previous reports, and assessed for mood symptoms during one year, a substantially shorter period than the previous studies. Even with our relatively small sample and brief follow-up period, it is worth noting that patients with CSWG spent 23% more days with depression than those without CSWG. Furthermore, the number of days with mood symptoms was highly variable. Forty-three percent of our patients were euthymic for the entire 12-month follow-up period, and among the 57% who experienced mood symptoms, the number of days symptomatic ranged from 11 to 330. The high variability of our data, which is to be expected early in the course of BD, made it extremely difficult to statistically detect even a clinically meaningful difference in the amount of time ill between patients with and  without CSWG. Thus, the lack of association between weight gain and depression might be best viewed as an  indeterminate result, since there is a strong likelihood that we had both inadequate statistical power and too brief a follow-up period to detect an association between CSWG and mood symptoms. Further longitudinal follow-up of our sample and additional patients recruited into STOP-EM will provide a more definitive assessment of the degree to which weight gain is associated with mood symptoms early in the course of BD.  A particular strength of the current study is its prospective design. Previous reports have gathered weight data cross-sectionally or retrospectively, and so have been unable to distinguish between the consequences of weight gain and the consequences of obesity in BD. It is therefore interesting that our results suggest that CSWG following a first manic episode, but not pre-existing or 12-month overweight or obesity, predict functional status. The meaning of this finding is not immediately apparent, since there are essentially no data available regarding the differential effects of obesity versus weight gain. It is clear that substantial weight gain occurs 46  very rapidly following the first manic episode in patients with BD (141), and we can speculate that this rapid weight change may have particularly detrimental effects on psychological and/or physical functioning in our young, previously normal-weight patients. Then, following a period of weight stabilization and adaptation, functioning might improve, even if weight does not return to normal. If this is indeed the case, we would expect functioning in patients with CSWG to eventually ?catch up? to that in patients without CSWG, with the overall effect that CSWG delays rather than prevents functional recovery. A final point worth noting is that the proportion of our patients with obesity was low at baseline and remained low at 12 months, even with the substantial weight gain they experienced. As we combined patients with overweight and obesity for our analyses, if obesity is associated with greater functional impairment than overweight, we had little chance of detecting this. Over the longer term, if further weight gain accrues during the course of BD, higher rates of obesity may lead to more pronounced functional impairment. Further follow-up of our patients and others enrolled in STOP-EM will be needed to address this question.   Our report must be interpreted in light of its limitations. Our sample may not be fully representative of all North American patients with BD, since Canada has a lower rate of obesity than the United States, and Vancouver has the lowest obesity rate of any city in Canada (115,148). As noted, we may have had inadequate statistical power and study duration to detect an association between weight gain and mood symptoms. As well, our naturalistic study design precludes the possibility of making causal inferences regarding CSWG and functioning. Furthermore, the STOP-EM program enrolled patients following treatment of their first manic episode, and we did not have MSIF data to assess pre-manic functional status, which has been shown in some studies to be a predictor of later functioning. We did, however, have retrospective data on premorbid socioeconomic status, which is at least partly reflective of functional status and which was used as a covariate in our analyses of functional outcomes. Finally, we were unable to include in our analyses some possible predictors of functioning, such as cognitive impairment and the presence of personality disorders.  Nonetheless, this report contains several notable findings. STOP-EM is the first program to prospectively measure weight gain and rates of overweight and obesity starting at the time of the 47  initial diagnosis of BD. It has broad inclusion criteria and few exclusion criteria, and the patients in our sample are thus highly reflective of those seen in routine clinical practice. We provided treatment according to current evidence-based clinical practice guidelines, and within this framework clinicians were given the latitude to tailor treatment to the needs of individual patients, making the results of our report generalizable to real-world clinical settings. This report is the first to prospectively measure the association of weight gain with functional outcomes in BD, and the first to assess the relationships between weight gain, mood symptoms, and functional outcomes early in BD. By measuring functioning using the MSIF, which does not include mood symptoms in its anchor points, and by covarying our analyses for mood symptoms and other predictors of functional outcomes, we were able to separate the effect of weight gain from other causes of functional impairment. Our main finding, that CSWG in the 12 months after a first manic episode is associated with impaired global functioning and particularly with the ability to attain satisfying roles, has immediate clinical relevance. It underscores the importance of understanding weight gain as affecting not only patients? physical health, but also their ability to truly recover from BD and to lead full, productive lives. Clinicians should consider the possibility and consequences of weight gain when making even the earliest treatment decisions for patients with BD change.   48  Table 3.1: Baseline sociodemographic characteristics of BD patients with and without CSWG from baseline to 12 months   Patients with CSWG Over 12 Months (N=19) Patients without CSWG Over 12 Months (N=27) P-value        Mean SD Mean SD  Age 23.4 4.4 22.2 3.6 0.319        N % N %  Gender      0.425 male 10 52.6% 11 40.7%  female 9 47.4% 16 59.3%  Ethnicity (*)      Caucasian 15 78.9% 20 76.9% 0.949 Asian 3 15.8% 4 15.4%  Other 1 5.3% 2 7.7%  Employment     0.259 Full time or part-time employment 10 52.6% 9 33.3%  Student 9 47.4% 16 59.3%  Unemployed 0 0% 2 7.4%   * N=45; 1 value missing 49  Table 3.2: Baseline clinical characteristics of BD patients with and without 12-month CSWG     Patients with CSWG Over 12 Months (N=19) Patients without CSWG Over 12 Months (N=27) P-value  Mean SD Mean SD  Age at onset of mood disorder (years) (*) 20.7 5.2 19.8 5.0 0.547 Duration of mood disorder (years) (*) 2.7 3.6 2.6 3.3 0.918 Duration of index mania (days) 70.6 60.3 56.4 45.6 0.368 Baseline YMRS score 4.1 5.7 3.6 6.4 0.781 Baseline MADRS score 6.6 8.2 5.7 7.6 0.684 Baseline HAM-D-29 score 7.6 9.2 6.3 8.6 0.621 Baseline MSIF Global Functioning score 4.4 2.6 4.0 2.3 0.609        N % N %  In remission at enrolment  11 57.9% 18 66.7% 0.544 Previous depressive episode 9 47.4% 17 63.0% 0.293 Previous hypomanic episode 3 15.8% 7 25.9% 0.488 Previous mood episode (depressive or hypomanic) 10 52.6% 19 70.4% 0.220 Previous suicide attempt 1 5.3% 4 14.8% 0.387  50    Patients with CSWG Over 12 Months (N=19) Patients without CSWG Over 12 Months (N=27) P-value Lifetime comorbidity      Anxiety 2 10.5% 3 11.1% 1.000 Alcohol abuse or dependence 2 10.5% 7 25.9% 0.270 Cannabis abuse or dependence 5 26.3% 11 40.7% 0.101 Other drug abuse or dependence 1 5.3% 3 11.1% 0.632   * N=45; 1 value missing  51  Table 3.3: Medication use over 12 months in BD patients with and without 12-month CSWG    Patients with CSWG Over 12 Months (N=19) Patients without CSWG Over 12 Months (N=27) P-value  N % N %  Medication Class      Mood Stabilizer 17 89.5% 26 96.3% 0.561 SGA 17 89.5% 22 81.5% 0.682 Mood Stabilizer + SGA 15 78.9% 22 81.5% 1.000       Total Number of Medications Prescribed Over 12 Months     0.733 Zero 0 0% 1 3.7%  One 3 15.8% 4 14.8%  Two 9 47.4% 11 40.7%  Three 5 26.3% 10 37.0%  Four 2 10.5% 1 3.7%   52  Table 3.4: Pre-manic illness characteristics in BD patients with and without baseline overweight/obesity     Overweight/ Obese (n=18) Normal-weight (N=34)   Mean SD Mean SD P value Current age 21.28 3.82 21.18 3.75 0.927 Age at illness onset 19.33 3.99 19.29 4.54 0.975 Duration of mood illness 2.94 2.98 2.88 3.66 0.951 Number of previous depressions (*) 1.00 1.17 1.24 1.60 0.593 Number of previous hypomanias 0.83 2.12 0.41 1.42 0.396 Number of previous mood episodes 1.78 2.24 1.65 2.16 0.838 Number of previous suicide attempts 0.33 0.97 0.06 0.24 0.122 Duration of index mania (days) 55.17 43.75 64.58 56.04 0.541 Baseline YMRS score 3.44 4.02 4.24 6.64 0.647 Baseline MADRS score 5.94 6.58 6.29 8.68 0.882 Baseline HAM-D score 6.50 8.29 7.12 9.50 0.818 Baseline BPRS score 21.89 4.48 22.59 6.18 0.674 Baseline PANSS score 7.50 1.29 8.06 1.72 0.234 Baseline CGI-S score 2.17 1.25 2.15 1.33 0.959        N % N % P value In remission at enrolment (YMRS ? 8, MADRS ? 8) 11 61.1% 21 61.8% 0.963 Current mood stabilizer 17 94.4% 28 82.4% 0.399 Current antipsychotic 14 77.7% 27 79.4% 1.000 Current combination treatment 14 77.7% 24 70.6% 0.746 Previous depression (*) 10 55.5% 19 55.9% 1.000 Previous hypomania 6 33.3% 6 17.6% 0.300 Previous mood episode 13 72.2% 20 58.8% 0.340 Previous suicide attempt 3 16.7% 2 5.9% 0.327 Lifetime alcohol abuse/dependence (**) 5 27.7% 5 15.6% 0.463 Lifetime cannabis abuse/ dependence (**) 12 66.7% 12 37.5% 0.048 Lifetime drug abuse/dependence (**) 0 0% 4 12.5% 0.283 Lifetime anxiety disorder (**) 1 5.6% 4 12.5% 0.642 Lifetime ADHD (**) 1 5.6% 3 9.4% 1.000 Family History of Diabetes (+) 1 5.6% 1 3.2% 1.000 Family History of Heart Disease (**) 3 16.7% 2 6.5% 0.342 Family History of Hypercholesterolemia 4 22.2% 1 3.2% 0.057  * N=51; 1 patient with missing data ** N=50; 2 patients with missing data + N=49; 3 patients with missing data 53  Figure 3.1 Mean number of days with mood symptoms during 12 months of maintenance treatment in a) BD patients with baseline normal weight vs overweight/obesity, b) patients with 12-month normal weight vs overweight/obesity, and c) patients with and without 12-month CSWG                     a)              b) 010203040506070Any MoodSymptomsDepressive Manic HypomanicPatients WithBaseline NormalWeight Patients WithBaselineOverweight orObesity Days With Mood Symptoms 010203040506070Any MoodSymptomsDepressive Manic HypomanicPati nts With 12-Month NormalWeight Patients With 12-Month Overweightor Obesity Days With Mood Symptoms 54                   c)   Days With Mood Symptoms 010203040506070Any MoodSymptomsDepressive Manic HypomanicPatients With CSWG Patients WithoutCSWG55  Figure 3.2 : Mean 12-month MSIF scores in BD patients with baseline normal weight vs overweight/obesity a) measuring functioning over the previous 30 days, and b) measuring functioning over the previous 6 months                  a)                    b) 00.511.522.533.5Role Position Required Support Role Performance GlobalPatients With BaselineNormal Weight Patients With BaselineOverweight or Obesity00.511.522.53Role Position Required Support Role Performance GlobalPatients With BaselineNormal WeightPatients With BaselineOverweight or Obesity12-Month MSIF Scores (Past 30 Days) 12-Month MSIF Scores (Past 6 Months) 56  Figure 3.3 Mean 12-month MSIF scores in BD patients with 12-month normal weight vs overweight or obesity a) measuring functioning over the previous 30 days, and b) measuring functioning over the previous 6 months                   a)                  b) 00.51.522.53Role Position Required Support Role Performance GlobalPatients With 12-MonthNormal Weight Patients With 12-MonthOverweight or Obesity00.511.522.53Role Position Required Support Role Performance GlobalPatients With 12-MonthNormal Weight Patients With 12-MonthOverweight or Obesity12 Month MSIF Scores (Past 30 Days) 12-Month MSIF Scores (Past 6 Months) 57  Figure 3.4 Mean 12-month MSIF scores in BD patients with and without 12-month CSWG a) measuring functioning over the previous 30 days, and b) measuring functioning over the previous 6 months                  a)               b) 12 Month MSIF Scores (Past-30-Days) 00.511.522.53Role Position Required Support Role Performance GlobalPatients With CSWG Patients Without CSWG12 Month MSIF Scores (Past 6 Months) p = 0.073      p = 0.011  00.511.522.533.5Role Position Required Support Role Performance GlobalPatients With CSWG Patients Without CSWGp = 0.004       p = 0.015 58  4.  The association of overweight/obesity with reduced white matter volume and temporal lobe volume at recovery from a first manic episode  4.1  Introduction  The results presented in Chapter 3 demonstrate that the association between weight gain and poor psychiatric outcomes in BD begins very early in the course of the illness, as STOP-EM patients who experienced CSWG in the 12 months following their first manic episode spent 23% more time with depression and had significantly greater functional impairment than patients without CSWG (149). This chapter and the following two examine the neurobiological correlates of overweight and obesity in BD patients, specifically by looking at BMI-related changes in brain structure (Chapters 4 and 5) and brain chemistry (Chapter 6) present at their enrolment into STOP-EM ? that is, at the time of their first diagnosis of BD. Such correlations, if detected, would suggest a possible mechanism to explain the relationship between elevated BMI and poor clinical outcomes in BD.   Despite the by now well-established association between elevated BMI and time spent with depression, and the increased frequency of serious consequences of depression such as suicidality in obese patients, the neurobiological mechanisms underlying these relationship are currently unknown. In the general population, however, where obesity is also associated with depressive symptoms (89), neuroimaging studies suggest that the brain may be a target organ for obesity-related damage. Compared to normal-weight individuals, people with elevated BMI have significantly decreased total brain volume, and particularly reduced GMV, even in adolescence and young adulthood (76-80). These findings remain true when people with neurological and other medical illnesses are excluded or when the presence of these conditions is controlled for (76,78), suggesting that they are not due simply to medical complications of obesity. Furthermore, obese individuals experience greater brain volume loss over time than those with normal weight (81), and animal models show an association between experimentally-induced weight gain and smaller brain volume (95), suggesting a causal relationship between weight gain and brain volume loss.   59  However, despite the high prevalence of obesity in patients with BD, and its relationship with reduced brain volume in the general population, no neuroimaging studies have compared brain volumes in patients with and without obesity. Here, we present MRI data on the association between BMI with and brain volumes in several large-scale regions of interest ? total grey matter (GM), total white matter (WM), and frontal, parietal, occipital, and temporal lobe volumes - in patients from our first-episode mania program. The same analyses were also carried out in age- and gender-matched healthy subjects to determine the specificity of weight-related changes in BD.   This is the first study to investigate the relationship between weight and brain structure in BD, or in fact any psychiatric illness, and so we did not have previous results to guide us in developing hypotheses. Nonetheless, we undertook these analyses with some general predictions. Based on previous work in healthy samples, we hypothesized that increased BMI would be associated with reduced brain volumes in BD patients; and based on the association of obesity with poor clinical outcomes in BD, we suspected that these would be in brain areas implicated in the pathophysiology of BD. In healthy subjects, in contrast, we predicted weight related brain volume loss in total grey matter and total brain volume, as previously reported.  4.2  Methods  4.2.1  STOP-EM  The BD patients and healthy subjects included here were drawn from the UBC STOP-EM program. The inclusion and exclusion criteria, assessment procedures, and measurement of weight and BMI for STOP-EM were as described in Chapter 2. At the time this report was written, substantially more patients than healthy subjects were enrolled in STOP-EM, and so to balance the groups we also included in our analyses consecutively-enrolled healthy subjects who were recruited for a study of first-episode psychosis (FEP) conducted by two of our collaborators (William G. Honer and Donna J. Lang).  60  4.2.2  MRI protocol and data extraction  All patients and healthy subjects had a cerebral MRI at enrolment. Images were obtained using a Philips Achieva 3T scanner. Patients and healthy subjects were positioned for T2-weighted whole brain transverse sagittal localizers (field of view (FOV) 23.0 cm, matrix = 256 x 195, voxel size = 1.12 x 0.90 mm, TR/TE = 3000/80 ms, flip angle = 90 degrees, slice thickness = 4 mm/1 mm gap, total = 28 slices) and T2-weighted whole brain coronal localizers (FOV = 23 cm, matrix =   256 x 256, voxel size =  1.20 x 0.90 mm, TR/TE = 3000/80 ms, flip angle = 90 deg, T/R head coil, slice thickness = 4 mm/1 mm gap, total = 24 slices). Structural volumetric acquisition included a three-dimensional axial inversion recovery-weighted spoiled gradient recalled sequence (3D SPGR T1-weighted), FOV = 25.6 cm, matrix = 256 x 256, isotropic voxels (1 x 1 x1 mm), autoshim, TR/TE = autoset shortest, T/R head coil, flip angle = 8 deg, and 1 mm thick contiguous 180 slices of the whole brain.   Brain volumes were quantified using FSL version 4.1 (FMRIB Software Library, www.fmrib.ox.ac.uk/fsl/) (150,151). First, the FSL Brain Extraction Tool (BET) v2.1 was used to remove all non-brain signal from the datasets. Then skull-stripped images were segmented into grey matter, white matter, and CSF using FMRIB's Automated Segmentation Tool (FAST 4) based on Montreal Neurological Institute (MNI) 152 tissue probability maps (152) (Figure 4.1). The stripped image sets were registered to MNI 152 standard space with FMRIB's Linear Image Registration Tool (FLIRT) v5.4.2. A global optimization procedure (12 degrees of freedom, correlation ratio cost function) was performed to provide robust affine registration of brain images to MNI template (152). The final subregional volumes were calculated using in-house scripts, and subregional lobar volumes were determined by mapping the MNI structural atlas onto the processed datasets to calculate left and right frontal, parietal, occipital, and temporal lobe volumes (Figure 4.1).   4.2.3  Data analysis and statistics  We included in our analysis all patients and healthy subjects who had baseline data for BMI and brain volumes at enrolment into STOP-EM. Since patients were early in the course of their 61  illnesses and were treated with medications for a relatively short time period, rates of overweight and obesity were low. They were thus combined so that patients with overweight or obesity were compared to those with normal weight, and similarly for healthy subjects. We normalized brain volumes using total cranial volume as measured by FSL 4.1 to control for differences in head size (153), to minimize artefacts introduced during the MRI scanning process (154), and in keeping with previous studies of BMI and brain volumes (76,78). We carried out statistical analyses using PASW for Windows, version 18.0 (SPSS Inc, Chicago, IL). All statistical tests were two-tailed and were carried out using a significance level of ?=0.05. Our analyses were exploratory and we did not correct for multiple comparisons.  We examined sociodemographic and clinical variables using t-tests, ?2 tests, and Fisher?s exact test as appropriate. Since numerous sociodemographic, clinical, and treatment factors influence brain volumes in BD (155-161), for our primary analysis we used standard linear regression analyses to determine whether BMI predicted normalized total brain volume (nTBV), normalized GMV (nGMV), normalized WM volume (nWMV), and normalized frontal, parietal, occipital, and temporal lobe volumes (including GM and WM) in patients when other mediators of brain volume were taken into account. We selected variables for entry into our regression models based on previous research (81,155-162). They included BMI; fasting serum cholesterol, HDL, LDL, triglycerides, and glucose; age; gender; the number of previous depressions and hypomanias; current YMRS and MADRS scores; psychosis during the index mania; alcohol, cannabis, and drug dependence; and treatment with lithium, divalproex, risperidone, olanzapine, and quetiapine. We first calculated Pearson correlation coefficients to determine which variables were significantly associated with nTBV, and those with significant correlations were then entered into the regression models (these are listed in the Results section). We conducted similar analyses in healthy subjects, using BMI, age, and gender as predictors.   As secondary analyses, we repeated our regression models with absolute brain volumes as the dependent variables, since there is a lack of consensus regarding whether normalized or absolute brain volumes should be reported in psychiatric studies (163). Cranial volume was entered as an additional predictor in these analyses. We also compared mean normalized and absolute TBV, GMV, WMV, and frontal, parietal, occipital, and temporal lobe volumes in patients with 62  overweight/obesity to patients with normal weight, and in healthy subjects with overweight/obesity to those with normal weight, using ANCOVA with the same covariates as in the above analyses. Finally, we directly assessed interactions between diagnosis and BMI in the brain regions of interest by using factorial multiple ANCOVA (MANCOVA) with diagnosis (patient vs. healthy subject) and BMI category (normal weight vs. overweight/obese) as main factors and age, gender, and cranial volume as covariates.  4.3  Results  4.3.1  Patients with BD and healthy subjects  Sociodemographic and clinical characteristics. Of the initial 71 patients and 30 healthy subjects recruited into STOP-EM, 57 patients and 30 healthy subjects had baseline data for BMI and brain volumes and were included in our analysis. Two patients with missing data declined to be weighed, and 12 refused an MRI. All 33 of the healthy subjects recruited into the FEP study had data for BMI and brain volumes, but eight did not meet inclusion criteria for STOP-EM healthy controls as they were older than 35 years (N=3) or had histories of serious head injuries (N=5), and so were excluded. This left us with a total of 57 patients and 55 healthy subjects. Sociodemographic and clinical characteristics of the patients and healthy subjects are listed in Table 4.1.   BMI, overweight, and obesity. Patients and healthy subjects had similar mean BMIs (patients: 24.1, SD=3.9; healthy subjects: 24.0 (3.9); p=NS). They did not differ with respect to the proportions with normal weight (64.9% vs 69.1%, respectively), overweight (26.3% vs 23.6%), or obesity (8.8% vs 7.2%) (p=NS).   TBV, GMV, and WMV. Patients and healthy subjects had identical mean nTBV (0.693 for both, p=NS) and similar mean nGMV (0.5% larger in patients, 0.391 vs 0.389, F=1.350, p=NS) and nWMV (0.7% smaller in patients, 0.302 vs 0.304, F=.448, p=NS). Mean absolute TBV, GMV, and WMV also did not differ between patients and healthy subjects (TBV 3.9% smaller in 63  patients, 1298 mL vs 1348 mL, F=0.001, p=NS; GMV 3.3% smaller in patients 732 mL vs 756 mL, F=0.421, p=NS; WMV 4.6% smaller in patients, 565 mL vs 591 mL, F=0.530, p=NS).   4.3.2  BMI and brain volumes  Linear regression models. In patients with BD, BMI, age, gender, and lithium use were significantly correlated with nTBV and so were entered as predictors in our all of our linear regression models. Increased BMI in patients significantly predicted decreased nWMV (?=-.252, t=-2.003, p=.050) and decreased normalized temporal lobe volume (?=-.376, t=-3.436, p=.001), and a non-significant trend suggested that it predicted decreased nTBV (?=-.241, t=-1.984, p=.053) (Table 4.2a, Figure 4.2). In healthy subjects, BMI, age, and gender were entered as predictors in our regression models, and increased BMI predicted decreased nTBV (?=-.287, t=-2.436, p=.018) and decreased nGMV (?=-.363, t=-3.132, p=.003) (Table 4.2a, Figure 4.2).   Regression analyses of absolute brain volumes produced largely similar results (Table 4.2b), with the exceptions that the association of increased BMI with decreased TBV now reached significance in patients (?=-.094, t=-2.021, p=.049), but was no longer significant in healthy subjects (?=-.08, t=-1.869, p=068).  Mean brain volumes. Overweight/obese patients had significantly smaller mean nWMV, normalized temporal lobe volume, and normalized occipital lobe volume than normal-weight patients (Table 4.3a, Figures 4.3 and 4.4). The effect sizes (Cohen?s d) for nWMV and normalized occipital volume were medium (0.51 and 0.66), while the effect size for normalized temporal lobe volume was large (0.97). The reduction in temporal lobe volume in overweight/obese patients was almost twice as great as that in the other brain lobes. In contrast, overweight/obese and normal-weight healthy subjects did not differ significantly with respect to any of the normalized brain volumes we measured (Table 4.3a, Figures 4.3 and 4.4).   For absolute brain volumes, the results and significance levels for patients with BD and healthy subjects were similar to those for normalized volumes (Table 4.3b).   64  Factorial MANCOVA. There was a significant interaction between diagnosis and BMI for temporal lobe volume (F=4.122, df=1, p=.045). The interaction between diagnosis and BMI did not reach statistical significance for GMV (F=.001, df=1, p=NS), WMV (F=2.640, df=1, p=NS), or any lobar volumes. No main effect of diagnosis was detected for any brain volume, and a main effect for BMI was detected only for temporal lobe volume (F=3.920, df=1, p=.050).  4.4  Discussion  This is the first investigation to demonstrate a relationship between elevated BMI and reduced brain volumes in BD, or any psychiatric illness. As predicted, increased BMI in BD patients was associated with decreased brain volumes in regions relevant to BD, specifically reduced WMV and temporal lobe volume, immediately following the first manic episode, or essentially at the first diagnosis of BD. These brain regions were different from those affected in overweight/obese healthy subjects, suggesting that these weight-related structural brain changes are unique to BD.   Our results are particularly noteworthy when placed in the context of a recent meta-analysis of 11 MRI studies in first-episode mania patients, which found reduced WMV, but not GMV, in patients compared to healthy subjects (163), and a qualitative review which reported that decreased temporal lobe volume is one of the most consistent findings in BD (164). These reports provide specific evidence that the BMI-related changes we observed in patients were due to patients having volume reductions in brain areas of known vulnerability in early BD. Furthermore, when taken together with previous MRI studies reporting that reduced brain volume is a poor prognostic factor in BD (165) and other psychiatric illnesses including major depressive disorder (166) and schizophrenia (167), our findings imply a possible biological mechanism underlying the association of obesity with a more severe illness course.   We were able to detect associations between BMI and brain volumes despite the numerous sociodemographic, clinical, and treatment variables that affect brain morphology in BD (155-161). For example, in our patients there were strong correlations between nTBV and age (r=-.304, p=.02), gender (r=-.327, p=.01), and lithium use (r=-.280, p=.035) which were similar in magnitude to the correlation between nTBV and BMI (r=-.334, p=.01). Similarly, in our healthy 65  subjects there were correlations between nTBV and age and gender, in keeping with previous reports (155,168). This underscores the importance of using linear regression to control for the effects of confounding variables on brain volumes. In fact, in our healthy subjects, we detected an association between increased BMI and decreased nTBV and nGMV only in our regression models, and not when comparing mean brain volumes between overweight/obese and normal-weight subjects. As well, our factorial MANCOVA was able to detect a significant interaction between diagnosis and BMI only for temporal lobe volume. We note that both our MANCOVA and our comparisons of mean brain volumes had reduced statistical power as a result of dichotomizing BMI into ?normal-weight? and ?overweight/obese?, compared to our regressions, which used the continuous variable of BMI as a predictor; and that our MANCOVA was unable to control for lithium use, which was a factor only in patients. For these reasons, we feel that linear regression was most appropriate for our primary analysis, particularly given that most previous studies investigating the effect of BMI on brain volumes have also employed linear regression (76,77,79,80,169,170).  However, it is important to bear in mind that our naturalistic study design precludes us from making causal inferences with respect to overweight/obesity and brain volumes. Thus, we cannot distinguish between the possibility that elevated BMI causes reduced brain volumes, and the alternate possibility that there is a subtype of BD characterized by both smaller brain volumes and increased BMI. Supporting the first hypothesis are intriguing data from animal studies demonstrating that mice with experimentally-induced weight gain have smaller brain volumes than mice without weight gain (95), and that rhesus monkeys randomly assigned to a caloric restriction/weight loss diet show reduced brain atrophy compared to monkeys with diet-as-usual and no weight loss (96). Furthermore, there are plausible biological mechanisms to explain how obesity might cause brain volume loss, such as weight-related neuroinflammation and the impact of obesity-related changes in adipokine levels on brain structure and function. In support of the second hypothesis, on the other hand, a recent study reported that a common allele of the fat mass and obesity-associated (FTO) gene, an established risk factor for obesity, was independently linked with frontal and temporal lobe volume reductions (171), suggesting that third variables such as genetic polymorphisms might account for both elevated BMI and reduced brain volumes. If obesity does in fact cause brain volume loss, we expect that overweight/obese 66  STOP-EM patients will experience further volume loss over time, that patients who gain weight and become overweight/obese will experience new-onset volume loss, and that weight loss will be associated with increased brain volumes. Longitudinal follow-up of our sample and additional patients enrolled in STOP-EM will clarify this point.  We also cannot definitively exclude that confounding variables that were present in patients but not healthy subjects contributed to the differences we observed between patients and healthy subjects in BMI-related brain volume reductions. The possible impact of medications deserves particular note. Lithium, for example, has consistently been associated with increased GMV in patients with BD (172,173). While lithium use was unlikely to have confounded our findings when overweight/obese patients were compared to normal weight patients, since we entered it as a predictor in our regression models, we cannot discount the possibility that our finding of BMI-related GMV reduction in healthy subjects but not in patients was confounded by lithium use in the patients. Some but not all studies suggest that divalproex and SGAs also lead to increased GMV (161,174). However, we did not find an association between these medications and nGMV, and so it is unlikely that they confounded our findings within patients, but again we cannot be certain that they did not play a role in the differential findings between patients and healthy subjects. The effect of medications on WMV in BD has received less attention than their impact on GMV, but the available evidence suggests that they have little impact (160,161), and they were thus unlikely to have affected differences in nWMV reduction between patients and healthy subjects.  Additional limitations of this report include the fact that we enrolled patients into STOP-EM following treatment of their first manic episode, and we did not have data regarding pre-manic BMI or brain volumes. Also, since rates of overweight and obesity we low, we combined them in both patients and healthy subjects for our analyses, and we thus could not examine their effects separately. Nonetheless, this report is noteworthy in several respects. STOP-EM is the first program to measure the association between BMI and brain volume in BD, or any psychiatric illness. Our main finding, that overweight and obesity are associated with unique reductions in brain volumes at the outset of BD, suggests a neurobiological rationale for the association between obesity and poor clinical outcomes. Furthermore, it underscores the importance of 67  understanding obesity as affecting not just the physical health of patients with BD, but also as potentially altering the neurobiological basis of the illness. Additional studies are clearly warranted to confirm our findings, and to determine whether similar relationships exist between BMI and brain volumes in other psychiatric illnesses. 68  Figure 4.1: Segmentation of MRI image into a) grey matter, white matter, and cerebrospinal fluid (CSF) and b) frontal, parietal, occipital, and temporal lobes, using FSL 4.1   (a)                    b)                    69  Table 4.1: Baseline sociodemographic and clinical characteristics of BD patients and healthy subjects taking part in a region-of-interest MRI study  Patients with BD (N=57) Healthy Subjects (N=55)  Overweight or Obese (N=20) Normal Weight (N=37)  Overweight or Obese (N=17) Normal Weight (N=38)          Mean (SD) Mean (SD) P-value Mean (SD) Mean (SD) P-value Age 23.8 (4.5) Range: 17-32 22.2 (4.4) Range: 16-34 .202 22.0 (3.8) Range: 18-29 22.3 (3.5) Range: 18-32 .744  % (N) % (N)  % (N) % (N)  Gender    .647   .155 male 15.8% (9)  33.3% (19)  21.8% (12) 34.5% (19)  female 19.3% (11) 31.6% (18)  9.1% (5) 34.5% (19)  Ethnicity   .512   .118 Caucasian 29.8% (17) 49.1% (28)  23.6% (13) 43.6% (24)  Asian 3.5% (2) 14.0% (8)  3.6% (2) 23.6% (13)  Other 1.8% (1) 1.8% (1)  3.6% (2) 1.8% (1)          Mean (SD) Mean (SD) P-value    Baseline YMRS score 3.1 (4.1) 3.7 (6.5) .684    Baseline MADRS score 5.4 (7.3) 6.2 (8.4) .708    Baseline BPRS score * 21.8 (4.1) 22.7 (6.1) .566    Baseline PANSS score 7.5 (1.2) 7.8 (1.7) .399    Duration of manic/mixed episode (days) * 55.3 (43.2) 64.6 (55.8) .529    Duration of mood disorder (years) 2.4 (3.0) 3.1 (4.7) .562     % (N) % (N) P-value    Previous depressive episode 50.0% (10) 54.1% (20) .770    70   * N=55; 2 values missing ** N=56; 1 value missing   Patients with BD (N=57) Healthy Subjects (N=55)  Overweight or Obese (N=20) Normal Weight (N=37)  Overweight or Obese (N=17) Normal Weight (N=38)          % (N) % (N) P-value    Lifetime comorbidity       Anxiety ** 5.0% (1) 11.1% (4) .645    Alcohol abuse or dependence ** 25.0% (5) 11.1% (4) .256    Cannabis abuse or dependence ** 40.0% (8) 30.6% (11) .474    Other drug abuse or dependence ** 0% (0) 13.9% (5) .148    Pharmacotherapy for index mania       Lithium 30.0% (6) 48.6% (18) .174    Divalproex 65.0% (13) 37.8% (14) .050    Risperidone 35.0% (7) 40.5% (15) .682    Olanzapine 20.0% (4) 18.9% (7) 1.000    Quetiapine 20.0% (4) 21.6% (8) 1.000    Mood stabilizer monotherapy 20.0% (4) 10.8% (4) .432    Antipsychotic monotherapy 0% (0) 8.1% (3) .545    Mood stabilizer plus antipsychotic 75% (15) 73.0% (27) .868    No medication  5.0% (1) 8.1% (3) 1.000    71  Table 4.2: Associations between baseline BMI and a) normalized and b) absolute brain volumes, detected with linear regression analyses, in BD patients and healthy subjects     (a)  Patients with BD Healthy Subjects Normalized Brain Volume Beta p-value R2 for model Normalized Brain Volume Beta p-value R2 for model TBV   BMI   Age   Gender   Lithium  -.241 -.185 -.268 -.223  .053 .138 .030 .067 Without BMI: .233 With BMI: .287 TBV   BMI   Age   Gender   -.287 -.125 -.517  .018 .288 .000 Without BMI: .238 With BMI: .318 GMV   BMI   Age   Gender   Lithium  -.115 -.436 -.094 -.198  .342 .001 .429 .097 Without BMI: .296 With BMI: .309 GMV   BMI   Age   Gender   -.363 -.387 -.324  .003 .001 .007 Without BMI: .214 With BMI: .341 WMV   BMI   Age   Gender   Lithium  -.252 .359 -.333 -.086  .050 .007 .010 .487 Without BMI: .176 With BMI: .235 WMV   BMI   Age   Gender   -.083 .176 -.454  .515 .164 .001 Without BMI: .213 With BMI: .219 Frontal lobe    BMI   Age   Gender   Lithium  -.190 -.178 -.265 -.155  .142 .172 .040 .221 Without BMI: .184 With BMI: .217 Frontal lobe    BMI   Age   Gender  -.198 .021 -.461  .122 .866 .001 Without BMI: .181 With BMI: .219 Parietal lobe    BMI   Age   Gender   Lithium  -.208 -.170 -.200 -.144  .116 .202 .128 .264 Without BMI: .141 With BMI: .181 Parietal lobe    BMI   Age   Gender  -.219 -.008 -.361  .102 .954 .008 Without BMI: .105 With BMI: .151 Occipital lobe    BMI   Age   Gender   Lithium  -.244 -.043 -.265 -.227  .060 .740 .040 .074 Without BMI: .167 With BMI: .222 Occipital lobe    BMI   Age   Gender  -.196 .016 -.522  .144 .895 .000 Without BMI: .237 With BMI: .274 Temporal lobe    BMI   Age   Gender   Lithium  -.376 -.246 -.339 -.068  .001 .030 .003 .529 Without BMI: .291 With BMI: .422 Temporal lobe    BMI   Age   Gender  -.188 -.330 -.433  .118 .007 .001 Without BMI: .282 With BMI: .316  72  (b)  Patients with BD Healthy Subjects Brain Volume Beta p-value R2 for model Brain Volume Beta p-value R2 for model TBV   BMI   Age   Gender   Lithium   ICV  -.094 -.070 -.094 -.086 .902  .049 .138 .076 .074 .000 Without BMI: .890 With BMI: .899  TBV   BMI   Age   Gender   ICV   -.088 -.044 -.250 .792   .068 .316 .000 .000  Without BMI: .898 With BMI: .905  GMV   BMI   Age   Gender   Lithium   ICV  -.068 -.236 -.069 -.102 .812  .311 .001 .361 .136 .000 Without BMI: .788 With BMI: .792  GMV   BMI   Age   Gender   ICV   -.146 -.168 -.187 .815   .009 .001 .006 .000  Without BMI: .859 With BMI: .877  WMV   BMI   Age   Gender   Lithium   ICV  -.112 .148 -.113 -.053 .881  .047 .010 .075 .350 .000 Without BMI: .845 With BMI: .857  WMV   BMI   Age   Gender   ICV   -.014 .098 -.293 .680   .845 .140 .001 .000  Without BMI: .791 With BMI: .791  Frontal lobe    BMI   Age   Gender   Lithium   ICV  -.087 -.081 -.086 -.086 .884  .113 .143 .166 .127 .000 Without BMI: .854 With BMI: .861  Frontal lobe    BMI   Age   Gender   ICV   -.095 .017 -.206 .795   .135 .779 .009 .000  Without BMI: .824 With BMI: .832  Parietal lobe    BMI   Age   Gender   Lithium   ICV  -.104 -.085 -.071 -.083 .861  .108 .191 .321 .206 .000 Without BMI: .800 With BMI: .810  Parietal lobe    BMI   Age   Gender   ICV   -.065 .000 -.254 .746   .304 .996 .002 .000  Without BMI: .825 With BMI: .829  Occipital lobe    BMI   Age   Gender   Lithium   ICV  -.095 -.018 -.093 -.088 .906  .057 .719 .095 .082 .000 Without BMI: .880 With BMI: .889  Occipital lobe    BMI   Age   Gender   ICV   -.075 .010 -.232 .793   .173 .197 .001 .000  Without BMI: .870 With BMI: .875  Temporal lobe    BMI   Age   Gender   Lithium   ICV  -.155 -.099 -.132 -.027 .860  .001 .034 .012 .556 .000 Without BMI: .882 With BMI: .904  Temporal lobe    BMI   Age   Gender   ICV   -.059 -.143 -.254 .741   .292 .008 .000 .000  Without BMI: .865 With BMI: .868  73  Figure 4.2: Relationship between BMI and a) normalized TBV in BD patients, b) normalized TBV in healthy subjects, c) normalized GMV in patients, d) normalized GMV in healthy subjects, e) normalized WMV in patients, f) normalized WMV in healthy subjects, g) normalized temporal lobe volume in patients, and h) normalized temporal lobe volume in healthy subjects      a)   b) BMI Normalized TBV BMI Normalized TBV 74    c)     d) BMI Normalized GMV BMI Normalized GMV 75    e)    f) BMI Normalized WMV BMI Normalized WMV 76    g)     h) BMI Normalized temporal lobe volume BMI Normalized temporal lobe volume 77    Table 4.3: Mean a) normalized and b) absolute brain volumes (mL) in overweight/obese and normal-weight BD patients and healthy subjects   (a)  (b)   Patients with BD Healthy Subjects Overweight/Obese Normal Weight p-value Overweight/Obese Normal Weight p-value Mean SD Mean SD Mean SD Mean SD nTBV .68452 .02041 .69831 .02476 NS .69258 .02755 .69369 .02148 NS nGM .38687 .02207 .39400 .02055 NS .38646 .01506 .39045 .01468 NS nWM .29785 .01113 .30431 .01407 .030 .30612 .01924 .30324 .01444 NS nFrontal .14477 .00540 .14757 .00694 NS .14665 .00796 .14570 .00606 NS nParietal .09778 .00388 .09959 .00515 NS .10004 .00466 .09984 .00384 NS nOccipital .06461 .00204 .06605 .00232 .037 .06554 .00308 .06535 .00239 NS nTemporal .08080 .00264 .08361 .00311 .002 .08372 .00355 .08343 .00347 NS  Patients with BD Healthy Subjects  Overweight/Obese Normal Weight p-value Overweight/Obese Normal Weight p-value  Mean SD Mean SD Mean SD Mean SD TBV 1272.1 92.3 1311.3 134.7 NS 1403.0 119.3 1323.1 114.4 NS GM 717.8 52.0 739.9 81.4 NS 782.3 59.2 744.9 67.8 NS WM 554.3 52.1 571.4 60.7 .040 620.7 66.5 578.3 53.0 NS Frontal 269.1 21.9 277.3 30.9 NS 297.3 29.9 278.0 26.1 NS Parietal 181.7 13.5 187.1 21.5 NS 202.6 17.8 190.3 15.8 NS Occipital 120.2 10.0 124.0 12.2 .040 132.8 12.5 124.7 11.5 NS Temporal 150.2 11.3 157.0 16.6 .002 169.6 14.6 159.1 14.5 NS r=-.129 p=.358 78  Figure 4.3: Mean normalized grey and white matter volumes in a) patients with BD and b) healthy subjects       (a)    b) 1.8% less p = NS Normalized Brain Volumes 2.1% less p = .030 0.9% more P = NS 1.0% less p = NS Grey Matter White Matter Grey Matter White Matter Normalized Brain Volumes Normal-weight Overweight/obese 79   Figure 4.4: Mean normalized frontal, parietal, occipital, and temporal lobe volumes in a) patients with BD and b) healthy subjects    (a)    b) 0.3% more p = NS 0.3% more p = NS 0.7% more p = NS 1.9% less p = NS 1.8% less p = NS 2.2% less p = .037 3.4% less p = .002 0.2% more p = NS Normalized Brain Volumes Normalized Brain Volumes Frontal lobes Parietal lobes Occipital lobes Temporal lobes Frontal lobes Parietal lobes Occipital lobes Temporal lobes Normal-weight Overweight/obese 80  5.  The association of overweight/obesity with reduced regional GM and WM volumes in frontal, temporal, and subcortical limbic brain structures at recovery from a first manic episode  5.1  Introduction  Chapter 4 described the findings from a region-of-interest MRI study that examined the relationship between BMI and large-scale brain volumes in the STOP-EM sample. This was the first study to examine neurobiological correlates of obesity in BD, or any psychiatric illness. Most importantly, it generated a specific testable hypothesis: that elevated BMI is associated with unique brain changes early in BD, such that the neuropathology of the illness is exacerbated in overweight/obese patients.   To test this hypothesis, and to further elucidate BMI-related brain changes in BD, we conducted, and present here, a voxel-based morphometry (VBM) analysis of regional GM and WM volumes in our first-episode mania sample. In contrast to a region-of-interest approach, which analyses differences in pre-specified brain regions, VBM is an exploratory MRI analysis method which permits assessment of regional GM and WM volume differences across the entire brain, independent of brain size and shape, while statistically correcting for multiple comparisons. Current models of BD propose that it arises from structural and functional alterations in frontal, temporal, and subcortical brain areas that are involved in generating and modulating emotional responses (175). This argument is buttressed by meta-analyses of MRI studies, which show volume reductions primarily in these areas (176,177). Notwithstanding the exploratory nature of VBM, we undertook the current analyses with an a priori hypothesis: that elevated BMI would be associated with reduced local GM and WM volumes in frontal, temporal, and subcortical emotion-generating and -modulating brain areas in patients, but not healthy subjects.   81  5.2  Methods   5.2.1  STOP-EM  The BD patients and healthy subjects included here were drawn from the UBC STOP-EM program, supplemented, as in Chapter 4, by consecutively-enrolled healthy subjects recruited for a study of FEP. The inclusion and exclusion criteria, assessment procedures, and measurement of weight and BMI for STOP-EM were as described in detail in Chapter 2.   5.2.2  MRI protocol and data extraction  As described in Chapter 4, T1-weighted MR images were acquired with a Philips Achieva 3T scanner (Amsterdam, the Netherlands), using a three-dimensional axial inversion recovery-weighted spoiled gradient recalled sequence and the following parameters: FOV = 25.6 cm, matrix = 256 ? 256, isotropic voxels (1 ? 1 ? 1 mm3), autoshim, TR/TE = autoset shortest, T/R head coil, flip angle = 8 degrees, and 1 mm thick contiguous 180 slices of the whole brain.  5.2.3  Image preprocessing and analysis  Data were processed using SPM8 software (Wellcome Department of Imaging Neuroscience, London, UK) with VBM implemented in the VBM8 toolbox (http://dbm.neuro.uni-jena.de/vbm.html). The Diffeomorphic Anatomical Registration Through Exponential Lie Algebra (DARTEL) algorithm was used for improved accuracy of registration (178). This non-linear warping technique minimizes between-subject structural variations and achieves accurate realignment of small inner brain structures (179,180).   T1 images were bias-corrected, segmented into GM, WM, and CSF, and affine-transformed to MNI space. A customized DARTEL template was created using the affine-registered GM and WM tissue segments of all subjects, and non-linear deformation was performed with the high-dimensional DARTEL algorithm using this template. To examine absolute tissue volumes corrected for individual brain sizes, images were modulated by multiplying by the non-linear 82  components derived from the normalization matrix, so as to preserve actual GM and WM volumes locally. The modulated images had a voxel size of 1.5 ? 1.5 ? 1.5 mm3, and were smoothed using an 8mm full width at half maximum (FWHM) Gaussian kernel.   5.2.4  Statistical methods  We examined sociodemographic and clinical variables with t-tests, ?2 tests, or Fisher?s exact test as appropriate, employing a two-tailed significance level of ?=0.05, using PASW for Windows, version 18.0 (SPSS Inc, Chicago, IL).   For our primary analysis, we used multiple linear regression models to examine the association between BMI and regional GM and WM volumes in patients and healthy subjects separately. This allowed us to use the continuous variable of BMI as a predictor, which our previous report found to be more strongly associated with brain volumes than the dichotomous variable of BMI category (overweight/obese vs normal weight), and to control for lithium use in patients, which has been positively associated with brain volumes in BD. For secondary analyses, we directly compared the impact of overweight/obesity on brain volumes in patients and healthy subjects using a factorial model with BMI category and diagnosis as factors, and a BMI-diagnosis interaction term; and used t-tests to compare brain volumes in overweight/obese patients vs. overweight/obese healthy subjects, and normal-weight patients vs. normal-weight healthy subjects. Age and gender were entered as covariates in all primary and secondary analyses.  For the inclusion of voxels in a cluster, we employed a significance level of p < .005 uncorrected, with a spatial extent threshold of 500 contiguous voxels. To ensure the validity of cluster-level statistics, we applied a nonisotropic smoothness correction (181). We assessed cluster significance using a stringent cluster-level threshold of p < .05, corrected for family-wise error (FWE) to account for multiple comparisons over the entire brain. We also assessed significance at the voxel level, utilizing a voxel-level threshold of p < .05, FWE-corrected. We assigned anatomic labels to significant GM and WM clusters using WFU Pickatlas (182), and identified the most likely WM tracts crossing WM clusters using the JHU White Matter 83  Tractography Atlas and the International Consortium for Brain Mapping DTI-81 Atlas (183,184).   5.3  Results  5.3.1  Patients with BD and healthy subjects  The sample for the current analyses was the same as that in Chapter 4, comprising 57 patients and 55 healthy subjects. Their sociodemographic and clinical characteristics are listed in Table 5.1, and their weight data in Table 5.2. Patients were euthymic, with very low YMRS and MADRS scores. Patients and healthy subjects had similar BMI distributions: approximately two-thirds of each group were normal-weight, one-quarter were overweight, and < 10% were obese.  5.3.2  BMI and regional GM and WM volumes   Regression models. Increased BMI in patients was significantly associated with one cluster of reduced GMV and two large clusters of reduced WMV, primarily involving frontal, temporal, and subcortical brain structures (Figure 5.1, Table 5.3). The GM cluster was 2600 voxels in size and comprised portions of the right superior, middle and inferior temporal gyri and right uncus. The larger WM cluster (11,476 voxels) was in the right frontal and temporal lobes and right subcortical WM, including the superior and inferior frontal gyri, anterior and posterior cingulate gyri, subcallosal gyrus, middle temporal gyrus, parahippocampal gyrus, fusiform gyrus, frontal and temporal sub-gyral areas, and deep subcortical areas. WM tracts traversing this cluster included the superior and inferior longitudinal fasciculi, inferior fronto-occipital fasciculus, anterior and posterior limbs of the internal capsule, anterior thalamic radiation, external capsule, and fornix. The smaller WM cluster (3906 voxels) was in the left temporal lobe and subcortical WM, including the middle temporal gyrus, fusiform gyrus, temporal subgyral areas, and deep subcortical areas, and was crossed by the superior and inferior longitudinal fasciculi, inferior fronto-occipital fasciculus, and anterior thalamic radiation.   84  BMI-related volume reductions in in other brain areas were minimal, consisting of small WM reductions in the right middle occipital gyrus and right occipital subgyral area which were traversed by the occipital portion of the inferior longitudinal fasciculus. In contrast to patients, increased BMI in healthy subjects was significantly associated with only a single cluster of reduced GMV mainly in the bilateral occipital lobes, and no significant WMV reductions (Figure 5.1, Table 5.3). The GM cluster (4200 voxels) included parts of the bilateral cunei, bilateral lingual gyri, right precuneus, and a small portion of the left parahippocampal gyrus. When the voxel-level significance threshold was applied to our data, two volume reductions in patients remained significant: the GM reduction in the right inferior temporal gyrus, and the WM reduction in the left temporal sub-gyral region. There were no significant BMI-related GMV or WMV increases in patients or healthy subjects.   Factorial models. There was a significant effect of diagnosis on GM, with patients having one cluster of reduced volume in the left temporal lobe, and WM, with patients having three clusters of volume reduction, one in the rostral left temporal lobe, the second in the caudal left temporal lobe that also included portions of the left parietal and occipital lobes, and the third in the right temporal lobe and right subcortical WM (Figure 5.2, Table 5.4). There was no effect of BMI on any regional GM or WM volumes. However, there was a significant BMI-diagnosis interaction, indicating a stronger association between increased BMI and reduced brain volume in patients, in two WM clusters that encompassed widespread frontal, temporal, and subcortical areas, and small parietal and occipital regions (Figure 5.2, Table 5.4). The affected areas closely mirrored the BMI-related WM reductions detected in patients by our regression model.   T-tests. Overweight/obese patients, compared to overweight/obese healthy subjects, had extensive WMV reductions in two clusters involving areas similar to those identified in our regression and factorial models (Figure 5.3, Table 5.5). There were no GMV reductions in overweight/obese patients compared to overweight/obese healthy subjects. Normal-weight patients did not have any GMV or WMV reductions compared to normal-weight healthy subjects.  85  5.4  Discussion  This report constitutes the first hypothesis-driven test of, and supporting evidence for, our theory that elevated BMI is associated with unique structural brain changes in BD, primarily affecting areas believed to be vulnerable in the illness. Our results demonstrate that patients with increased BMI had GMV and WMV reductions in frontal, temporal, and subcortical limbic areas implicated in the pathophysiology of BD. In contrast, age- and gender-matched healthy subjects with a similar BMI distribution had weight-related volume reductions in bilateral occipital GM areas only. Our results thus strongly support our hypothesis, and provide additional evidence of a neurobiological rationale for the well-known link between obesity and poor psychiatric outcomes in BD.  There is little doubt that the BMI-related volume reductions in our patients were in brain areas relevant to BD. Current models of the illness posit that it results from impaired prefrontal modulation of overactive temporal and subcortical limbic structures, and numerous neuroimaging and post-mortem brain investigations converge in showing altered size and activity of, and connectivity between, these brain areas (175). It is thus noteworthy that the cluster of weight-related GM reduction detected in our regression analysis was in the right temporal cortex, an area which plays an important role in emotional processing (185). Moreover, it overlapped the area of GM identified as being smaller in BD patients than healthy subjects in a recent meta-analysis of VBM studies (176).   Reduced WMV is one of the most consistent findings early in BD (163), and our BMI-related WM reductions were extensive and were traversed by WM tracts that link important limbic brain structures, including the superior longitudinal fasciculus, which connects the prefrontal cortex to the striatum and temporal lobes; the inferior longitudinal fasciculus, which forms a conduit between the anterior/medial temporal lobe and the occipital lobe; the inferior fronto-occipital fasciculus, which links the orbitofrontal and temporo-occipital cortices; and the anterior thalamic radiation, which carries nerve fibres between the prefrontal cortex and the anterior and medial thalamic nuclei. Most notably, these WM clusters overlapped the two areas of WM found to be abnormal in BD in a recent meta-analysis of diffusion-weighted MRI studies (177) and contained 86  many of the same WM tracts. WM abnormalities in these regions have also been identified as likely endophenotypes for BD (186).  Our results in both patients and healthy subjects are all the more remarkable considering their weight distributions. The mean BMIs of both groups were very similar, and well within the normal range; only a quarter of each group was overweight, and less than 10% were obese. This has two implications. First, the distinct pattern of BMI-related volume reductions in patients was not due to differences in rates of overweight/obesity between patients and healthy subjects. Second, weight-related volume reductions were detectable in patients with, at most, modestly elevated BMI. This is in keeping with results from previous studies which found unfavourable psychiatric outcomes associated with small BMI increases within the normal range in BD. For example, one clinical trial showed that response rates to lithium decreased by 7% for every 1-point increase in BMI above a normal-range BMI of 22 (100). It is also consistent with at least some general population studies, which have reported that the risk of weight-related medical complications increases in a ?dose-dependent? fashion beginning at the high end of the normal BMI range (11).   The data presented here must be interpreted in light of the limitations of our report. In addition to those listed previously, the most significant is that our cross-sectional, case-control design precludes us from definitively establishing a causal relationship between increased BMI and decreased brain volumes. While a sizable body of evidence from human and animal studies suggests that obesity causes brain volume reductions and impaired brain functioning, there are other possible explanations for our findings (171). As well, the assignment of cortical labels and white matter tracts to the clusters of reduced GM and WM volume we identified was probabilistic. Finally, we enrolled patients into STOP-EM following treatment of their first manic episode, and so did not have data regarding pre-manic BMI or brain volumes.   Nonetheless, the current report is the first to specifically test the hypothesis that elevated BMI is associated with neurobiological changes in emotion-generating and modulating brain areas in people with BD. Our finding of BMI-related frontal, temporal, and subcortical volume reductions in patients, but not healthy subjects, strongly supports this hypothesis. Our results 87  provide further evidence of a neurobiological explanation for the well-established association between obesity and poor clinical outcomes in BD, and underscore the need for further research on obesity-related brain changes and outcomes in BD, and other psychiatric illnesses.  88  Table 5.1: Baseline sociodemographic and clinical characteristics of BD patients and healthy subjects taking part in a VBM MRI study  Patients with BD (N=57) Healthy Subjects (N=55)  Overweight or Obese (N=20) Normal Weight (N=37)  Overweight or Obese (N=17) Normal Weight (N=38)            Mean  SD Mean  SD P-value Mean (SD) Mean (SD) P-value Age 23.8   4.5 22.2 4.4 .202  22.0 (3.8)  22.3 (3.5)  .744   %  N %  N  % (N) % (N)  Gender      .647   .155 male 15.8%   9 33.3%  19  21.8% (12) 34.5% (19)  female 19.3%  11 31.6%  18  9.1% (5) 34.5% (19)  Ethnicity     .512   .118 Caucasian 29.8%  17 49.1%  28  23.6% (13) 43.6% (24)  Asian 3.5% 2 14.0%  8  3.6% (2) 23.6% (13)  Other 1.8% 1 1.8%  1  3.6% (2) 1.8% (1)    Mean  SD Mean  SD P-value    Baseline YMRS score 3.1  4.1 3.7  6.5 .684    Baseline MADRS score 5.4  7.3 6.2  8.4 .708    Baseline BPRS score * 21.8  4.1 22.7  6.1 .566    Duration of manic/mixed episode (days) * 55.3  43.2 64.6  55.8 .529    Duration of mood disorder (years) 2.4  3.0 3.1  4.7 .562    89   * N=55; 2 values missing ** N=56; 1 value missing  Patients with BD (N=57) Healthy Subjects (N=55)  Overweight or Obese (N=20) Normal Weight (N=37)  Overweight or Obese (N=17) Normal Weight (N=38)   %  N %  N P-value    Previous depressive episode 50.0%  10 54.1%  20 .770    Previous hypomanic episode 30.0%  6 13.5%  5 .168    Lifetime comorbidity         Anxiety ** 5.0%  1 11.1%  4 .645    Alcohol abuse or dependence ** 25.0%  5 11.1%  4 .256    Cannabis abuse or dependence ** 40.0%  8 30.6%  11 .474    Other drug abuse or dependence ** 0%  0 13.9%  5 .148    Pharmacotherapy for index mania         Lithium 30.0%  6 48.6%  18 .174    Divalproex 65.0%  13 37.8%  14 .050    Risperidone 35.0%  7 40.5%  15 .682    Olanzapine 20.0%  4 18.9%  7 1.000    Quetiapine 20.0%  4 21.6%  8 1.000    Mood stabilizer monotherapy 20.0%  4 10.8%  4 .432    Antipsychotic monotherapy 0%  0 8.1%  3 .545    Mood stabilizer plus antipsychotic 75%  15 73.0%  27 .868    No medication  5.0%  1 8.1%  3 1.000    90  Table 5.2: Baseline BMI and rates of overweight and obesity in BD patients and healthy subjects in a VBM MRI study     Patients with BD (N=57) Healthy Subjects (N=55) p-value      Mean (SD) Mean (SD)  BMI 24.1 (3.9) 24.0 (3.9) .900     BMI Category % (N) % (N) .891 Normal weight 64.9% (37) 69.1% (38)  Overweight 26.3% (15) 23.6% (13)  Obese 8.8% (5) 7.2% (4)     91  Figure 5.1: BMI-related brain volume reductions (p < .005 uncorrected, spatial extent threshold > 500 voxels), demonstrated on T1-weighted sections and glass brains, in a) GM in BD patients, b) WM in BD patients, and c) GM in healthy subjects        a)   b) LEFT LEFT LEFT LEFT LEFT LEFT LEFT LEFT 92   c)  No BMI-related WMV reductions were detected in healthy subjects. LEFT LEFT LEFT LEFT 93  Table 5.3: BMI-related GMV and WMV reductions in BD patients and healthy subjects, detected in linear regression models (p < .005 uncorrected, spatial extent threshold ? 500 voxels)  Region (Brodmann area)   Peak coordinates  Z value Voxel-level p, FWE corrected Cluster size Cluster-level p, FWE corrected White Matter Tracts   Decreased gray matter volume  in patients           R inferior temporal gyrus (20) R superior temporal gyrus (38) R uncus (28) R middle temporal gyrus (21) 56 39 35 66 -9 8 2 -13 -24 -20 -29 -9 4.72 3.94 3.55 3.30 .021 .334 .759 .949 2600 .009    -   Decreased white matter volume in patients Right hemisphere cluster R inferior frontal gyrus R fusiform gyrus R superior frontal gyrus R temporal sub-gyral WM R sub-lobar extra-nuclear WM R occipital sub-gyral WM R middle temporal gyrus R inferior temporal gyrus R frontal sub-gyral WM R middle occipital gyrus R midbrain  R anterior cingulate R parahippocampal gyrus R subcallosal gyrus R posterior cingulate       47 50 27 51 35 29 51 59 15 41 11 17 30 17 23       24 -6 45 -23 -7 -73 -15 -25 17 -66 26 42 -52 8 -69       7 -20 10 -15 6 -3 -13 -20 -11 -8 -18 -3 -8 -9    6       4.18 4.12 4.10 3.90 3.83 3.58 3.54 3.41 3.32 3.27 3.03 2.99 2.93 2.87 2.72      .082 .101 .106 .207 .248 .470 .514 .659 .748 .796 .948 .962 .977 .987 .998     11,476     .000      R Superior longitudinal fasciculus R Inferior longitudinal fasciculus R Inferior fronto-occipital fasciculus R Anterior limb of internal capsule R Posterior limb of internal capsule R Anterior thalamic radiation R External capsule R Fornix    94  Region (Brodmann area)         Peak coordinates          v lue Z value Voxel-level p, FWE corrected   Cluster size Cluster-level p, FWE corrected White Matter Tracts    Left hemisphere cluster L temporal sub-gyral WM L fusiform gyrus L middle temporal gyrus L sub-lobar extra-nuclear WM L midbrain  Decreased gray matter volume in healthy subjects L cuneus (18) R precuneus (31) L cuneus (18) L lingual gyrus (18) R lingual gyrus (19) L parahippocampal gyrus (19) R cuneus (18)  -51 -45 -50 -30 -9    -3 3 -5 -9 12 -21 8  -21 -40 -40 -21 -31    -73 -72 -88 -64 -61 -55 -85  -20 -17 -12 -3 -20    12 16 7 1 -0 -5 19  4.38 3.83 3.72 3.06 3.05    4.07 3.92 3.61 3.59 3.43 3.35 3.28  .038 .249 .337 .935 .940    .254 .391 .740 .762 .898 .940 .971  3906        4200  .018        .002  L Superior longitudinal fasciculus L Inferior longitudinal fasciculus L Inferior fronto-occipital fasciculus L Anterior thalamic radiation     - _________________________________________________________________________________________________________________________   L, left; R, right; FWE, family-wise error.   95  Figure 5.2: Effect of diagnosis on a) GMV and b) WMV (p < .005 uncorrected, spatial extent threshold > 500 voxels). c) Effect of BMI-diagnosis interaction on white matter volume     a)   b) LEFT LEFT LEFT LEFT LEFT LEFT LEFT LEFT LEFT 96   c)  Colored areas show volume reductions in patients with BD compared to healthy subjects (a, b), or regions where the impact of overweight/obesity on reduced brain volume was greater in patients than healthy subjects (c).  There was no main effect of BMI on GMV or WMV, and no BMI-diagnosis interaction on gray matter volume.     LEFT LEFT LEFT LEFT 97  Table 5.4: The effect of diagnosis and diagnosis-BMI interaction on GMV and WMV, detected with factorial models (p < .005 uncorrected, spatial extent threshold ? 500 voxels)   Region (Brodmann area)   Peak coordinates  Z value Voxel-level p, FWE corrected Cluster size Cluster-level p, FWE corrected White Matter Tracts   Decreased gray matter volume  in patients                   L uncus (20) L superior temporal gyrus (38) L inferior temporal gyrus (20) L fusiform gyrus (20) L middle temporal gyrus (38)  Decreased white matter volume in patients R hemisphere cluster R middle temporal gyrus R temporal sub-gyral WM R midbrain  R fusiform gyrus R sub-lobar extranuclear WM R insula R parahippocampal gyrus  Rostral L temporal cluster L temporal sub-gyral WM L middle temporal gyrus L superior temporal gyrus  -35 -27 -44 -48 -38      59 50 3 42 21 33 18   -41 -38 -45  -12 17 -9 -18 6      -31 -7 -31 -12 -22 -24 -34   -7 -3 -19  -35 -38 -36 -30 -42      -17 -23 -5 -29 9 12 4   -30 -38 1  4.24 3.45 2.95 2.85 2.73      4.39 4.16 3.59 3.48 3.15 3.14 3.10   4.74 4.65 3.24  .120 .842 .999 1.000 1.000      .035 .080 .450 .566 .882 .889 .909   .008 .012 .809  2542          3722         2146      .034          .013         .041      -          R inferior longitudinal fasciculus R inferior fronto-occipital fasciculus R posterior limb of internal capsule R anterior thalamic radiation R fornix     L superior longitudinal fasciculus L inferior longitudinal fasciculus      98     Region (Brodmann area)           Peak Coordinates        Z-score Voxel-level p, FWE corrected Cluster size Cluster-level p, FWE- corrected White Matter Tracts   Caudal L temporal cluster L superior temporal gyrus L supramarginal gyrus L cingulate gyrus L parietal sub-gyral WM L inferior parietal lobule L temporal sub-gyral WM L occipital subgyral WM L parahippocampal gyrus L sub-lobar extranuclear WM L posterior cingulate gyrus  BMI-diagnosis interaction ? greater impact of elevated BMI on WM reduction in patients R frontal cluster R medial frontal gyrus R middle frontal gyrus R anterior cingulate gyrus R superior frontal gyrus R inferior frontal gyrus  Bilateral fronto-temporal-subcortical cluster R cingulate gyrus L temporal sub-gyral WM R parietal sub-gyral WM L sub-lobar extranuclear WM L cingulate gyrus    -45 -39 -12 -24 -45 -27 -38 -23 -20 -15      8 29 15 21 24    20 -48 24 -20 -9   -40 -48 -43 -54 -39 -37 -46 -36 -34 -54      36 47 48 50 36    -43 -24 -46 -24 -12   10 30 31 31 22 3 -12 4 7 18      -17 9 -2 7 -8    34 -17 42 22 37   3.93 3.56 3.49 3.49 3.43 3.22 3.17 2.99 2.80 2.73      4.00 3.86 3.84 3.84 2.66    4.69 4.60 4.44 4.33 4.23   .176 .477 .551 .553 .616 .826 .866 .959 .993 .997      .137 .216 .229 .231 .999    .010 .014 .028 .043 .052   2762               1836     23,713   .014               .044       .000   L superior longitudinal fasciculus L inferior longitudinal fasciculus L inferior fronto-occipital fasciculus L cingulum L anterior thalamic radiation          R inferior fronto-occipital fasciculus R forceps minor     L and R superior longitudinal fasciculus L and R inferior longitudinal fasciculus L and R inferior fronto-occipital   99  L, left; R, right; FWE, family-wise error. 100  Figure 5.3: WMV reductions in overweight/obese BD patients compared to overweight/ obese healthy subjects (p < .005 uncorrected, spatial extent threshold > 500 voxels)      There were no gray matter volume reductions in overweight/obese patients, and no gray or white matter volume reductions in normal-weight patients compared to normal-weight healthy subjects. LEFT LEFT LEFT LEFT 101  Table 5.5: WMV reductions in overweight/obese BD patients compared to overweight/obese healthy subjects (p < .005 uncorrected, spatial extent threshold ? 500 voxels)   Region (Brodmann area)   Peak coordinates  Z value Voxel-level p, FWE corrected Cluster size Cluster-level p, FWE corrected White Matter Tracts   R frontal cluster           R medial frontal gyrus R superior frontal gyrus  R frontal sub-gyral WM R anterior cingulate gyrus  8 24 30 15  38 51 26 47  -17 10 7 -3  4.09 4.08 3.28 2.93  .128 .133 .824 .983  3675 .016 R superior longitudinal fasciculus R anterior thalamic radiation R forceps minor  Bilateral fronto-temporal-subcortical cluster L cingulate gyrus L temporal sub-gyral WM R posterior cingulate gyrus R cingulate gyrus R temporal sub-gyral WM L precuneus L sub-lobar extranuclear WM L parietal sub-gyral WM L frontal sub-gyral WM L posterior cingulate gyrus L middle temporal gyrus R insula L anterior cingulate gyrus R middle temporal gyrus R precuneus R occipital sub-gyral WM R frontal sub-gyral WM R lingual gyrus   -11 -50 23 15 44 -17 -8 -21 -29 -9 -50 35 -9 50 12 27 33 26    -42 -24 -59 -45 -52 -60 -3 -58 23 -49 -18 -22 30 -36 -55 -75 -4 -67    33 -12 7 34 -3 31 24 24 10 19 -8 10 22 -3 55 -2 45 -5    4.77 4.60 4.56 4.41 4.29 4.10 4.07 4.05 4.05 4.01 4.01 3.93 3.77 3.70 3.62 3.52 3.50 3.48    .010 .020 .023 .041 .064 .123 .137 .146 .146 .165 .165 .210 .332 .398 .476 .578 .605 .620    32,640                 .000           L and R superior longitudinal fasciculus L and R inferior longitudinal fasciculus L and R inferior fronto-occipital fasciculus L and R cingulum L and R body of corpus callosum L and R forceps major R forceps minor   ______________________________________________________________________________________________________________________ 102  L, left; R, right; FWE, family-wise error. There were no GMV reductions in overweight/obese patients compared to overweight/obese healthy subjects; and no GMV or WMV reductions in normal-weight patients compared to normal-weight healthy subjects. 103  6.  BMI ? diagnosis interaction in hippocampal N-acetylaspartate at recovery from a first manic episode  6.1  Introduction  Chapters 4 and 5 reported the results of structural MRI studies examining the relationship between BMI and brain structure in STOP-EM patients. These findings suggested that elevated BMI is associated with unique structural brain changes early in the course of BD, such that the neuropathology characteristic of the illness is exacerbated in overweight/obese patients.   The purpose of the current analyses is to extend these results by examining whether elevated BMI is also associated with altered neurochemistry early in BD. To do this, we used single-voxel 1H-MRS to compare hippocampal N-acetylaspartate (NAA) in overweight/obese and normal-weight patients with BD at recovery from their first manic episode. To determine whether weight-related alterations in these neurochemicals were unique to BD, we carried out the same investigations in a comparison group of age- and gender-matched overweight/obese and normal-weight healthy subjects.   NAA is the second most abundant amino acid in the brain after glutamate. It has a number of important roles, including acting as a ?molecular water pump? which removes excess cellular water generated by metabolic processes; functioning as an acetyl donor in myelin production; and, as N-actetylaspartylglutamate, serving as a neurotransmitter and a storage depot for glutamate (187). In contrast to the extensive literature on the impact of obesity on brain structure and cognitive functioning, there is relatively little information on neurochemical changes associated with elevated BMI. The small number of studies that examined NAA unfortunately have produced inconsistent results, and none of them examined hippocampal NAA specifically, providing little information to guide hypothesis formulation in our healthy subjects (188-190). However, NAA has been extensively studied in BD, and a number of studies suggest that NAA levels in the hippocampus and prefrontal cortex are abnormally low in BD patients compared to healthy subjects (191,192). Not all studies have supported this finding (193,194), suggesting that reduced NAA may occur in a sub-group of patients. To the best of our knowledge, however, no 104  previous studies have investigated whether NAA levels differ in overweight/obese and normal-weight patients.   We focused on the hippocampus as numerous lines of evidence implicate it in the pathophysiology of BD (195), and our previous MRI studies suggested that temporal lobe structures may be particularly susceptible to obesity-related changes. Based on our structural imaging findings, we hypothesized that overweight/obese patients would have reduced hippocampal NAA compared to normal-weight patients, but that NAA would not vary with BMI in healthy subjects. Taken together with our MRI studies, a positive result would suggest that elevated BMI is associated with widespread neurobiological changes early in the course of BD, primarily affecting areas/processes vulnerable in the illness.   6.2  Methods  6.2.1  STOP-EM  The BD patients and healthy subjects included here were drawn from the UBC STOP-EM program. The inclusion and exclusion criteria, assessment procedures, and measurement of weight and BMI for STOP-EM were as described in Chapter 2. The consecutively-enrolled FEP healthy subjects, who had been added to our samples in Chapters 4 and 5, did not have MRS data, and so were not included in the current analyses.  6.2.2  MRI protocol and data extraction  Each patient and healthy subject had a cerebral MRI at enrolment. MR images were obtained as outlined in Chapters 4 and 5. Hippocampal volume was quantified using FSL v4.1 (FMRIB Software Library, www.fmrib.ox.ac.uk/fsl/) (150,151). The FSL Brain Extraction Tool (BET) v2.1 was used to remove all non-brain signal from the datasets. Skull-stripped images were then segmented into grey matter, white matter, and CSF using FMRIB's Automated Segmentation Tool (FAST 4) based on MNI 152 tissue probability maps (152). The stripped image sets were registered to MNI 152 standard space with FMRIB's Linear Image Registration Tool (FLIRT) 105  v5.4.2. A global optimization procedure (12 degrees of freedom, correlation ratio cost function) was performed to provide robust affine registration of brain images to MNI template (152). Hippocampal volumes were determined by mapping the MNI structural atlas onto the processed datasets. The volumes of hippocampal and non-hippocampal GM and WM in the MRS voxel were calculated using in-house scripts.  6.2.3  Acquisition and processing of MRS signals  MRS signals were also acquired at enrolment with the Philips 3.0 T unit. T2-weighted coronal, sagittal, and axial MR images for anatomical parameters were first obtained. 30mm x 15mm x 15 mm voxels to the hippocampus on both sides of the brain were then created using a Point Resolved Spectroscopy (PRESS) sequence (TR = 2000ms, TE = 35ms; Figure 1). Using the sagittal image, the voxel was placed with the long axis angled along the hippocampus. The coronal and axial images were then used to adjust the position of the voxel in the medial/lateral and superior/inferior directions so as to include the entire hippocampus and avoid CSF.   Data analysis was performed using LCModel software (196). Metabolite signals were normalized to the unsuppressed water spectrum in order to obtain relative concentrations, and reported as institutional units (IU) (Figure 6.1).  6.2.4  Data analysis and statistics  We included in our analyses all patients and healthy subjects with baseline data for BMI and hippocampal NAA. Since patients were at an early illness stage and had received pharmacotherapy for a relatively short time, rates of overweight and obesity were low. They were therefore combined so that patients with overweight or obesity were considered together, and similarly for healthy subjects. All of our comparisons were two-tailed and were carried out using a significance level of ?=0.05.   Statistical analyses were carried out using IBM SPSS Statistics for Windows 19.0 (SPSS Inc, Chicago, IL). We examined sociodemographic and clinical variables using t-tests, ANOVA, ?2 106  tests, and Fisher?s exact test as appropriate. We compared mean BMIs between all patients and all healthy subjects using ANCOVA, controlled for age and gender, and mean hippocampal volumes, controlled for age, gender, and cranial volume.   For our primary analyses, we used ANCOVA to examine whether mean bilateral hippocampal NAA differed between overweight/obese patients and normal-weight patients with BD. We covaried these analyses for age and gender, and to ensure that our findings were not due to BMI-related differences in the composition of the hippocampal voxel, we also covaried for the volumes of hippocampal and non-hippocampal GM and WM in the voxel. We carried out the same analyses in our healthy subjects.   We also constructed a linear regression model for patients to determine whether BMI significantly predicted NAA when other potential sociodemographic and clinical moderators were controlled for. The variables selected for entry into our regression models included BMI; age; gender; the number of previous depressive and hypomanic episodes; the duration of the first manic episode; current YMRS and MADRS scores; psychosis during the initial mania; alcohol, cannabis, and drug dependence; and treatment with lithium, divalproex, risperidone, olanzapine, and quetiapine. We first calculated Pearson correlation coefficients to determine which of these variables were significantly associated with hippocampal NAA, and those with significant correlations were then entered into the regression model.   Finally, to directly examine whether elevated BMI was associated with distinct effects on NAA in patients and healthy subjects, we used factorial ANCOVA to assess the interaction between diagnosis (patient vs. healthy subject) and BMI category (normal weight vs. overweight/obese) on mean hippocampal NAA. Hippocampal NAA was the dependent variable in these analysis, diagnosis and BMI category were factors, and age, gender, and the volumes of hippocampal and non-hippocampal GM and WM in the voxel were covariates.     107  6.3  Results  6.3.1  Patients with BD and healthy subjects  Sociodemographic and clinical characteristics. Of the initial 71 patients and 30 healthy subjects enrolled in STOP-EM, 57 patients and 27 healthy subjects had baseline BMI data and underwent the MRS procedure. One patient with missing data did not wish to be weighed, while 13 patients and 3 healthy subjects declined the MRS procedure. In addition, we were unable to acquire usable MRS data for hippocampal NAA from 1 patient and 1 healthy subject. This left us with a total sample 56 patients and 26 healthy subjects. Overweight/obese and normal-weight patients and healthy subjects were well-matched on important sociodemographic and clinical characteristics, which are listed in Table 6.1.   BMI and rates of overweight/obesity. Patients and healthy subjects had similar mean BMIs (patients: 24.06 (SD=3.95); healthy subjects: 23.59 (2.83); F=.279, df=1, p=.599) and did not differ with respect to the proportions with normal weight (64.9% vs. 77.8%), overweight (26.3% vs. 18.5%), or obesity (8.8% vs. 3.7%) (?2=1.566, df=2, p=.457).   Hippocampal volumes and hippocampal NAA. Patients and healthy subjects had comparable mean bilateral hippocampal volumes (8.78 mL (0.81) vs. 9.03 mL (0.82); F=.051; df=1; p=.821) and hippocampal NAA concentrations (5.65 institutional units [IU] (0.56) vs. 5.56 IU (0.58), F=.655, df=1, p=.421).  6.3.2  The relationship between BMI and hippocampal NAA  Hippocampal volumes. Overweight/obese and normal-weight patients had similar mean hippocampal volumes (8.62 mL (0.58) vs. 8.85 mL (0.91); F=.643; df=1; p=.426), as did overweight/obese and normal-weight healthy subjects (9.14 mL (0.38) vs. 9.01 mL (0.87); F=.006; df=1; p=.941).  108  Hippocampal NAA. Our primary analysis demonstrated that in overweight/obese patients, hippocampal NAA was 6.1% lower (5.43 IU (0.58) vs. 5.78 IU (0.52), F=5.325, df=1, p=.025) (Figure 6.2). In contrast, in healthy subjects, there was no significant difference between overweight/obese and normal-weight individuals in hippocampal NAA (5.81 IU (0.63) vs. 5.49 IU (0.56), F=2.552, df=1, p=.128) (Figure 6.2). The effect size (ES; Cohen?s d) for reduced NAA in overweight/obese patients was moderate at 0.64.  Linear regression model. BMI and the use of quetiapine were significantly correlated with mean hippocampal NAA in patients with BD (Figure 6.3). When these were entered into a regression model, BMI remained a significant predictor of hippocampal NAA (BMI: ?=-.259, t=-2.062, p=.044; quetiapine use: ?=.304, t=2.418, p=.019). In healthy subjects, the correlation between BMI and hippocampal NAA was smaller and not statistically significant (Figure 6.3).   Factorial ANCOVAs. There were no main effects of diagnosis or BMI category on hippocampal NAA (diagnosis: F=.054, df=1, p=.817; BMI: F=.044, df=1, p=.835). However, we did detect a significant diagnosis x BMI interaction for NAA (F=5.696, df=1, p=.020), indicating a differential effect of BMI on NAA in patients compared to healthy subjects.   6.4  Discussion  Our results demonstrate that overweight/obese STOP-EM patients had significantly lower hippocampal NAA than normal-weight patients at recovery from their first manic episode ? the first time, in other words, that they could be diagnosed with BD. As well, approximately 75% of the patients classified as ?overweight/obese? were in fact overweight, suggesting that these neurochemical changes were present even with moderately elevated BMI. A preponderance of previous MRS studies have shown that reduced hippocampal NAA is a common neurobiological finding in patients with BD. Thus, taken together with our previous MRI studies, our results support our hypothesis that the structural and chemical brain changes typical of BD are more pronounced in patients with elevated BMI.  109  The strength of the association between BMI and hippocampal NAA was surprisingly large. In fact, the difference in NAA that we detected between overweight/obese and normal-weight patients (-6.1%) was over three times greater than the difference between all BD patients and all healthy subjects (1.6%). Our findings were not simply due to BMI-related hippocampal structural changes, such as reduced hippocampal volume or alterations in the tissue composition of the MRS voxel, since hippocampal volumes did not differ based on BMI, and the volumes of hippocampal and non-hippocampal grey and white matter in the voxel were entered as covariates in our analyses. The findings were also unlikely to result from differences in sociodemographic, clinical, or treatment variables between overweight/obese and normal-weight patients, since the two groups were well-matched on these. Moreover, BMI remained a predictor of NAA in a regression analysis which controlled for other significantly-correlated variables.   In contrast to the results in patients, we did not detect weight-related changes in hippocampal neurochemistry in healthy subjects. Although our sample included a smaller number of healthy subjects than patients, resulting in reduced statistical power to detect differences, we are confident that the findings in patients are unique to BD since our factorial ANCOVAs detected diagnosis-BMI interactions for hippocampal NAA, indicating a distinct effect of BMI on neurochemistry in patients. Moreover, the neurochemical changes seen in overweight/obese healthy subjects were actually in the opposite direction to those in patients ? ie. increased NAA. Although the lack of statistical significance of this finding prevents me from drawing firm conclusions about them, it is worth noting that it are similar to that reported in major depressive disorder, ie. modestly increased NAA, at least early in MDD (197-199). Thus, it is consistent with the association between obesity and depression in otherwise-healthy individuals.   The design of STOP-EM, which is essentially a case-control study, precludes us from determining whether there is a causal relationship between elevated BMI and the structural and functional brain changes we observed in our patients. There are two possible explanations for our results: 1) there exists a subtype of BD characterized by a propensity for weight gain, pronounced neurobiological changes, and poor clinical outcomes; and 2) excessive weight gain in BD causes brain volume loss, alters neurochemistry, and negatively impacts the course of BD. There is some evidence consistent with the first hypothesis, including the fact that a common 110  polymorphism of the fat mass and obesity-related (FTO) gene, an established risk factor for obesity, is associated with reduced frontal and occipital lobe volumes independent of its effect on BMI (171), demonstrating that a ?third variable? may mediate both obesity and structural brain changes. However, a number of lines of evidence support the second hypothesis and suggest that elevated BMI causes brain changes. Obesity damages numerous bodily organs, including the heart, liver, kidneys, and peripheral nervous system (200,201). In a prospectively-followed sample of healthy adults, people with elevated BMI in middle age experienced greater brain volume loss over the subsequent 6 years (81), and obesity has also been shown to precede the development of neuropsychiatric illnesses such as depression and dementia (89,202,203). Most definitively, animal studies, which permit random assignment to obesigenic conditions, have shown that Rhesus macaque monkeys assigned to a low-calorie diet, which resulted in low BMI through adulthood, experienced less brain volume loss over their lifespans than monkeys assigned to ?diet-as-usual? and who had greater BMI (96).  The results we have presented here must be interpreted in light of the limitations of our study. In addition to the naturalistic design of STOP-EM and the relatively small number of healthy subjects in our sample, these include the fact that we enrolled patients after the resolution of their first manic episode, and thus we did not have data regarding their pre-manic BMI or brain chemistry. As well, since rates of overweight and obesity we low in patients and healthy subjects, they were combined for our analyses, meaning that we could not examine their effects separately. Furthermore, we did not collect data on diet or exercise, which might have independent effects on neurobiological functioning.  Notwithstanding the above limitations, our results are important in several respects. STOP-EM is the first program to measure the association between BMI and brain chemistry in BD, or any psychiatric illness. Our main finding, that elevated BMI is associated with unique alterations in neurochemistry very early in the course of BD, provides supporting data for our hypothesis that the neurobiological changes characteristic of BD are more pronounced in overweight/obese individuals. Further studies are needed to confirm our findings, and to determine whether they hold true in patients at later stages of the illness. As well, given the high rates of obesity in people with other psychiatric disorders such as major depressive disorder and schizophrenia, 111  further work should also address whether similar BMI-brain relationships impact on these illnesses. 112  Figure 6.1: Sample MRS spectra from a) a BD patient and b) a healthy subject   a)   b) 113  Table 6.1: Sociodemographic and clinical characteristics of BD patients and healthy subjects in a single-voxel hippocampal MRS study   Patients with BD (N=57) Healthy Subjects (N=27)  Overweight/Obese (N=20) Normal Weight (N=37) P-value Overweight/Obese (N=6) Normal Weight (N=21) P-value         Mean (SD) Mean (SD)  Mean (SD) Mean (SD)  Age 23.75 (4.51) 22.38 (4.37) .268 25.00 (7.64) 21.95 (3.28) .156 Years of education 14.60 (1.82) 13.74 (2.37) .166 14.67 (3.88) 14.86 (2.35) .881         % (N) % (N)  % (N) % (N)  Gender    .792   1.000 male 15.8% (9) 31.6% (18)  11.1% (3) 37.0% (10)  female 19.3% (11) 33.3% (19)  11.1% (3) 40.7% (11)  Ethnicity   .512   .633 Caucasian 29.8% (17) 49.1% (28)  18.5% (5) 51.6% (14)  Asian 3.5% (2) 14.0% (8)  3.7% (1) 25.9% (7)  Other 1.8% (1) 1.8% (1)  0% (0) 0% (0)   Mean (SD) Mean (SD)     Baseline rating scale scores       YMRS 3.05 (4.07) 3.73 (6.46) .672    MADRS  5.40 (7.29) 6.24 (8.44) .708    PANSS  7.45 (1.23) 7.84 (1.66) .364    Duration of mood disorder (years) 2.40 (3.02) 3.08 (4.71) .562    114      Patients with BD (N=57) Healthy Subjects (N=27)  Overweight/Obese (N=20) Normal Weight (N=37) P-value Overweight/Obese (N=6) Normal Weight (N=21) P-value         % (N) % (N)     Previous depressive episode 50.0% (10) 54.1% (20) .770    Previous hypomanic episode 30.0% (6) 13.5% (5) .168    Lifetime comorbidity       Anxiety * 5.0% (1) 11.1% (4) .645    Alcohol abuse or dependence * 25.0% (5) 11.1% (4) .256    Cannabis abuse or dependence * 40.0% (8) 30.5% (11) .474    Other drug abuse or dependence * 0% (0) 13.9% (5) .148    115    * N=56; one value missing  Patients with BD (N=57) Healthy Subjects (N=27)  Overweight/Obese (N=20) Normal Weight (N=37) P-value Overweight/Obese (N=6) Normal Weight (N=21) P-value         Mean (SD) Mean (SD)     Pharmacotherapy for index mania       Lithium 30.0% (6) 45.94% (17) .242    Divalproex 65.0% (13) 40.5% (15) .078    Risperidone 35.0% (7) 43.2% (16) .545    Olanzapine 20% (4) 18.9% (7) 1.000    Quetiapine 20% (4) 18.9% (7) 1.000    Mood stabilizer + antipsychotic 75% (15) 73.0% (27) .868    No medication  5.0% (1) 8.1% (3) 1.000    116  Figure 6.2: Mean hippocampal NAA in overweight/obese and normal-weight BD patients and healthy subjects      p=.025 p=NS BD Patients Healthy Subjects Overweight/obese Normal-weight 117  Figure 6.3: Correlation between BMI and mean hippocampal NAA in BD patients and healthy subjects    r=-.263 p=.050 r=.118 p=.565 BMI Mean hippocampal NAA 118  7.  Summary and discussion  7.1  Summary of research findings  This thesis comprehensively documents the dramatic weight gain that accrues early in the course of BD, and the adverse clinical and neurobiological outcomes that accompany it. Specifically, I found that at recovery from their initial manic episode, there was no difference between BD patients and healthy subjects in mean BMI, which was well within the normal range for both groups, or in rates of overweight or obesity, which were low. However, during just 6 months of maintenance treatment, patients experienced significantly greater weight gain than healthy comparator subjects, and almost half experienced CSWG. Patients? weight gain was maintained during the subsequent 6 months, and combined rates of overweight and obesity at 6 months and 12 months exceeded 50%, and were almost double those of healthy subjects. Even this early in the illness, psychiatric outcomes were demonstrably worse in patients with CSWG, who spent 23% more time with depression during the first 12 months of maintenance treatment, and had significantly poorer 12-month psychosocial functioning,  I also found clear neurobiological correlates of elevated BMI in patients, with a highly consistent pattern: the pathophysiologic brain changes believed to underlie BD were exacerbated in overweight/obese patients. Current models of BD, supported by meta-analyses of MRI studies, posit that the illness results from structural and functional alterations in prefrontal, temporal, and subcortical brain areas that are involved in generating and modulating emotional responses. Total white matter (WM) reductions, likely reflecting altered connectivity between these areas, are particularly prominent early in the illness. BD is also characterized by neurochemical changes, including reduced hippocampal N-acetylaspartate (NAA), an amino acid that is often taken as a marker of neuronal density and health. In our sample, overweight/obese patients had significantly reduced total WM and total temporal lobe volume compared to normal-weight patients. The WM reductions almost exclusively affected tracts linking important prefrontal, subcortical, and temporal limbic structures. Total gray matter (GM) did not differ based on BMI, but overweight/obese patients did have regional GM reductions in the right temporal lobe, a brain area repeatedly identified as being abnormal in BD patients. Moreover, overweight/obese 119  patients had significantly reduced hippocampal NAA. Interestingly, none of the above changes were detected in overweight/obese healthy subjects, who, consistent with previous reports, had smaller total brain volume and total GM than normal-weight subjects, and no BMI-related differences in brain chemistry. My results therefore demonstrate that the brain changes that accompany elevated BMI following the first manic episode are unique to the illness.   7.2  Clinical implications  7.2.1  Weight gain  The results presented here extend those of previous reports on obesity in BD. Studies in North American samples have reported that 55%-75% of patients with BD are overweight or obese (14,16-20,27), with the rate of obesity in most studies substantially greater than the North American average of approximately 30% (14,16-20,27). However, these reports were limited by retrospective or cross-sectional designs, and/or the assessment of weight in patients with affective illnesses of up to 21 years duration (14), which made it difficult to determine the course of weight gain relative to the affective component of BD. Using a longitudinal study design in a first-episode mania sample allowed us to examine the timing of weight gain prospectively from the time of first diagnosis of BD, while at the same time minimizing many of the confounding treatment and clinical variables that have plagued previous reports of obesity in BD, such as the multiple medication trials patients with lengthy illnesses are exposed to, the varying numbers of mood episodes they experience, their differences in primary illness polarities (eg. predominantly depressive vs predominantly manic), and others.   Two particular findings regarding weight gain in STOP-EM patients deserve highlighting. The first is that at enrolment, patients with BD and healthy subjects had very similar BMIs and rates of overweight and obesity. This is despite the fact that almost two-thirds of patients had experienced previous depressions and/or hypomanias, and had a mean duration of affective illness of approximately three years. This fact, along with the well-known propensity of many mood stabilizing and antipsychotic medications to promote weight gain (45), immediately suggests that BD-specific pharmacotherapy, rather than the presence of a mood disorder, was 120  primarily responsible for the weight gain we observed. This supposition is further supported by the fact that in our regression analyses, treatment with two particular medications, risperidone and olanzapine, which are known to be among the most obesigenic mood stabilizers, significantly predicted CSWG. However, we were unable to definitively determine the impact of medications on weight, or to assess the alternative possibility, that weight gain is an inherent component of post-manic (but not pre-manic) BD, as only one patient in our sample was medication-free for the entire 12 months of follow-up. The absence of excess weight at enrolment was all the more striking since most patients received treatment with obesigenic medications for several weeks during their acute manic episode, ie. preceding their enrolment in STOP-EM. The STOP-EM program did not begin collecting data until recovery from the initial mania, and so the duration or type of medication treatment patients received prior to recovery are not known with certainty, although in almost all cases the medications used during the acute mania are the same as those continued into maintenance therapy. The most likely explanation for this apparent lack of weight gain during acute treatment lies in the common clinical observation that the high energy and activity levels characteristic of mania frequently cause weight loss until treatment is initiated; weight gain during acute treatment then simply leads to weight normalization.   The second noteworthy point is that the weight gain that patients experienced during just 6 months of maintenance treatment resulted in their combined rate of overweight and obesity increasing from 35% to 55% in that time. The proportion of STOP-EM patients who were overweight or obese after 6 months was thus already at the lower end of the rates reported for North American BD samples of patients with established illnesses. This, along with the fact that weight appeared to stabilize during the second 6 months of treatment, suggests that a large fraction of the total weight gain experienced by BD patients occurs very early in the illness course. This underscores the importance of clinicians considering the possibility of weight gain from the moment they begin treating people with BD, and taking steps to prevent it and reverse it when it occurs. It is important to consider, however, that while the percentage of patients in our sample who were overweight almost doubled over 6 months, from 26% to 47%, the rate of obesity remained unchanged at 9%. This relatively low rate of obesity contrasts with previous studies, which reported that 27%-45% of patients with BD were obese. Longer term follow-up of 121  our cohort will be required to clarify how much further weight gain occurs, when it occurs, and the degree to which obesity rates increase over more prolonged illness durations.  7.2.2  Mood symptoms and psychosocial functioning  The true nature of the relationship between elevated BMI and mood symptoms remains unclear, despite the results presented here. We found that patients with CSWG spent 23% more time depressed than those without CSWG, arguably a clinically relevant difference, but one that was not statistically significant. Similarly, we did not detect significant differences between patients with baseline overweight/obesity and those with baseline normal weight in the frequency of pre-manic mood episodes or suicide attempts, although there were numerical differences in the expected directions: overweight/obese patients had approximately one-third more previous mood episodes, and almost three times as many prior suicide attempts as normal-weight patients. The lack of statistical significance of these findings might be explained by a number of factors, including 1) our relatively small sample size, 2) the short duration of mood illness in STOP-EM patients compared to those in previous studies, 3) the short follow-up period of 12 months, and 4) the fact that few patients enrolled in STOP-EM were obese (about 10%), and so they were combined with overweight patients for our analyses. If there is a ?dose-response? relationship between weight and previous mood episodes, then combining obese and overweight patients may have obscured differences due to obesity. Thus, while the results of these analyses were negative, whether they are a ?true negative?, or rather negative due to inadequate sample size, follow-up period, and statistical power is an open question.   Notwithstanding the above, we did have several interesting findings, particularly with respect to the relationship between weight and psychosocial functioning. It has become clear that simply treating mood symptoms in people with BD is insufficient to return them to full health, and a strong consensus has emerged in the field that re-establishing full premorbid functioning is the ultimate goal of treatment. However, although obesity has been associated with impaired psychosocial functioning in non-psychiatric samples (107,108), and numerous studies have demonstrated that functional impairment in BD patients persists even in the absence of mood symptoms (103-105,146), the impact of weight gain on functioning in patients with BD has 122  received surprisingly little attention. In fact, this are the first prospective data to demonstrate an association between CSWG and impaired functioning in BD The fact that it occurs so early in the illness makes it especially noteworthy. Thus, our results suggest that CSWG is an overlooked, but potentially modifiable, cause of functional impairment in BD. Further studies to replicate these findings, and particularly studies investigating whether preventing or reversing weight gain improves functioning in patients with BD, are clearly needed.  Despite the paucity of studies on weight gain and functioning in BD, our findings are very much in keeping with previous reports of functional impairment in overweight and obese individuals in non-psychiatric samples. Overweight adolescents and young adults have been shown to have poorer educational achievement, reduced household income, and lower rates of marriage than their normal-weight peers (204). Much of this disability is related to stigma directed toward overweight and obese people, which has been detected in peers, employers, and even health care professionals and the individuals? own parents (205-208). Biological factors may also be etiologically relevant, given the reduced brain volumes and decrements in cognitive functioning seen in obese individuals, particularly impairments in memory and executive functions, even as young adults and in middle age (77,86,87). Whatever the cause or causes, most studies have reported a greater negative impact of obesity on functioning in females (204,208), similar to our finding that CSWG had a stronger association with functional impairment in female patients with BD, even though males experienced greater weight gain.  A final noteworthy finding is that the association of CSWG with mood symptoms and functional outcomes in BD patients was substantially greater than the association between these outcomes and baseline or 12-month overweight/obesity. Our results thus suggest a distinction between the effects of acute weight gain and the effects of stably elevated BMI. This is a distinction that has received little, if any, attention in previous studies on the health consequences of weight. One can envision psychological mechanisms which might be at play ? the rapid, marked weight gain that occurs in the first six months of maintenance treatment might be particularly distressing for patients and their friends and families, but the weight stabilization that tends to follow in the next six months might lead to a degree of psychological adaptation, even if weight does not return to normal. Biological factors could also be relevant. For instance, the mechanisms underlying acute 123  inflammation differ markedly from those of chronic inflammation, and the biological effects of acute weight gain and chronic overweight/obesity might differ accordingly. Further research is needed to clarify these points.  7.3  Implications for understanding the neurobiology of BD  Over the past decade, ground-breaking human and animal studies have provided evidence that obesity harms the brain. In humans, obese individuals, even children as young as five years of age, have decrements in GM volume and cognitive impairment, as well as an increased risk of developing serious neurological illnesses later in life, including multiple sclerosis, Parkinson?s disease, and Alzheimer?s disease. Animal studies have definitively established that obesity is causally linked with adverse neurobiological changes, strongly suggesting that the same is also true in humans. However, despite our understanding of psychiatric illnesses as brain disorders; the high obesity rates endemic to many psychiatric conditions; and the unambiguous association of obesity with more severe psychiatric symptoms; research has lagged behind in studying the impact of obesity on the pathophysiology of these conditions. The data presented here are thus the first to demonstrate a negative impact of elevated BMI on the brain for any psychiatric illness. In particular, the finding that the pathophysiological changes characteristic of BD are exacerbated in overweight/obese individuals immediately suggests a neurobiological mechanism to explain their greater illness severity. Our findings are all the more striking since approximately three-quarters of STOP-EM patients in the ?overweight/obese? group were in fact overweight, demonstrating that definite neurobiological changes were present in patients with only modestly elevated BMI.  The results in our healthy subjects were markedly different from those in our patients, but were highly consistent with previous studies in healthy samples, including children and adolescents, which, like ours, reported BMI-related GM reductions that were most pronounced in the occipital lobes (171,209). The mechanisms that mediate the association between increased BMI and reduced brain volumes are not known with certainty, for either patients or healthy subjects, and so the explanation for their strikingly different patterns of BMI-related brain volume reductions in is unclear. In any case, the marked discrepancy in findings between patients and 124  healthy subjects shows that the impact of elevated BMI on limbic brain areas is unique to BD patients. This, in turn, suggests two intriguing, and competing, possibilities regarding the impact of weight on the neurobiology of BD. The first is that elevated BMI interacts with the pathophysiologic diathesis underlying BD, perhaps creating a generalized brain vulnerability which allows the separate processes underlying BD to progress more rapidly. The second is that increased BMI has specific pathological effects that differ between people with BD and healthy subjects ? in other words, that these BMI-related changes in fact are the pathophysiology of the illness, or at least a part of it.  However, a note of caution should be interjected in interpreting our findings. The naturalistic design of STOP-EM, and the correlational nature of our results, mean that we cannot assume a causal relationship between elevated BMI and the neurobiological changes we detected in either patients or healthy subjects. There are, in fact, three possible explanations for them. The first is that elevated overweight/obesity causes the structural and functional brain changes observed in STOP-EM participants. On the balance of probabilities, I believe that this is the most likely situation. As previously noted, a causal relationship has been demonstrated between weight and brain changes in rodents and primates. Furthermore, human studies have shown that obese people in non-psychiatric samples who are randomly assigned to diet and weight loss interventions show improved cognition following their weight loss, compared to other obese individuals randomly assigned to not receive the intervention, also suggesting causality (88). However, a second possibility is that there is a subtype of BD characterized by both a propensity for weight gain and a neurobiologically more rapidly progressive illness, which cannot be definitely ruled out based on our data. It should be noted that if this is the case, it is unlikely to be a result of differential medication use between overweight/obese and normal-weight patients, since the types of medications prescribed were similar between the two groups, and weight gain did not differ based on the number of medications prescribed. The final possibility is that a ?third variable? might mediate both obesity and brain changes. Such an occurrence has already been established in non-psychiatric samples, as evidenced by the fact that  a common polymorphism of the fat mass and obesity-related (FTO) gene, an established risk factor for obesity, is associated with reduced brain volumes independent of its effect on BMI in humans (171). Whether elevated BMI does in fact cause the adverse clinical and neurobiological outcomes 125  observed in BD, and whether intervening to minimize or reverse weight gain will lead to improved outcomes, are the most important questions to arise from this line of research, and I have written a grant application for a clinical trial designed to answer them (see ?Future Research Directions? below).  If overweight/obesity does cause the brain changes we detected, our findings have immediate implications for understanding disease severity and neuroprogression in BD. Clinicians have long recognized that mood episode frequency tends to increase during the early course of BD, typically over the initial three to four episodes, with a corresponding shortening of well intervals and increased illness severity in the form of treatment refractoriness (46). In keeping with this, longitudinal brain imaging studies have also shown progressive structural brain changes early in BD, particularly involving volume reductions in the frontal lobes (210-212). However, not all patients experience this deteriorating course, and so identifying factors that predict illness progression, particularly modifiable ones, is a major priority in the field. I am currently conducting longitudinal analyses in the STOP-EM sample to determine if patients who experience CSWG during maintenance treatment also have more volume loss in limbic brain areas over time than patients without CSWG, and the initial results suggest that this is indeed the case. (Unfortunately, the analyses are not yet complete enough to include in this thesis.) If so, this would strongly implicate weight gain as a modifiable driver of neuroprogression in BD.  Our findings further suggest that taking BMI into account in future investigations may help researchers to more accurately elucidate the pathophysiology of BD, and perhaps equally importantly, explain some of the heterogeneity that has plagued neurobiological investigations in BD and other psychiatric illnesses. To the first point, to the degree that obesity is associated with different MRI findings in BD patients and healthy subjects, then comparing these differences might provide information regarding which brain regions are important to the neurobiology of BD, or vulnerable to damage in patients. With respect to the second point, numerous confounding factors obscure illness-specific brain changes in neurobiological studies of BD. Perhaps the best documented of these is the normalizing effect of mood stabilizing medications such as lithium. A recent meta-analysis of structural neuroimaging studies clearly illustrated this, in reporting that BD patients who were not treated with lithium had smaller hippocampal 126  volumes than healthy subjects, while lithium treated patients actually had significantly larger hippocampal volumes (213). Thus, to the extent that obesity might be associated with similar MRI findings in patients and healthy subjects in some brain areas, future studies will benefit from taking this into account to avoid the misidentification of brain volume changes in BD which are in reality artefacts due simply to the greater rate of obesity in patients.   The association of elevated BMI with neurobiological markers of disease severity and progression in BD also suggests new avenues of exploration regarding the pathophysiology of BD. Over the past five decades, the primary hypotheses that have been proposed to account for the symptoms and clinical course of the illness have generally emphasized structural and functional changes in limbic brain areas (175), and aberrations in the neurotransmitters that mediate activity in brain reward circuits, such as dopamine (214). While evidence from numerous clinical and brain imaging studies support these theories, it has become clear that they do not provide a complete explanation for all aspects of BD. The past 10-15 years has seen a veritable explosion of new, complementary theories, which propose central roles for neuroinflammation, oxidative stress, mitochondrial dysfunction, and alterations in neurotrophin levels in the pathophysiology of BD. It is noteworthy that aberrations in the normal functioning of many of these processes have been described in obese people, in ways that may be relevant to the neurobiology of BD.   Studies in healthy individuals have shown that the links between obesity, reduced brain volumes, and dementia persist even when vascular risk factors such as hypercholesterolemia and diabetes are controlled for (202,203), indicating that other mechanisms must be etiologically relevant. Many of the biochemicals whose serum levels are altered by obesity, such as cytokines, adipokines, neurotrophins such as brain-derived neurotrophic factor (BDNF), and oxidative stress markers, have pronounced effects on brain structure and function, perhaps none more so that inflammatory cytokines. One of the most exciting recent developments in BD research has been the understanding of the degree to which it is an inflammatory illness. Numerous studies have shown that the ratio of serum inflammatory to anti-inflammatory cytokines is markedly increased during manic episodes, and possibly depressive episodes, with serum inflammation tending to normalize during periods of recovery (215). Inflammation has thus been hypothesized 127  to be a major factor contributing to neuroprogression in BD (215,216), and chronic obesity-relayed inflammation, which persists even during recovery, is one possible cause for increased disease severity and progression in BD. This is a parsimonious hypothesis since neuroinflammation in fact leads to more rapid progression of many neuropsychiatric illnesses, including multiple sclerosis, Parkinson?s disease, and Alzheimer?s disease. Research into the inflammatory component of these illnesses is far advanced compared to BD, and it has been shown that periods of increased systemic inflammation, for example due to physical illnesses or stress, can precipitate relapses of multiple sclerosis and progression of Alzheimer?s disease (217). A similar role for inflammation in BD thus could provide a unifying thread between it and other neuropsychiatric illnesses with respect to causality, severity, and progression.   As reviewed above, adipose tissue is a key immune organ which produces cytokines such as TNF-?, IL-6, IL-10, and IL-18 (48). In normal-weight individuals, serum levels of inflammatory and anti-inflammatory cytokines are kept finely balanced, but in obese individuals the infiltration of adipose tissue with activated macrophages leads to chronically increased production of pro-inflammatory molecules such as TNF-? and IL-6, and reduced synthesis of anti-inflammatory molecules, such as IL-10 (50-54). Despite the presence of the blood-brain barrier (BBB), which theoretically creates an ?immune-privileged? state in the brain, serum inflammation impacts on brain immune functioning via a number of mechanisms: 1) the presence of saturable transport mechanisms for cytokines across the BBB; 2) the presence of cytokine receptors on the vagus nerve, which communicates information about the periphery of the body to the brain, including its inflammatory status; 3) receptors for cytokines and other inflammatory mediators on the epithelial cells that constitute the BBB; and 4) the absence of a functional BBB in the circumventricular organs, which permits unfettered passage of cytokines into the brain in these regions (218,219). Furthermore, the integrity of the BBB is compromised by inflammatory conditions such as obesity, leading to increased permeability to substances that would not normally be permitted entry into the brain (220). These are not just theoretical considerations: a positron emission tomography (PET) study in mice demonstrated that experimentally-induced obesity caused brain inflammation (221). The authors also replicated this finding in human subjects, but the sample size was small and the results thus preliminary. Given the association of 128  obesity with reduced WM in STOP-EM patients, it is particularly noteworthy that inflammatory cytokines have been associated with WM loss in both humans and animal models (222).   Two additional attractive features of the inflammatory hypothesis of BD deserve note. The first is that, unlike other theories that purport to explain the pathophysiology of BD, it provides a credible rationale for the finding that people with BD have increased morbidity and mortality from cardiovascular disease, stroke, and diabetes mellitus, and the converse, that comorbidity with these illnesses tends to worsen the course of BD. These medical conditions are also caused or worsened by inflammation, and thus inflammation secondary to BD - whether a direct result of mood episodes, or related to high obesity rates in patients - makes them more likely to manifest themselves, while their own inflammatory components conversely worsen the course of BD. It could also explain the otherwise paradoxical finding that although mood stabilizers predispose to weight gain and obesity, they actually lower mortality from comorbid medical conditions - this is likely due to direct anti-inflammatory properties of these medications, which may outweigh any indirect inflammation they cause by promoting obesity (223-226). The second attractive feature of the inflammatory hypothesis is of even broader relevance, as it offers an explanation, in fact a potential organizing principle, for the well-documented increase in prevalence and severity of mood disorders over the past several decades. Numerous long-term studies have reported a ?cohort effect? in BD, an increasing prevalence from one generation to the next (46). It may be argued that this is due to an ascertainment bias, a result of increased help-seeking behaviour among sufferers and increased awareness of psychiatric illnesses among clinicians, and this is no doubt a contributing factor. However, it is also true that many environmental factors that precipitate acute mood episodes and/or worsen the long-term course of BD, such as obesity, psychosocial stressors, sleep deprivation, and the aforementioned medical illnesses, all share an inflammatory component, and have increased in prevalence over the past century (227,228). To give one example, sleep deprivation has become markedly more common in a world of 24-7 business, entertainment, and news, and people in the western world now sleep on average one-and-a-half to two hours less per night than they did a century ago (229).    129  In addition to inflammation, other obesity-related biochemical aberrations, particularly involving molecules related to appetite control, such as leptin and ghrelin, are also highly relevant to BD. Until recently, these hormones were believed to interact primarily or exclusively with hypothalamic receptors that regulate feeding behaviour. The traditional understanding of food intake control posits a central role for the hypothalamus, especially its ventromedial and ventrolateral subdivisions. These areas receive neural input indirectly from the vagus nerve via the nucleus tractus solitarius, and also have receptors for hunger and satiety hormones, including leptin and ghrelin. They are believed to regulate appetite via their output, which originates in two distinct populations of neurons, one that produces the orexigenic peptides neuropeptide Y (NPY) and agouti-related peptide (AGRP), and the other that produces the anorexigenic peptides pro-opiomelanocortin (POMC) and cocaine and amphetamine regulated transcript (CART) (230).   While this view remains valid, it provides too limited a description of appetite regulation. Few, if any, biological functions are more important to an organism?s survival than procuring sufficient nutrition. This was even truer in the evolutionary past, when periods of famine were commonplace. Thus, large portions of the central nervous system play roles in regulating appetite, satiety, food-related rewards, and appetitive behaviors related to food procurement. Recently, the role of the limbic brain in regulating food-related rewards has come into sharp focus (231). If anything, this aspect of the central control of food intake has become increasingly relevant over time. One of the most profound changes in western society over the past four decades has been the degree to which food consumption patterns have changed, especially with respect to increased consumption of processed, sugar- and fat-laden foods that are designed to be highly palatable. In addition to the appetite-control mechanisms outlined above, the consumption of highly palatable foods strongly recruits brain reward circuits, including some of the areas most implicated in the pathophysiology of BD, such as the amygdala (emotional salience of food), hippocampus (location of food supplies), ventral striatum (food-related rewards), and frontal lobes (reconciling biological drives such as appetite with social contingencies, for example) (232). Mood-relevant neurochemicals such as dopamine are also involved (233).   Thus, there is broad overlap between brain areas recruited when procuring and consuming food, and particularly highly palatable foods, and those implicated in the pathophysiology of BD. In 130  keeping with this, adipokines have much more widespread effects on widespread limbic brain areas relevant to BD than was previously appreciated. Alterations in adipokine serum levels or functioning, such as reduced levels of ghrelin, a peptide hormone produced in the gut, and leptin resistance, are well documented in obese individuals (234). Receptors for leptin and ghrelin are widely distributed in the brain, including in the frontal and temporal lobes and the midbrain (235,236). Furthermore, emerging evidence suggests that these two molecules have neuroprotective effects (237,238), promote hippocampal plasticity (239,240), and modulate dopamine activity (241,242), all brain functions that are important in BD. Leptin also has direct effects on mitochondrial functioning, which has been shown to be altered in obesity and is also implicated in BD (50,243,244). This is particularly relevant to our current results, since NAA is produced primarily in the mitochondrial matrix from Krebs cycle intermediates (245,246). Alterations in all of these functions would be expected as a result of the leptin resistance and reduced ghrelin levels that characterize obesity. Thus, obesity may lie at the heart of the nexus of inflammation, mitochondrial dysfunction, and reward circuit malfunction which is increasingly understood to play a major role in the pathophysiology of BD (216,247). However, to date, only a small number of studies have examined serum adipokines in BD. They reported significant alterations in leptin and ghrelin levels in BD patients compared to healthy subjects, but they did not examine whether this was due mainly to higher rates of obesity in the patients, or alternatively whether obesity had a greater impact on serum adipokines in BD (248,249).  Finally, obesity also leads to altered production of a number of other serum biochemicals, including elevated levels of reactive oxygen species such as peroxynitrite (56) which cause oxidative damage to DNA, proteins, and lipids and which may be central to the pathophysiology of BD (250), and reduced serum BDNF levels. BDNF plays a critical role in promoting neuronal health and survival, and in synaptic plasticity, and low BDNF levels are implicated in the pathophysiology of a number of psychiatric illnesses, including BD and schizophrenia. Alterations in BDNF in BD appear to be mood-state-specific, with low levels during manic and depressive episodes, and normalization during euthymia (251). Similarly to inflammatory cytokines, obesity is associated with chronic alterations in BDNF levels, which, if they persist during euthymia in BD patients, might partly explain the greater disease severity in obese individuals. Interestingly, given our BMI-related white matter findings in patients, reduced 131  BDNF levels have been associated with WM volume reductions in neurological illnesses such as MS (252). Thus, our findings might be indicative of altered cytokine, adipokine, and/or neurotrophin activity in BD compared to healthy subjects. The role of obesity-related serum factors on brain structure and function in BD clearly warrants further attention.   7.4  Limitations  The naturalistic design of STOP-EM creates a number of shortcomings, particularly with respect to assigning casualty to the weight-related neurobiological changes we detected, and these were reviewed above. In addition, the broad inclusion criteria and limited exclusion criteria for the program may be viewed as both a strength and a limitation. The enrolment of people with comorbid conditions, including other psychiatric illnesses and drug abuse, permitted the ascertainment of a clinically relevant and highly generalizable first-episode mania sample. Nonetheless, the presence of these other conditions, each of which has its own neurobiological basis that partly overlaps with and is partly distinct from BD, created ?noise? in our data with respect to examining the pathophysiology of BD. The same is also true of the fact that this was a treated sample. As noted above, pharmacologic treatment of BD tends to have normalizing effect on brain changes, frequently obscuring illness-specific neurobiological findings. For these reasons, we statistically controlled for the presence of the most important confounding variables, particularly when they were correlated with our outcomes of interest.  Despite the efforts to make STOP-EM patients as reflective as possible of the population of people with first-episode mania, that fact that patients were recruited and assessed at an urban academic medical centre introduces obvious generalizability issues. Furthermore, the sample may not be fully representative of all North American patients with BD with respect to weight, since Canada has a lower rate of obesity than the United States, and Vancouver has the lowest obesity rate of any Canadian city (115,148). Also, since rates of overweight and obesity in our sample were low, we combined them in both patients and healthy subjects, and we thus could not examine their effects separately. Additional limitations include the fact that patients were enrolled following recovery from their first manic episode, meaning that we did not have information on weight, brain volumes, or medication use during treatment of the acute mania or 132  related to prior depressive or hypomanic episodes. Much of the sociodemographic and clinical data that was collected at enrolment, for example regarding previous course of mood illness, was ascertained retrospectively.   With respect to our neurobiological outcomes of interest, we did not have complete weight or brain volume data for all of the patients, and we cannot exclude the possibility that the missing values might have affected our findings. We did not collect data on diet or exercise, which might have independent effects on brain structure and function in BD. We cannot exclude the possibility that confounding variables that were present in patients but not healthy subjects, particularly the use of mood stabilizing and antipsychotic medications, contributed to differences between patients and healthy subjects in BMI-related brain neurobiological changes. In our VBM analysis, the assignment of cortical labels and white matter tracts to the clusters of reduced GM and WM volume we identified was probabilistic. Finally, we may have had inadequate statistical power and study duration to detect an association between weight gain and more subtle brain changes. Nonetheless, the fact that we were able to detect weight-related brain changes despite the presence of these study limitations, and the multiple demographic and psychosocial determinants of brain volume, speaks to both the soundness of our methods and the robustness of our findings.  7.5  Relevance and future research directions  In conclusion, the research presented here provides a plausible neurobiological rationale for a commonly observed clinical phenomenon: the association of obesity with a more severe psychiatric illness in people with BD. In so doing, it suggests a possible template for examining how the presence of other comorbid conditions, such as anxiety disorders, substance abuse, and ADHD, also lead to worse illness courses in BD. Finally, the results of this study potentially have implications for understanding the pathophysiology of a broad spectrum of psychiatric illnesses, given the high obesity rates in other diseases such as schizophrenia and major depressive disorder (89,253).  133  There is obvious interest in further elucidating weight-related neurobiological changes in BD. As a starting point, replicating the findings presented here in separate samples would increase confidence in their validity. It would be of particular interest to determine if they also hold true in non-first-episode BD samples. In addition, I have already begun to pursue several lines of inquiry which logically arise from the results that I have obtained so far, and these are outlined in the sections below.   7.5.1  Mediators in the relationship between BMI and brain structure and function  The mechanisms that mediate the relationship between increased BMI and decrements in brain structure and function in BD are poorly understood. Nonetheless, elucidating them will aid in comprehending the pathophysiology of the illness, and may also suggest novel therapeutic strategies to prevent brain damage related to weight gain.   In the general population, the link between obesity and neuropathology persists even when vascular risk factors such as hypertension, hypercholesterolemia, hyperlipidemia, and diabetes are controlled for (202,203), suggesting that it is not solely due to obesity-related medical illnesses, and that other factors are etiologically relevant. What might these factors be? As reviewed in the sections above, there is a growing understanding that obese individuals have alterations in the levels of numerous serum biochemicals, many of which have known effects on brain structure and function (48). In particular, serum levels of pro-inflammatory adipokines, such as leptin and resistin, are elevated, and serum levels of anti-inflammatory adipokines, such as adiponectin, are reduced in obese individuals (51-53). Obesity also leads to increased production of reactive oxygen species such as peroxynitrite (56) which cause oxidative damage to DNA, proteins, and lipids. Finally, serum levels of brain derived neurotrophic factor (BDNF) and ghrelin, a peptide hormone produced in the gut, are decreased in obese individuals (234,254), while cortisol activity is elevated (255).   However, to date, only a very small literature has investigated these serum biomarkers in BD (248,249), and these studies were limited by small sample sizes, and did not examine for an interaction between diagnosis (patient vs. healthy subject) and BMI (overweight/obese vs. 134  normal weight) in serum levels of these factors. Preliminary data from our first-episode mania patients suggests that there is a negative correlation between BMI and serum levels of both BDNF and the anti-inflammatory cytokine IL-10 (DJ Bond, unpublished observation), but this is based on only a subset of our patients, and we have not yet examined these correlations in our healthy subjects. Based on our previous results showing a diagnosis x BMI interaction for regional brain volumes and hippocampal metabolites, I hypothesize that there will be similar interactions for serum levels of adipokines, cytokines, BDNF, and cortisol and for mitochondrial functioning, and that these interactions may constitute the mediating factors between elevated BMI and brain changes in BD patients.  Therefore, an analysis is currently underway to examine the interaction between obesity and diagnosis in determining serum levels of these serum biomarkers in STOP-EM patients and healthy subjects. To date, we have obtained blood samples from 56 patients and 23 healthy subjects. Twenty-three patients and 11 healthy subjects were overweight or obese at the time the samples were collected, while the 33 patients and 12 patients had normal weights.   I will carry out assays of the following serum factors and assays of mitochondrial functioning:  Adipokines ? Leptin ? Ghrelin ? Resistin ? Adiponectin ? Vasoactive intestinal peptide (VIP) ? Tumour necrosis factor-alpha (TNF-?) ? Visfatin ? IGF-1 ? NLRP3 ? RBP4 ? ANGPLT2 ? CXCL5 135  ? NAMPT  Cytokines ? IL-1 ? IL-6 ? IL-10 ? IL-18  Neurotrophins ? BDNF  HPA Axis Functioning ? Cortisol  Mitochondrial Function ? Mitochondrial complex I-IV activation.  ? Expression of mitochondrial subunits (measured using real time PCR (mRNA) and protein levels) ? DNA methylation in mitochondrial electron transport chain subunit genes.  My primary outcome is to determine whether elevated BMI has different effects on these serum factors and measures of mitochondrial functioning in patients with BD compared to healthy subjects. As in the analyses in Chapters 2-6,, rates of overweight and obesity are low, and they will thus be combined so that patients with overweight or obesity were compared to those with normal weight, and similarly for healthy subjects. My primary analysis will be a factorial MANCOVA with the serum levels of each of the above molecules/mitochondrial measures as the dependent variables, and diagnosis (patient vs. healthy subject) and BMI (overweight/obese vs. normal weight) as factors. Age and gender will be entered as covariates. I will conduct post-hoc analyses for serum levels with significant results, directly comparing levels in overweight/obese patients to normal-weight patients, and in overweight/obese healthy subjects to normal-weight healthy subjects. For exploratory purposes, I will also create linear regression 136  models to examine whether serum levels of the above molecules are differentially associated with brain volumes in BD patients and healthy subjects. The dependent variables will be the brain volumes of interest, including total brain volume, grey matter volume, white matter volume, and frontal, parietal, occipital, and temporal lobe volumes. Predictors entered into the regression models will include age, gender, diagnosis, the serum level of the molecule of interest, and a diagnosis x serum level interaction term. We hope that this will assist us in generating hypotheses regarding obesity-related serum factors and brain volumes that may be tested in future studies.  This will be the first study to examine the interaction between diagnosis and BMI on BMI-related serum factors and mitochondrial function in BD, or any psychiatric illness. I anticipate that my results will help illuminate the biological factors that mediate the relationship between elevated BMI and the unique neurobiological changes we have detected in overweight/obese BD patients. Furthermore, an expanding evidence base suggests that adipokines and cytokines impact on the structure and function of brain reward circuits relevant to BD, and this will be the first study to examine these relationships directly in BD patients  7.5.2  Randomized clinical trial of a weight loss intervention in people with BD  The most important question regarding the BMI-related neurobiological changes I have found is whether they are in fact caused by obesity. A causal relationship between obesity and brain changes has been demonstrated in numerous animal studies, in which rats or primates randomly assigned to obesigenic conditions showed decrements in brain volume and function compared to animals randomized to non-weight-gain study arms (95-98). However, it is not yet apparent that this is also the case in BD, primarily because the naturalistic study design employed by STOP-EM allowed me to demonstrate only an association between obesity and brain changes.   Ethical and practical considerations preclude randomly assigning people to obesigenic conditions to assess the impact of weight gain on brain structure and function, but randomly assigning already-overweight/obese patients to weight loss treatments provides an alternative approach to assessing causality. A single proof-of-principle study in humans without psychiatric illnesses has 137  already suggested that obesity-related brain changes may be reversible with weight loss, though it had a number of significant limitations, including a non-randomized design, a small sample size, and a very intensive weight-loss regimen that would be difficult to implement in clinical practice (256). Nonetheless, if BMI-related neurobiological changes in BD patients were slowed or reversed by weight loss, this would definitively establish a causal relationship between weight and brain changes. Surprisingly, given the high obesity rates in people with BD, the evidence base for weight-loss interventions in people with BD is limited. However, the feasibility of weight loss treatments in people with serious mental illnesses has been convincingly demonstrated by multiple placebo-controlled trials in schizophrenia, which have shown that several lifestyle interventions and medications, alone or in combination, are effective in reducing weight in this population, including in people treated with obesigenic medications (257,258). The available preliminary data in BD suggest promise for a number of these same interventions. In addition, one small study assessing the benefits of weight loss in BD provided preliminary data suggesting that it improved psychiatric outcomes, as patients who lost weight had reduced depressive symptoms and improved psychosocial functioning (259).   I therefore propose to carry out a 52-week randomized clinical trial in which overweight/obese patients with BD  will be randomly assigned to either treatment as usual (TAU), or TAU + a weight loss program consisting of diet and exercise counselling and the hypoglycemic medication metformin 500-2000 mg/day. I hypothesize that compared to patients who receive TAU alone, patients who receive the adjunctive weight-loss intervention will lose significantly more weight, experience less brain volume loss and less pronounced alterations in neurochemistry in frontal and temporal limbic areas, spend less time depressed, and have better psychosocial functioning. Combination treatment with diet/exercise management and metformin was chosen as the intervention because it is the most efficacious weight-loss treatment in people with schizophrenia (260), with a mean weight loss approximately equal to the mean weight gain experienced by our first-episode mania patients in 1 year of maintenance treatment, and which was associated with significant brain volume loss over that time period.  The protocol for this trial will specify that BD patients who are receiving treatment with ?1 standard medication for BD, are overweight/ obese, and are euthymic for at least 4 weeks can be 138  enrolled. Half of the patients will be randomly assigned to TAU, and the other half to TAU + the intervention. Patients will undergo MRI and MRS scans at baseline, 26 weeks, and 52 weeks to measure GMV, WMV, and NAA and glutatmate and NAA levels in prefrontal and temporal limbic areas implicated in BD. As secondary outcomes, we will gather important clinical information regarding the amount of time patients spend with mania and depression, and their level of psychosocial functioning. We will also collect serum samples to measure possible mediators of weight-related brain changes, including inflammatory cytokines and adipokines, as well predictors of obesity-related medical complications, such as serum glucose, cholesterol, and triglycerides. The randomized design of our trial will allow us to establish whether weight change causes neurobiological alterations in people with BD. Our primary outcomes, structural and chemical brain changes measured using MRI and MRS, would not be expected to vary based on patient or clinician expectation, negating the need for a double-blinded design, which would be difficult to implement given the diet and exercise intervention we propose.   A grant application and protocol for the proposed trial have already been written. The results of this study will have immediate implications for understanding the neurobiology of BD, and clinical implications for treating the illness. From a biological point of view, if my results demonstrate that patients randomly assigned to the weight-loss arm experience less brain volume loss and less progressive neurochemical brain changes in limbic brain areas, this will establish a causal relationship between weight and neuroprogression in BD. Clinically, if these patients also experience fewer relapses and improved psychosocial functioning, this will provide a powerful incentive to clinicians and patients to incorporate weight management as a core treatment for the affective component of BD. If, however, my results demonstrate that structural and chemical brain changes do not differ between patients in the weight loss and TAU arms, this will support the hypothesis that there is a subtype of BD characterized by both a more neurobiologically progressive illness and a propensity for weight gain, and confirm that weight change does not cause illness progression in BD. Patients might nonetheless experience improvements in relapse rates and psychosocial functioning, and if so this would suggest that weight change affects psychiatric symptoms through other mechanisms, such as improved self-esteem and reduced obesity-related stigma, opening up areas for future research. Determining its impact on 139  psychiatric symptoms and neurobiological outcomes will allow clinicians to provide patients considering weight loss treatments with a realistic assessment of the likely benefits of this. 140  References  1.   Schaffer A, Cairney J, Cheung A, Veldhuizen S, Levitt A. Community survey of bipolar disorder in Canada: Lifetime prevalence and illness characteristics. Can J Psychiatry 2006; 51:9-16 2.   Perlis RH, Miyahara S, Marangell LB, Wisniewski SR, Ostacher M, DelBello MP, Bowden CL, Sachs GS, Nierenberg AA, STEP-BD Investigators. Long-term implications of early onset in bipolar disorder: Data from the first 1000 participants in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). Biol Psychiatry 2004; 55:875-881 3.   Judd LL, Akiskal HS, Schettler PJ, Endicott J, Maser J, Solomon DA, Leon AC, Rice JA, Keller MB. The long-term natural history of the weekly symptomatic status of bipolar I disorder. Arch Gen Psychiatry 2002; 59:530-537 4.   Judd LL, Schettler PJ, Solomon DA, Maser JD, Coryell W, Endicott J, Akiskal HS. Psychosocial disability and work role function compared across the long-term course of bipolar I, bipolar II and unipolar major depressive disorders. J Affect Disord 2008; 108:49-58 5.   Murray CJL, Lopez AD. The global burden of disease: A comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020. Edited by Anonymous Cambridge, Harvard University Press, 1996 6.   Dilsaver SC. An estimate of the minimum economic burden of bipolar I and II disorders in the united states: 2009. J Affect Disord 2011; 129:79-83 7.   Nguyen DM, El-Serag HB. The epidemiology of obesity. Gastroenterol Clin North Am 2010; 39:1-7 8.   Shields M, Carroll MD, Ogden CL. Adult obesity prevalence in Canada and the United States. NCHS Data Brief 2011; 56:1-8 141  9.   Misra A, Kurana L. Obesity and the metabolic syndrome in developing countries. J Clin Endocrinol Metab 2008; 93(11 Suppl 1):S9-S30 10.   World Health Organization. Obesity and overweight (fact sheet 311). Available at: Http://www Who int/mediacentre/factsheets/fs311/en/index Html. Accessed March 1, 2012;  11.   Field AE, Coakley EH, Must A, Spadano JL, Laird N, Dietz WH, Rimm E, Colditz GA. Impact of overweight on the risk of developing common chronic diseases during a 10-year period. Arch Intern Med 2001; 161:1581-1586 12.   Homel P, Casey D, Allison DB. Changes in body mass index for individuals with and without schizophrenia, 1987-1996. Schizophr Res 2002; 55:277-284 13.   de Wit L, Luppino F, van Straten A, Penninx B, Zitman F, Cuijpers P. Depression and obesity: A meta-analysis of community-based studies. Psychiatry Res 2010; 178:230-235 14.   McElroy SL, Frye MA, Suppes T, Dhavale D, Keck PE, Leverich GS, Altshuler L, Denicoff KD, Nolen WA, Kupka R, Grunze H, Walden J, Post RM. Correlates of overweight and obesity in 644 patients with bipolar disorder. J Clin Psychiatry 2002; 63:207-213 15.   Simon GE, Von Korff M, Saunders K, Miglioretti DL, Crane PK, van Belle G, Kessler RC. Association between obesity and psychiatric disorders in the US adult population. Arch Gen Psychiatry 2006; 63:824-830 16.   Fagiolini A, Frank E, Houck PR, Mallinger AG, Swartz HA, Buysse DJ, Ombao H, Kupfer DJ. Prevalence of obesity and weight change during treatment in patients with bipolar I disorder. J Clin Psychiatry 2002; 63:528-533 17.   Fagiolini A, Kupfer DJ, Houck PR, Novick DM, Frank E. Obesity as a correlate of outcome in patients with bipolar I disorder. Am J Psychiatry 2003; 160:112-117 142  18.   Fagiolini A, Frank E, Scott JA, Turkin S, Kupfer DJ. Metabolic syndrome in bipolar disorder: Findings from the bipolar disorder center for Pennsylvanians. Bipolar Disord 2005; 7:424-430 19.   Shah A, Shen N, El-Mallakh RS. Weight gain occurs after onset of bipolar illness in overweight bipolar patients. Ann Clin Psychiatry 2006; 18:239-241 20.   Fiedorowicz JG, Palagummi NM, Forman-Hoffman VL, Miller del D, Haynes WG. Elevated prevalence of obesity, metabolic syndrome, and cardiovascular risk factors in bipolar disorder. Ann Clin Psychiatry 2008; 20:131-137 21.   Goldstein BI, Liu SM, Zivkovic N, Schaffer A, Chien LC, Blanco C. The burden of obesity among adults with bipolar disorder in the united states. Bipolar Disord 2011; 13:387-395 22.   Sicras A, Rejas J, Navarro R, Serrat J, Blanca M. Metabolic syndrome in bipolar disorder: A cross-sectional assessment of a health management organization database. Bipolar Disord 2008; 10:607-616 23.   Birkenaes AB, Opjordsmoen S, Brunborg C, Engh JA, Jonsdottir H, Ringen PA, Simonsen C, Vaskinn A, Birkeland KI, Friis S, Sundet K, Andreassen OA. The level of cardiovascular risk factors in bipolar disorder equals that of schizophrenia: A comparative study. J Clin Psychiatry 2007; 68:917-923 24.   Kim B, Kim SJ, Son JI, Joo YH. Weight change in the acute treatment of bipolar I disorder: A naturalistic observational study of psychiatric inpatients. J Affect Disord 2008; 105:45-52 25.   Lee NY, Kim SH, Cho B, Lee YJ, Chang JS, Kang UG, Kim YS, Ahn YM. Patients taking medications for bipolar disorder are more prone to metabolic syndrome than Korea's general population. Prog Neuropsychopharmacol Biol Psychiatry 2010; 34:1243-1249 143  26.   Elmslie JL, Silverstone JT, Mann JI, Williams SM, Romans SE. Prevalence of overweight and obesity in bipolar patients. J Clin Psychiatry 2000; 61:179-184 27.   Wang PW, Sachs GS, Zarate CA, Marangell LB, Calabrese JR, Goldberg JF, Sagduyu K, Miyahara S, Ketter TA. Overweight and obesity in bipolar disorders. J Psychiatr Res 2006; 40:762-764 28.   Goldstein BI, Birmaher B, Axelson DA, Goldstein TR, Esposito-Smythers C, Strober MA, Hunt J, Leonard H, Gill MK, Iyengar S, Grimm C, Yang M, Ryan ND, Keller MB. Preliminary findings regarding overweight and obesity in pediatric bipolar disorder. J Clin Psychiatry 2008; 69:1953-1959 29.   Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel III). JAMA 2001; 285:2486-2497 30.   Lakka HM, Laaksonen DE, Lakka TA, Niskanen LK, Kumpusalo E, Tuomilehto J, Salonen JT. The metabolic syndrome and total and cardiovascular disease mortality in middle-aged men. JAMA 2002; 288:2709-2716 31.   Kurl S, Laukkanen JA, Niskanen L, Laaksonen D, Sivenius J, Nyyssonen K, Salonen JT. Metabolic syndrome and the risk of stroke in middle-aged men. Stroke 2006; 37:806-811 32.   Ford ES, Li C, Sattar N. Metabolic syndrome and incident diabetes: Current state of the evidence. Diabetes Care 2008; 31:1898-1904 33.   Cardenas J, Frye MA, Marusak SL, Levander EM, Chirichigno JW, Lewis S, Nakelsky S, Hwang S, Mintz J, Altshuler LL. Modal subcomponents of metabolic syndrome in patients with bipolar disorder. J Affect Disord 2008; 106:91-97 144  34.   McIntyre RS, Woldeyohannes HO, Soczynska JK, Miranda A, Lachowski A, Liauw SS, Grossman T, Lourenco MT, Kim B, Alsuwaidan MT, Kennedy SH. The rate of metabolic syndrome in euthymic Canadian individuals with bipolar I/II disorder. Adv Ther 2010; 27:828-836 35.   Garcia-Portilla MP, Saiz PA, Benabarre A, Sierra P, Perez J, Rodriguez A, Livianos L, Torres P, Bobes J. The prevalence of metabolic syndrome in patients with bipolar disorder. J Affect Disord 2008; 106:197-201 36.   Salvi V, Albert U, Chiarle A, Soreca I, Bogetto F, Maina G. Metabolic syndrome in Italian patients with bipolar disorder. Gen Hosp Psychiatry 2008; 30:318-323 37.   Salvi V, D'Ambrosio V, Bogetto F, Maina G. Metabolic syndrome in Italian patients with bipolar disorder: A 2-year follow-up study. J Affect Disord 2012; 136:599-603 38.   Yumru M, Savas HA, Kurt E, Kaya MC, Selek S, Savas E, Oral ET, Atagun I. Atypical antipsychotics related metabolic syndrome in bipolar patients. J Affect Disord 2007; 98:247-252 39.   Chang HH, Chou CH, Chen PS, Gean PW, Huang HC, Lin CY, Yang YK, Lu RB. High prevalence of metabolic disturbances in patients with bipolar disorder in Taiwan. J Affect Disord 2009; 117:124-129 40.   Guan N, Liu H, Diao F, Zhang J, Zhang M, Wu T. Prevalence of metabolic syndrome in bipolar patients initiating acute-phase treatment: A 6-month follow up. Psychiatry Clin Neurosci 2010; 64:625-633 41.   Ezzaher A, Haj MD, Mechri A, Neffati F, Douki W, Gaha L, Najjar MF. Metabolic syndrome in Tunisian bipolar I patients. Afr Health Sci 2011; 11:414-420 145  42.   Grover S, Aggarwal M, Chakrabarti S, Dutt A, Avasthi A, Kulhara P, Malhotra N, Somaiya M, Chauhan N. Prevalence of metabolic syndrome in bipolar disorder: An exploratory study from north India. Prog Neuropsychopharmacol Biol Psychiatry 2012; 36:141-146 43.   Maina G, Salvi V, Vitalucci A, D'Ambrosio V, Bogetto F. Prevalence and correlates of overweight in drug-na?ve patients with bipolar disorder. J Affect Disord 2008; 110:149-155 44.   Sobczak S, Honig A, Christophe A, Maes M, Helsdingen RW, De Vriese SA, Riedel WJ. Lower high-density lipoprotein cholesterol and increased omega-6 polyunsaturated fatty acids in first-degree relatives of bipolar patients. Psychol Med 2004; 34:103-112 45.   Torrent C, Amann B, S?nchez-Moreno J, Colom F, Reinares M, Comes M, Rosa AR, Scott J, Vieta E. Weight gain in bipolar disorder: Pharmacological treatment as a contributing factor. Acta Psychiatr Scand 2008; 118:4-18 46.   Goodwin FK, Jameson KR. Manic Depressive Illness: Bipolar Disorders and Recurrent Depression. New York, Oxford University Press, 2007 47.   Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Positional cloning of the mouse obese gene and its human homologue. Nature 1994; 372:425-432 48.   Ouchi N, Parker JL, Lugus JJ, Walsh K. Adipokines in inflammation and metabolic disease. Nat Rev Immunol 2011; 11:85-97 49.   Rajala MW, Scherer PE. Minireview: The adipocyte--at the crossroads of energy homeostasis, inflammation, and atherosclerosis. Endocrinology 2003; 144:3765-3773 50.   Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante AW Jr. Obesity is associated with macrophage accumulation in adipose tissue. J Clin Invest 2003; 112:1796-1808 51.   Steppan CM, Lazar MA. The current biology of resistin. J Intern Med 2004; 255:439-447 146  52.   Friedman JM, Halaas JL. Leptin and the regulation of body weight in mammals. Nature 1998; 395:763-770 53.   Ryo M, Nakamura T, Kihara S, Kumada M, Shibazaki S, Takahashi M, Nagai M, Matsuzawa Y, Funahashi T. Adiponectin as a biomarker of the metabolic syndrome. Circ J 2004; 68:975-981 54.   Donath MY, Shoelson SE. Type 2 diabetes as an inflammatory disease. Nat Rev Immunol 2011; 11:98-107 55.   Dentali F, Squizzato A, Ageno W. The metabolic syndrome as a risk factor for venous and arterial thrombosis. Semin Thromb Hemost 2009; 35:451-457 56.   Fernandez-Sanchez A, Madrigal-Santillan E, Bautista M, Esquivel-Soto J, Morales-Gonzalez A, Esquivel-Chirino C, Durante-Montiel I, Sanchez-Rivera G, Valadez-Vega C, Morales-Gonzalez JA. Inflammation, oxidative stress, and obesity. Int J Mol Sci 2011; 12:3117-3132 57.   Zhang C. The role of inflammatory cytokines in endothelial dysfunction. Basic Res Cardiol 2008; 103:398-406 58.   Viazzi F, Leoncini G, Adami GF, Papadia FS, Bezante GP, Conti N, Baratto E, Scopinaro N, Deferrari G, Pontremoli R. Sub-clinical organ damage in hypertension and obesity. Nutr Metab Cardiovasc Dis 2011; 21:597-602 59.   Di Bonito P, Moio N, Scilla C, Cavuto L, Sibilio G, Forziato C, Sanguigno E, Saitta F, Iardino MR, Capaldo B. Preclinical manifestations of organ damage associated with the metabolic syndrome and its factors in outpatient children. Atherosclerosis 2010; 213:611-615 60.   Perego F, Renesto E, Arquati M, Scandiani L, Cogliati C, Torzillo D, Zocchi L, Casazza G, Duca P, Chirchiglia S, Lacaita G, Panteghini M, Cortellaro M. Target organ damage in a 147  population at intermediate cardiovascular risk, with adjunctive major risk factors: CArdiovascular PREvention sacco study (CAPRESS). Intern Emerg Med 2011; 6:337-347 61.   Hotamisligil GS, Shargill NS, Spiegelman BM. Adipose expression of tumor necrosis factor-alpha: Direct role in obesity-linked insulin resistance. Science 1993; 259:87-91 62.   Uysal KT, Wiesbrock SM, Marino MW, Hotamisligil GS. Protection from obesity-induced insulin resistance in mice lacking TNF-alpha function. Nature 1997; 389:610-614 63.   Guh DP, Zhang W, Bansback N, Amarsi Z, Birmingham CL, Anis AH. The incidence of co-morbidities related to obesity and overweight: A systematic review and meta-analysis. BMC Public Health 2009; 9:88 64.   Prospective Studies Collaboration, Whitlock G, Lewington S, Sherliker P, Clarke R, Emberson J, Halsey J, Qizilbash N, Collins R, Peto R. Body-mass index and cause-specific mortality in 900 000 adults: Collaborative analyses of 57 prospective studies. Lancet 2009; 373:1083-1096 65.   Berrington de Gonzalez A, Hartge P, Cerhan JR, Flint AJ, Hannan L, MacInnis RJ, Moore SC, Tobias GS, Anton-Culver H, Freeman LB, Beeson WL, Clipp SL, English DR, Folsom AR, Freedman DM, Giles G, Hakansson N, Henderson KD, Hoffman-Bolton J, Hoppin JA, Koenig KL, Lee IM, Linet MS, Park Y, Pocobelli G, Schatzkin A, Sesso HD, Weiderpass E, Willcox BJ, Wolk A, Zeleniuch-Jacquotte A, Willett WC, Thun MJ. Body-mass index and mortality among 1.46 million white adults. N Engl J Med 2010; 363:2211-2219 66.   Carney CP, Jones LE. Medical comorbidity in women and men with bipolar disorders: A population-based controlled study. Psychosom Med 2006; 68:684-691 148  67.   Johannessen L, Strudsholm U, Foldager L, Munk-Jorgensen P. Increased risk of hypertension in patients with bipolar disorder and patients with anxiety compared to background population and patients with schizophrenia. J Affect Disord 2006; 95:13-17 68.   Lilliker SL. Prevalence of diabetes in a manic-depressive population. Compr Psychiatry 1980; 21:270-275 69.   Cassidy F, Ahearn E, Carroll BJ. Elevated frequency of diabetes mellitus in hospitalized manic-depressive patients. Am J Psychiatry 1999; 156:1417-1420 70.   Regenold WT, Thapar RK, Marano C, Gavirneni S, Kondapavuluru PV. Increased prevalence of type 2 diabetes mellitus among psychiatric inpatients with bipolar I affective and schizoaffective disorders independent of psychotropic drug use. J Affect Disord 2002; 70:19-26 71.   van Winkel R, De Hert M, Van Eyck D, Hanssens L, Wampers M, Scheen A, Peuskens J. Prevalence of diabetes and the metabolic syndrome in a sample of patients with bipolar disorder. Bipolar Disord 2008; 10:342-348 72.   Osby U, Brandt L, Correia N, Ekbom A, Spar?n P. Excess mortality in bipolar and unipolar disorder in Sweden. Arch Gen Psychiatry 2001; 58:844-850 73.   Laursen TM. Life expectancy among persons with schizophrenia or bipolar affective disorder. Schizophr Res 2011; 131:101-104 74.   Chang CK, Hayes RD, Perera G, Broadbent MT, Fernandes AC, Lee WE, Hotopf M, Stewart R. Life expectancy at birth for people with serious mental illness and other major disorders from a secondary mental health care case register in London. PLoS One 2011; 6:e19590 75.   Angst F, Stassen HH, Clayton PJ, Angst J. Mortality of patients with mood disorders: Follow-up over 34-38 years. J Affect Disord 2002; 68:167-181 149  76.   Ward MA, Carlsson CM, Trivedi MA, Sager MA, Johnson SC. The effect of body mass index on global brain volume in middle-aged adults: A cross sectional study. BMC Neurol 2005; 5:23 77.   Taki Y, Kinomura S, Sato K, Inoue K, Goto R, Okada K, Uchida S, Kawashima R, Fukuda H. Relationship between body mass index and gray matter volume in 1,428 healthy individuals. Obesity 2008; 16:119-124 78.   Gunstad J, Paul RH, Cohen RA, Tate DF, Spitznagel MB, Grieve S, Gordon E. Relationship between body mass index and brain volume in healthy adults. Int J Neurosci 2008; 118:1582-1593 79.   Soreca I, Rosano C, Jennings JR, Sheu LK, Kuller LH, Matthews KA, Aizenstein HJ, Gianaros PJ. Gain in adiposity across 15 years is associated with reduced gray matter volume in healthy women. Psychosom Med 2009; 71:485-490 80.   Walther K, Birdsill AC, Glisky EL, Ryan L. Structural brain differences and cognitive functioning related to body mass index in older females. Hum Brain Mapp 2010; 31:1052-1064 81.   Enzinger C, Fazekas F, Matthews PM, Ropele S, Schmidt H, Smith S, Schmidt R. Risk factors for progression of brain atrophy in aging: Six-year follow-up of normal subjects. Neurology 2005; 64:1704-1711 82.   Gunstad J, Paul RH, Cohen RA, Tate DF, Spitznagel MB, Gordon E. Elevated body mass index is associated with executive dysfunction in otherwise healthy adults. Compr Psychiatry 2007; 48:57-61 83.   Cserj?si R, Luminet O, Poncelet AS, L?n?rd L. Altered executive function in obesity. exploration of the role of affective states on cognitive abilities. Appetite 2009; 52:535-539 150  84.   Fergenbaum JH, Bruce S, Lou W, Hanley AJ, Greenwood C, Young TK. Obesity and lowered cognitive performance in a Canadian first nations population. Obesity 2009; 17:1957-1963 85.   Cournot M, Marqui? JC, Ansiau D, Martinaud C, Fonds H, Ferri?res J, Ruidavets JB. Relation between body mass index and cognitive function in healthy middle-aged men and women. Neurology 2006; 67:1208-1214 86.   Boeka AG, Lokken KL. Neuropsychological performance of a clinical sample of extremely obese individuals. Arch Clin Neuropsychol 2008; 23:467-474 87.   Gunstad J, Paul RH, Cohen RA, Tate DF, Gordon E. Obesity is associated with memory deficits in young and middle-aged adults. Eat Weight Disord 2006; 11:e15-19 88.   Siervo M, Arnold R, Wells JC, Tagliabue A, Colantuoni A, Albanese E, Brayne C, Stephan BC. Intentional weight loss in overweight and obese individuals and cognitive function: A systematic review and meta-analysis. Obes Rev 2011; 12:968-983 89.   Luppino FS, de Wit LM, Bouvy PF, Stijnen T, Cuijpers P, Penninx BW, Zitman FG. Overweight, obesity, and depression: A systematic review and meta-analysis of longitudinal studies. Arch Gen Psychiatry 2010; 67:220-229 90.   Anstey KJ, Cherbuin N, Budge M, Young J. Body mass index in midlife and late-life as a risk factor for dementia: A meta-analysis of prospective studies. Obes Rev 2011; 12:e426-37 91.   Munger KL, Chitnis T, Ascherio A. Body size and risk of MS in two cohorts of US women. Neurology 2009; 73:1543-1550 92.   Hedstrom AK, Olsson T, Alfredsson L. High body mass index before age 20 is associated with increased risk for multiple sclerosis in both men and women. Mult Scler 2012;  151  93.   Abbott RD, Ross GW, White LR, Sanderson WT, Burchfiel CM, Kashon M, Sharp DS, Masaki KH, Curb JD, Petrovitch H. Environmental, life-style, and physical precursors of clinical Parkinson's disease: Recent findings from the Honolulu-Asia aging study. J Neurol 2003; 250 Suppl 3:III30-9 94.   Hu G, Jousilahti P, Nissinen A, Antikainen R, Kivipelto M, Tuomilehto J. Body mass index and the risk of Parkinson disease. Neurology 2006; 67:1955-1959 95.   Pedrini S, Thomas C, Brautigam H, Schmeidler J, Ho L, Fraser P, Westaway D, Hyslop PS, Martins RN, Buxbaum JD, Pasinetti GM, Dickstein DL, Hof PR, Ehrlich ME, Gandy S. Dietary composition modulates brain mass and solubilizable a-beta levels in a mouse model of aggressive Alzheimer's amyloid pathology. Mol Neurodegener 2009; 4:40 96.   Colman RJ, Anderson RM, Johnson SC, Kastman EK, Kosmatka KJ, Beasley TM, Allison DB, Cruzen C, Simmons HA, Kemnitz JW, Weindruch R. Caloric restriction delays disease onset and mortality in rhesus monkeys. Science 2009; 325:201-204 97.   Farr SA, Yamada KA, Butterfield DA, Abdul HM, Xu L, Miller NE, Banks WA, Morley JE. Obesity and hypertriglyceridemia produce cognitive impairment. Endocrinology 2008; 149:2628-2636 98.   Yamada-Goto N, Katsuura G, Ochi Y, Ebihara K, Kusakabe T, Hosoda K, Nakao K. Impairment of fear-conditioning responses and changes of brain neurotrophic factors in diet-induced obese mice. J Neuroendocrinol 2012;  99.   Kupka RW, Altshuler LL, Nolen WA, Suppes T, Luckenbaugh DA, Leverich GS, Frye MA, Keck PE Jr, McElroy SL, Grunze H, Post RM. Three times more days depressed than manic or hypomanic in both bipolar I and bipolar II disorder. Bipolar Disord 2007; 9:531-535 152  100.   Kemp DE, Gao K, Chan PK, Ganocy SJ, Findling RL, Calabrese JR. Medical comorbidity in bipolar disorder: Relationship between illnesses of the endocrine/metabolic system and treatment outcome. Bipolar Disord 2010; 12:404-413 101.   McIntyre RS, Mandel FS, Pappadopulos E. Baseline metabolic status is a moderator of outcome in bipolar disorder patients: Analysis of pooled data from ziprasidone monotherapy clinical trials. Presented at the 164th Annual Meeting of the American Psychiatric Association, Honolulu, HI, may 14-18, 2011 2011;  102.   Yim CY, Soczynska JK, Kennedy SH, Woldeyohannes HO, Brietzke E, McIntyre RS. The effect of overweight/obesity on cognitive function in euthymic individuals with bipolar disorder. Eur Psychiatry 2012; 27:223-228 103.   Keck PE Jr, McElroy SL, Strakowski SM, West SA, Sax KW, Hawkins JM, Bourne ML, Haggard P: 12-month outcome of patients with bipolar disorder following hospitalization for a manic or mixed episode. Am J Psychiatry 1998; 155:646-652 104.   Strakowski SM, Keck PE Jr, McElroy SL, West SA, Sax KW, Hawkins JM, Kmetz GF, Upadhyaya VH, Tugrul KC, Bourne ML. Twelve-month outcome after a first hospitalization for affective psychosis. Arch Gen Psychiatry 1998; 55:49-55 105.   Conus P, Cotton S, Abdel-Baki A, Lambert M, Berk M, McGorry PD. Symptomatic and functional outcome 12 months after a first episode of psychotic mania: Barriers to recovery in a catchment area sample. Bipolar Disord 2006; 8:221-231 106.   DelBello MP, Hanseman D, Adler CM, Fleck DE, Strakowski SM. Twelve-month outcome of adolescents with bipolar disorder following first hospitalization for a manic or mixed episode. Am J Psychiatry 2007; 164:582-590 153  107.   Rodbard HW, Fox KM, Grandy S. Impact of obesity on work productivity and role disability in individuals with and at risk for diabetes mellitus. Am J Health Promot 2009; 23:353-360 108.   Samsa GP, Kolotkin RL, Williams GR, Nguyen MH, Mendel CM. Effect of moderate weight loss on health-related quality of life: An analysis of combined data from 4 randomized trials of sibutramine vs placebo. Am J Manag Care 2001; 7:875-883 109.   Calkin C, van De Velde C, Ruzickova M, Slaney C, Garnham J, Hajek T, O'Donovan C, Alda M. Can body mass index help predict outcome in patients with bipolar disorder? Bipolar Disord 2009; 11:650-656 110.   Kolotkin RL, Corey-Lisle PK, Crosby RD, Swanson JM, Tuomari AV, L'italien GJ, Mitchell JE. Impact of obesity on health-related quality of life in schizophrenia and bipolar disorder. Obesity 2008; 16:749-754 111.   McIntyre RS. Understanding needs, interactions, treatment, and expectations among individuals affected by bipolar disorder or schizophrenia: The UNITE global survey. J Clin Psychiatry 2009; 70 Suppl 3:5-11 112.   Yatham LN, Kauer-Sant'anna M, Bond DJ, Lam RW, Tam E. Course and outcome after the first manic episode in patients with bipolar disorder: Prospective 12-month data from the Systematic Treatment Optimization Program for Early Mania (STOP-EM) project. Can J Psychiatry 2009; 54:105-112 113.   Yatham LN, Kennedy SH, Parikh SV, Schaffer A, Beaulieu S, Alda M, O'Donovan C, Macqueen G, McIntyre RS, Sharma V, Ravindran A, Young LT, Milev R, Bond DJ, Frey BN, Goldstein BI, Lafer B, Birmaher B, Ha K, Nolen WA, Berk M. Canadian Network for Mood and Anxiety Treatments (CANMAT) and International Society for Bipolar Disorders (ISBD) 154  collaborative update of CANMAT guidelines for the management of patients with bipolar disorder: Update 2013. Bipolar Disord 2013; 15:1-44 114.   Shields M, Tjepkema M. Trends in adult obesity. Health Reports (Statistics Canada, Catalogue 82-003) 2006; 17:53-59 115.   Shields M, Tjepkema M. Regional differences in obesity. Health Rep 2006; 17:61-67 116.   National Centre for Social Research Department of Epidemiology and Public Health at the Royal Free and University College Medical School. Health survey for England, 2003. 2004;  117.   Peselow ED, Dunner DL, Fieve RR, Lautin A. Lithium carbonate and weight gain. J Affect Disord 1980; 2:303-310 118.   Bowden CL, Calabrese JR, McElroy SL, Gyulai L, Wassef A, Petty F, Pope HG Jr, Chou JC, Keck PE Jr, Rhodes LJ, Swann AC, Hirschfeld RM, Wozniak PJ. A randomized, placebo-controlled 12-month trial of divalproex and lithium in treatment of outpatients with bipolar I disorder. Arch Gen Psychiatry 2000; 57:481-489 119.   Chengappa KN, Chalasani L, Brar JS, Parepally H, Houck P, Levine J. Changes in body weight and body mass index among psychiatric patients receiving lithium, valproate, or topiramate: An open-label, nonrandomized chart review. Clin Therapeutics 2002; 24:1576-1584 120.   Sachs G, Bowden C, Calabrese JR, Ketter T, Thompson T, White R, Bentley B. Effects of lamotrigine and lithium on body weight during maintenance treatment of bipolar I disorder. Bipolar Disord 2006; 8:175-181 121.   Hennen J, Perlis RH, Sachs G, Tohen M, Baldessarini RJ. Weight gain during treatment of bipolar I patients with olanzapine. J Clin Psychiatry 2004; 65:1679-1687 122.   Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, Hergueta T, Baker R, Dunbar GC. The mini-international neuropsychiatric interview (M.I.N.I.): The 155  development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry 1998; 59 Suppl 20:22-33 123.   Roy-Byrne P, Post RM, Uhde TW, Porcu T, Davis D. The longitudinal course of recurrent affective illness: Life chart data from research patients at the NIMH. Acta Psychiatr Scand 1985; 71:1-34 124.   Young RC, Biggs JT, Ziegler VE, Meyer DA: A rating scale for mania: Reliability, validity and sensitivity. Br J Psychiatry 1978; 133:429-435 125.   WIlliams JBW, Link MJ, Rosenthal NE, Terman M. Structured Interview Guide for the Hamilton Depression Rating Scale, Seasonal Affective Disorders Version (SIGH-SAD). New York, New York State Psychiatric Institute, 1988 126.   Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry 1979; 134:382-389 127.   Overall JE Gorham DR. The brief psychiatric rating scale. Psychol Rep 1962; 10:799-812 128.   Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull 1987; 13:261-276 129.   Spearing MK, Post RM, Leverich GS, Brandt D, Nolen W. Modification of the clinical global impressions (CGI) scale for use in bipolar illness (BP): The CGI-BP. Psychiatry Res 1997; 73:159-171 130.   Jaeger J, Berns SM, Czobor P. The multidimensional scale of independent functioning: A new instrument for measuring functional disability in psychiatric populations. Schizophr Bull 2003; 29:153-168 156  131.   Chengappa KN, Baker RW, Shao L, Yatham LN, Tohen M, Gershon S, Kupfer DJ. Rates of response, euthymia and remission in two placebo-controlled olanzapine trials for bipolar mania. Bipolar Disord 2003; 5:1-5 132.   McIntyre RS, Fallu A, Konarski JZ. Measurable outcomes in psychiatric disorders: Remission as a marker of wellness. Clin Ther 2006; 28:1882-1891 133.   World Health Organization. Obesity: Preventing and managing the global epidemic (publication WHO/NUT/NCD/98.1). 1997;  134.   Torrent C, Amann B, S?nchez-Moreno J, Colom F, Reinares M, Comes M, Rosa AR, Scott J, Vieta E. Weight gain in bipolar disorder: Pharmacological treatment as a contributing factor. Acta Psychiatric Scand 2008; 118:4-18 135.   Finkelstein EA, Fiebelkorn IC, Wang G. National medical spending attributable to overweight and obesity: How much, and who's paying? Health Aff 2003 Suppl Web Exclusives:W3-219-26; Suppl Web Exclusives (W3):219-226 136.   Fagiolini A, Kupfer DJ, Rucci P, Scott JA, Novick DM, Frank E.. Suicide attempts and ideation in patients with bipolar I disorder. J Clin Psychiatry 2004; 65:509-514 137.   Gardner HH, Kleinman NL, Brook RA, Rajagopalan K, Brizee TJ, Smeeding JE. The economic impact of bipolar disorder in an employed population from an employer perspective. J Clin Psychiatry 2006; 67:1209-1218 138.   American Diabetes Association, American Psychiatric Association, American Association of Clinical Endocrinologists, North American Association for the Study of Obesity. Consensus development conference on antipsychotic drugs and obesity and diabetes. J Clin Psychiatry 2004; 65:267-272 157  139.   Kolotkin RL, Crosby RD, Corey-Lisle PK, Li H, Swanson JM. Performance of a weight-related measure of quality of life in a psychiatric sample. Qual Life Res 2006; 15:587-596 140.   Berns S, Uzelac S, Gonzalez C, Jaeger J. Methodological considerations of measuring disability in bipolar disorder: Validity of the multidimensional scale of independent functioning. Bipolar Disord 2007; 9:3-10 141.   Bond DJ, Kauer-Sant?Anna M, Lam RW, Yatham LN. Weight gain, obesity, and metabolic indices following a first manic episode: Prospective 12-month data from the Systematic Treatment Optimization Program for Early Mania (STOP-EM). J Affect Disord 2010; 124:108-117 142.   Kessing LV, Hansen MG, Andersen PK. Course of illness in depressive and bipolar disorders: naturalistic study, 1994-1999. Br J Psychiatry 2004; 185:372-377 143.   Kora K, Saylan M, Akkaya C, Karamustafalioglu N, Tomruk N, Yasan A, Oral T. Predictive factors for time to remission and recurrence in patients treated for acute mania: Health outcomes of manic episodes (HOME) study. Prim Care Companion J Clin Psychiatry 2008; 10:114-119 144.   Kessing LV, Andersen PK, Mortensen PB. Predictors of recurrence in affective disorder: A case register study. J Affect Disord 1998; 49:101-108 145.   Perlis RH, Ostacher MJ, Patel JK, Marangell LB, Zhang H, Wisniewski SR, Ketter TA, Miklowitz DJ, Otto MW, Gyulai L, Reilly-Harrington NA, Nierenberg AA, Sachs GS, Thase ME. Predictors of recurrence in bipolar disorder: Primary outcomes from the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). Am J Psychiatry 2006; 163:217-224 158  146.   Tohen M, Zarate CA Jr, Hennen J, Khalsa HM, Strakowski SM, Gebre-Medhin P, Salvatore P, Baldessarini RJ. The McLean-Harvard first-episode mania study: Prediction of recovery and first recurrence. Am J Psychiatry 2003; 160:2099-2107 147.   Altshuler LL, Post RM, Black DO, Keck PE Jr, Nolen WA, Frye MA, Suppes T, Grunze H, Kupka RW, Leverich GS, McElroy SL, Walden J, Mintz J. Subsyndromal depressive symptoms are associated with functional impairment in patients with bipolar disorder: Results of a large, multisite study. J Clin Psychiatry 2006; 67:1551-1560 148.   Tjepkema M. Adult obesity. Health Reports (Statistics Canada, Catalogue 82-003) 2006; 17:9-25 149.   Bond DJ, Kunz M, Torres IJ, Lam RW, Yatham LN. The association of weight gain with mood symptoms and functional outcomes following a first manic episode: Prospective 12-month data from the Systematic Treatment Optimization Program for Early Mania (STOP-EM). Bipolar Disord 2010; 12:616-626 150.   Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H, Bannister PR, De Luca M, Drobnjak I, Flitney DE, Niazy RK, Saunders J, Vickers J, Zhang Y, De Stefano N, Brady JM, Matthews PM. Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 2004; 23 (Suppl 1):S208-S219 151.   Woolrich MW, Jbabdi S, Patenaude B, Chappell M, Makni S, Behrens T, Beckmann C, Jenkinson M, Smith SM. Bayesian analysis of neuroimaging data in FSL. Neuroimage 2009; 45(1 Suppl):S173-S186 152.   Jenkinson M, Smith S. A global optimization method for robust affine registration of brain images. Med Image Anal 2001; 5:143-156 159  153.   Mathalon DH, Sullivan EV, Rawles JM, Pfefferbaum A. Correction for head size in brain-imaging measurements. Psychiatry Res 1993; 50:121-139 154.   Whitwell JL, Crum WR, Watt HC, Fox NC. Normalization of cerebral volumes by use of intracranial volume: Implications for longitudinal quantitative MR imaging. Ajnr 2001; 22:1483-1489 155.   Giedd JN, Blumenthal J, Jeffries NO, Castellanos FX, Liu H, Zijdenbos A, Paus T, Evans AC, Rapoport JL. Brain development during childhood and adolescence: A longitudinal MRI study. Nat Neurosci 1999; 2:861-863 156.   Paul CA, Au R, Fredman L, Massaro JM, Seshadri S, Decarli C, Wolf PA. Association of alcohol consumption with brain volume in the Framingham study. Arch Neurol 2008; 65:1363-1367 157.   Quickfall J Crockford D. Brain neuroimaging in cannabis use: A review. J Neuropsychiatry Clin Neurosci 2006; 18:318-332 158.   Videbech P, Ravnkilde B. Hippocampal volume and depression: A meta-analysis of MRI studies. Am J Psychiatry 2004; 161:1957-1966 159.   Nery FG, Chen HH, Hatch JP, Nicoletti MA, Brambilla P, Sassi RB, Mallinger AG, Keshavan MS, Soares JC. Orbitofrontal cortex gray matter volumes in bipolar disorder patients: A region-of-interest MRI study. Bipolar Disord 2009; 11:145-153 160.   Moore GJ, Bebchuk JM, Wilds IB, Chen G, Manji HK. Lithium-induced increase in human brain grey matter. Lancet 2000; 356:1241-1242 161.   Atmaca M, Ozdemir H, Cetinkaya S, Parmaksiz S, Belli H, Poyraz AK, Tezcan E, Ogur E. Cingulate gyrus volumetry in drug free bipolar patients and patients treated with valproate or valproate and quetiapine. J Psychiatr Res 2007; 41:821-827 160  162.   Ward MA, Bendlin BB, McLaren DG, Hess TM, Gallagher CL, Kastman EK, Rowley HA, Asthana S, Carlsson CM, Sager MA, Johnson SC. Low HDL cholesterol is associated with lower gray matter volume in cognitively healthy adults. Front Aging Neurosci 2010; 2:article 29 163.   Vita A, De Peri L, Sacchetti E. Gray matter, white matter, brain, and intracranial volumes in first-episode bipolar disorder: A meta-analysis of magnetic resonance imaging studies. Bipolar Disord 2009; 11:807-814 164.   Konarski JZ, McIntyre RS, Kennedy SH, Rafi-Tari S, Soczynska JK, Ketter TA. Volumetric neuroimaging investigations in mood disorders: Bipolar disorder versus major depressive disorder. Bipolar Disord 2008; 10:1-37 165.   Doris A, Belton E, Ebmeier KP, Glabus MF, Marshall I. Reduction of cingulate gray matter density in poor outcome bipolar illness. Psychiatry Res 2004; 130:153-159 166.   MacQueen GM. Magnetic resonance imaging and prediction of outcome in patients with major depressive disorder. J Psychiatry Neurosci 2009; 34:343-349 167.   Puri BK. Progressive structural brain changes in schizophrenia. Expert Rev Neurother 2010; 10:33-42 168.   Thompson PM, Giedd JN, Woods RP, MacDonald D, Evans AC, Toga AW. Growth patterns in the developing brain detected by using continuum mechanical tensor maps. Nature 2000; 404:190-193 169.   Raji CA, Ho AJ, Parikshak NN, Becker JT, Lopez OL, Kuller LH, Hua X, Leow AD, Toga AW, Thompson PM. Brain structure and obesity. Hum Brain Map 2009; doi:10.1002/hbm.20870 170.   Debette S, Beiser A, Hoffmann U, Decarli C, O'Donnell CJ, Massaro JM, Au R, Himali JJ, Wolf PA, Fox CS, Seshadri S. Visceral fat is associated with lower brain volume in healthy middle-aged adults. Ann Neurol 2010; 68:136-144 161  171.   Ho AJ, Stein JL, Hua X, Lee S, Hibar DP, Leow AD, Dinov ID, Toga AW, Saykin AJ, Shen L, Foroud T, Pankratz N, Huentelman MJ, Craig DW, Gerber JD, Allen AN, Corneveaux JJ, Stephan DA, DeCarli CS, DeChairo BM, Potkin SG, Jack CR Jr, Weiner MW, Raji CA, Lopez OL, Becker JT, Carmichael OT, Thompson PM. A commonly carried allele of the obesity-related FTO gene is associated with reduced brain volume in the healthy elderly. Proc Natl Acad Sci USA 2010; 107:8404-8409 172.   Kempton MJ, Geddes JR, Ettinger U, Williams SC, Grasby PM. Meta-analysis, database, and meta-regression of 98 structural imaging studies in bipolar disorder. Arch Gen Psychiatry 2008; 65:1017-1032 173.   Lyoo IK, Dager SR, Kim JE, Yoon SJ, Friedman SD, Dunner DL, Renshaw PF. Lithium-induced gray matter volume increase as a neural correlate of treatment response in bipolar disorder: A longitudinal brain imaging study. Neuropsychopharmacology 2010; doi:10.1038/npp.2010.41 174.   Lieberman JA, Tollefson GD, Charles C, Zipursky R, Sharma T, Kahn RS, Keefe RS, Green AI, Gur RE, McEvoy J, Perkins D, Hamer RM, Gu H, Tohen M. Antipsychotic drug effects on brain morphology in first-episode psychosis. Arch Gen Psychiatry 2005; 62:361-370 175.   Strakowski SM, Adler CM, Almeida J, Altshuler LL, Blumberg HP, Chang KD, Delbello MP, Frangou S, McIntosh A, Phillips ML, Sussman JE, Townsend JD. The functional neuroanatomy of bipolar disorder: A consensus model. Bipolar Disord 2012; 14:313-325 176.   Selvaraj S, Arnone D, Job D, Stanfield A, Farrow TF, Nugent AC, Scherk H, Gruber O, Chen X, Sachdev PS, Dickstein DP, Malhi GS, Ha TH, Ha K, Phillips ML, McIntosh AM. Grey matter differences in bipolar disorder: A meta-analysis of voxel-based morphometry studies. Bipolar Disord 2012; 14:135-145 162  177.   Vederine FE, Wessa M, Leboyer M, Houenou J. A meta-analysis of whole-brain diffusion tensor imaging studies in bipolar disorder. Prog Neuropsychopharmacol Biol Psychiatry 2011; 35:1820-1826 178.   Ashburner J. A fast diffeomorphic image registration algorithm. Neuroimage 2007; 38:95-113 179.   Klein A, Andersson J, Ardekani BA, Ashburner J, Avants B, Chiang MC, Christensen GE, Collins DL, Gee J, Hellier P, Song JH, Jenkinson M, Lepage C, Rueckert D, Thompson P, Vercauteren T, Woods RP, Mann JJ, Parsey RV. Evaluation of 14 nonlinear deformation algorithms applied to human brain MRI registration. Neuroimage 2009; 46:786-802 180.   Yassa MA, Stark CE. A quantitative evaluation of cross-participant registration techniques for MRI studies of the medial temporal lobe. Neuroimage 2009; 44:319-327 181.   Hayasaka S, Phan KL, Liberzon I, Worsley KJ, Nichols TE. Nonstationary cluster-size inference with random field and permutation methods. Neuroimage 2004; 22:676-687 182.   Maldjian JA, Laurienti PJ, Kraft RA, Burdette JH. An automated method for neuroanatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets. Neuroimage 2003; 19:1233-1239 183.   Hua K, Zhang J, Wakana S, Jiang H, Li X, Reich DS, Calabresi PA, Pekar JJ, van Zijl PC, Mori S. Tract probability maps in stereotaxic spaces: Analyses of white matter anatomy and tract-specific quantification. Neuroimage 2008; 39:336-347 184.   Mori S, Oishi K, Jiang H, Jiang L, Li X, Akhter K, Hua K, Faria AV, Mahmood A, Woods R, Toga AW, Pike GB, Neto PR, Evans A, Zhang J, Huang H, Miller MI, van Zijl P, Mazziotta J. Stereotaxic white matter atlas based on diffusion tensor imaging in an ICBM template. Neuroimage 2008; 40:570-582 163  185.   Sabatinelli D, Fortune EE, Li Q, Siddiqui A, Krafft C, Oliver WT, Beck S, Jeffries J. Emotional perception: Meta-analyses of face and natural scene processing. Neuroimage 2011; 54:2524-2533 186.   Chaddock CA, Barker GJ, Marshall N, Schulze K, Hall MH, Fern A, Walshe M, Bramon E, Chitnis XA, Murray R, McDonald C. White matter microstructural impairments and genetic liability to familial bipolar I disorder. Br J Psychiatry 2009; 194:527-534 187.   Baslow MH. N-acetylaspartate in the vertebrate brain: Metabolism and function. Neurochem Res 2003; 28:941-953 188.   Gonzales MM, Tarumi T, Eagan DE, Tanaka H, Vaghasia M, Haley AP. Indirect effects of elevated body mass index on memory performance through altered cerebral metabolite concentrations. Psychosom Med 2012; 74:691-698 189.   Gazdzinski S, Kornak J, Weiner MW, Meyerhoff DJ. Body mass index and magnetic resonance markers of brain integrity in adults. Ann Neurol 2008; 63:652-657 190.   Gazdzinski S, Durazzo TC, Mon A, Meyerhoff DJ. Body mass index is associated with brain metabolite levels in alcohol dependence--a multimodal magnetic resonance study. Alcohol Clin Exp Res 2010; 34:2089-2096 191.   Atmaca M, Yildirim H, Ozdemir H, Poyraz AK, Tezcan E, Ogur E.. Hippocampal 1H MRS in first-episode bipolar I patients. Prog Neuropsychopharmacol Biol Psychiatry 2006; 30:1235-1239 192.   Reynolds LM Reynolds GP. Differential regional N-acetylaspartate deficits in postmortem brain in schizophrenia, bipolar disorder and major depressive disorder. J Psychiatr Res 2011; 45:54-59 164  193.   Senaratne R, Milne AM, MacQueen GM, Hall GB. Increased choline-containing compounds in the orbitofrontal cortex and hippocampus in euthymic patients with bipolar disorder: A proton magnetic resonance spectroscopy study. Psychiatry Res 2009; 172:205-209 194.   Iosifescu DV, Moore CM, Deckersbach T, Tilley CA, Ostacher MJ, Sachs GS, Nierenberg AA. Galantamine-ER for cognitive dysfunction in bipolar disorder and correlation with hippocampal neuronal viability: A proof-of-concept study. CNS Neurosci Ther 2009; 15:309-319 195.   Frey BN, Andreazza AC, Nery FG, Martins MR, Quevedo J, Soares JC, Kapczinski F. The role of hippocampus in the pathophysiology of bipolar disorder. Behav Pharmacol 2007; 18(5-6):419-430 196.   Provencher SW. Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magn Reson Med 1993; 30:672-679 197.   Milne A, MacQueen GM, Yucel K, Soreni N, Hall GB. Hippocampal metabolic abnormalities at first onset and with recurrent episodes of a major depressive disorder: A proton magnetic resonance spectroscopy study. Neuroimage 2009; 47:36-41 198.   Block W, Tr?ber F, von Widdern O, Metten M, Schild H, Maier W, Zobel A, Jessen F. Proton MR spectroscopy of the hippocampus at 3 T in patients with unipolar major depressive disorder: Correlates and predictors of treatment response. Int J Neuropsychopharmacol 2009; 12:415-422 199.   Portella MJ, de Diego-Adeli?o J, G?mez-Ans?n B, Morgan-Ferrando R, Vives Y, Puigdemont D, P?rez-Egea R, Ruscalleda J, Enric ?lvarez, P?rez V. Ventromedial prefrontal spectroscopic abnormalities over the course of depression: A comparison among first episode, remitted recurrent and chronic patients. J Psychiatr Res 2011; 45:427-434 165  200.   Lambert E, Sari CI, Dawood T, Nguyen J, McGrane M, Eikelis N, Chopra R, Wong C, Chatzivlastou K, Head G, Straznicky N, Esler M, Schlaich M, Lambert G. Sympathetic nervous system activity is associated with obesity-induced subclinical organ damage in young adults. Hypertension 2010; 56:351-358 201.   Viazzi F, Leoncini G, Adami GF, Papadia FS, Bezante GP, Conti N, Baratto E, Scopinaro N, Deferrari G, Pontremoli R. Sub-clinical organ damage in hypertension and obesity. Nutr Metab Cardiovasc Dis 2011; 21:597-602 202.   Whitmer RA, Gunderson EP, Quesenberry CP Jr, Zhou J, Yaffe K. Body mass index in midlife and risk of alzheimer disease and vascular dementia. Curr Alzheimer Res 2007; 4:103-109 203.   Rosengren A, Skoog I, Gustafson D, Wilhelmsen L. Body mass index, other cardiovascular risk factors, and hospitalization for dementia. Arch Intern Med 2005; 165:321-326 204.   Gortmaker SL, Must A, Perrin JM, Sobol AM, Dietz WH. Social and economic consequences of overweight in adolescence and young adulthood. Nejm 1993; 329:1008-1012 205.   Pierce JW, Wardle J. Self-esteem, parental appraisal and body size in children. J Child Psychol Psychiatry 1993; 34:1125-1136 206.   Roehling MV. Weight-based discrimination in employment: Psychological and legal aspects. Pers Psychol 1999; 52:969-1017 207.   Kristeller JL, Hoerr RA. Physician attitudes toward managing obesity: Differences among six specialty groups. Prev Med 1997; 26:542-549 208.   Tang-P?ronard JL, Heitmann BL. Stigmatization of obese children and adolescents, the importance of gender. Obes Rev 2008 9 522-34; 9:522-534 166  209.   Brain Development Cooperative Group. Total and regional brain volumes in a population-based normative sample from 4 to 18 years: The NIH MRI study of normal brain development. Cereb Cortex 2012; 22:1-12 210.   Kalmar JH, Wang F, Spencer L, Edmiston E, Lacadie CM, Martin A, Constable RT, Duncan JS, Staib LH, Papademetris X, Blumberg HP. Preliminary evidence for progressive prefrontal abnormalities in adolescents and young adults with bipolar disorder. J Int Neuropsychol Soc 2009; 15:476-481 211.   Lisy ME, Jarvis KB, DelBello MP, Mills NP, Weber WA, Fleck D, Strakowski SM, Adler CM. Progressive neurostructural changes in adolescent and adult patients with bipolar disorder. Bipolar Disord 2011; 13:396-405 212.   Schneider MR, Delbello MP, McNamara RK, Strakowski SM, Adler CM. Neuroprogression in bipolar disorder. Bipolar Disord 2012; 14:356-374 213.   Hajek T, Kopecek M, Hoschl C, Alda M. Smaller hippocampal volumes in patients with bipolar disorder are masked by exposure to lithium: A meta-analysis. J Psychiatry Neurosci 2012; 37:333-343 214.   Cousins DA, Butts K, Young AH. The role of dopamine in bipolar disorder. Bipolar Disord 2009; 11:787-806 215.   Goldstein BI, Kemp DE, Soczynska JK, McIntyre RS. Inflammation and the phenomenology, pathophysiology, comorbidity, and treatment of bipolar disorder: A systematic review of the literature. J Clin Psychiatry 2009; 70:1078-1090 216.   Berk M, Kapczinski F, Andreazza AC, Dean OM, Giorlando F, Maes M, Y?cel M, Gama CS, Dodd S, Dean B, Magalh?es PV, Amminger P, McGorry P, Malhi GS. Pathways underlying 167  neuroprogression in bipolar disorder: Focus on inflammation, oxidative stress and neurotrophic factors. Neurosci Biobehav Rev 2011; 35:804-817 217.   Fung A, Vizcaychipi M, Lloyd D, Wan Y, Ma D. Central nervous system inflammation in disease related conditions: Mechanistic prospects. Brain Res 2012; 1446:144-155 218.   Dantzer R, O'Connor JC, Freund GG, Johnson RW, Kelley KW. From inflammation to sickness and depression: When the immune system subjugates the brain. Nat Rev Neurosci 2008; 9:46-56 219.   Capuron L, Miller AH. Immune system to brain signaling: Neuropsychopharmacological implications. Pharmacol Ther 2011; 130:226-238 220.   Simka M. Blood brain barrier compromise with endothelial inflammation may lead to autoimmune loss of myelin during multiple sclerosis. Curr Neurovasc Res 2009; 6:132-139 221.   Drake C, Boutin H, Jones MS, Denes A, McColl BW, Selvarajah JR, Hulme S, Georgiou RF, Hinz R, Gerhard A, Vail A, Prenant C, Julyan P, Maroy R, Brown G, Smigova A, Herholz K, Kassiou M, Crossman D, Francis S, Proctor SD, Russell JC, Hopkins SJ, Tyrrell PJ, Rothwell NJ, Allan SM. Brain inflammation is induced by co-morbidities and risk factors for stroke. Brain Behav Immun 2011; 25:1113-1122 222.   Schmitz T Chew LJ. Cytokines and myelination in the central nervous system. Scientific World Journal 2008; 2:1119-1147 223.   Lee HJ, Ertley RN, Rapoport SI, Bazinet RP, Rao JS. Chronic administration of lamotrigine downregulates COX-2 mRNA and protein in rat frontal cortex. Neurochem Res 2008; 33:861-866 168  224.   Bosetti F, Weerasinghe GR, Rosenberger TA, Rapoport SI. Valproic acid down-regulates the conversion of arachidonic acid to eicosanoids via cyclooxygenase-1 and -2 in rat brain. J Neurochem 2003; 85:690-696 225.   Ghelardoni S, Tomita YA, Bell JM, Rapoport SI, Bosetti F. Chronic carbamazepine selectively downregulates cytosolic phospholipase A2 expression and cyclooxygenase activity in rat brain. Biol Psychiatry 2004; 56:248-254 226.   Pollmacher T, Haack M, Schuld A, Kraus T, Hinze-Selch D. Effects of antipsychotic drugs on cytokine networks. J Psychiatr Res 2000; 34:369-382 227.   Motivala SJ. Sleep and inflammation: Psychoneuroimmunology in the context of cardiovascular disease. Ann Behav Med 2011; 42:141-152 228.   Rohleder N. Acute and chronic stress induced changes in sensitivity of peripheral inflammatory pathways to the signals of multiple stress systems - 2011 Curt Richter award winner. Psychoneuroendocrinology 2012; 37:307-316 229.   Patel SR. Reduced sleep as an obesity risk factor. Obes Rev 2009; 10 Suppl 2:61-68 230.   Lenard NR, Berthoud HR. Central and peripheral regulation of food intake and physical activity: Pathways and genes. Obesity 2008; 16 Suppl 3:S11-S22 231.   Kenny PJ. Reward mechanisms in obesity: New insights and future directions. Neuron 2011; 69:664-679 232.   Ely AV, Winter S, Lowe MR. The generation and inhibition of hedonically-driven food intake: Behavioral and neurophysiological determinants in healthy weight individuals. Physiol Behav 2013;  233.   Vucetic Z, Reyes TM. Central dopaminergic circuitry controlling food intake and reward: Implications for the regulation of obesity. Wiley Interdiscip Rev Syst Biol Med 2010; 2:577-593 169  234.   Zhang N, Yuan C, Li Z, Li J, Li X, Li C, Li R, Wang SR. Meta-analysis of the relationship between obestatin and ghrelin levels and the ghrelin/obestatin ratio with respect to obesity. Am J Med Sci 2011; 341:48-55 235.   Harvey J. Leptin: A diverse regulator of neuronal function. J Neurochem 2007; 100:307-313 236.   Andrews ZB. The extra-hypothalamic actions of ghrelin on neuronal function. Trends Neurosci 2011; 34:31-40 237.   Signore AP, Zhang F, Weng Z, Gao Y, Chen J. Leptin neuroprotection in the CNS: Mechanisms and therapeutic potentials. J Neurochem 2008; 106:1977-1990 238.   Moon M, Kim HG, Hwang L, Seo JH, Kim S, Hwang S, Kim S, Lee D, Chung H, Oh MS, Lee KT, Park S. Neuroprotective effect of ghrelin in the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine mouse model of Parkinson's disease by blocking microglial activation. Neurotox Res 2009; 15:332-347 239.   Harvey J, Solovyova N, Irving A. Leptin and its role in hippocampal synaptic plasticity. Prog Lipid Res 2006; 45:369-378 240.   Chen L, Xing T, Wang M, Miao Y, Tang M, Chen J, Li G, Ruan DY. Local infusion of ghrelin enhanced hippocampal synaptic plasticity and spatial memory through activation of phosphoinositide 3-kinase in the dentate gyrus of adult rats. Eur J Neurosci 2011; 33:266-275 241.   Opland DM, Leinninger GM, Myers MG Jr. Modulation of the mesolimbic dopamine system by leptin. Brain Res 2010; 1350:65-70 242.   Andrews ZB, Erion D, Beiler R, Liu ZW, Abizaid A, Zigman J, Elsworth JD, Savitt JM, DiMarchi R, Tschoep M, Roth RH, Gao XB, Horvath TL. Ghrelin promotes and protects 170  nigrostriatal dopamine function via a UCP2-dependent mitochondrial mechanism. J Neurosci 2009; 29:14057-14065 243.   Porter RK, Andrews JF. Effects of leptin on mitochondrial 'proton leak' and uncoupling proteins: Implications for mammalian energy metabolism. Proc Nutr Soc 1998; 57:455-460 244.   Barazzoni R, Bosutti A, Stebel M, Cattin MR, Roder E, Visintin L, Cattin L, Biolo G, Zanetti M, Guarnieri G. Ghrelin regulates mitochondrial-lipid metabolism gene expression and tissue fat distribution in liver and skeletal muscle. Am J Physiol Endocrinol Metab 2005; 288:E228-235 245.   Hertz L, Dringen R, Schousboe A, Robinson SR. Astrocytes: Glutamate producers for neurons. J Neurosci Res 1999; 57:417-428 246.   Arun P, Moffett JR, Namboodiri AM. Evidence for mitochondrial and cytoplasmic N-acetylaspartate synthesis in SH-SY5Y neuroblastoma cells. Neurochem Int 2009; 55:219-225 247.   Konradi C, Eaton M, MacDonald ML, Walsh J, Benes FM, Heckers S. Molecular evidence for mitochondrial dysfunction in bipolar disorder. Arch Gen Psychiatry 2004; 61:300-308 248.   Atmaca M, Kuloglu M, Tezcan E, Ustundag B, Bayik Y. Serum leptin and cholesterol levels in patients with bipolar disorder. Neuropsychobiology 2002; 46:176-179 249.   Kurt E, Guler O, Serteser M, Cansel N, Ozbulut O, Altinba? K, Alata? G, Sava? H, Gecici O. The effects of electroconvulsive therapy on ghrelin, leptin and cholesterol levels in patients with mood disorders. Neurosci Lett 2007; 426:53 250.   Andreazza AC, Kauer-Sant'anna M, Frey BN, Bond DJ, Kapczinski F, Young LT, Yatham LN. Oxidative stress markers in bipolar disorder: A meta-analysis. J Affect Disord 2008; 111:135-144 171  251.   Fernandes BS, Gama CS, Cereser KM, Yatham LN, Fries GR, Colpo G, de Lucena D, Kunz M, Gomes FA, Kapczinski F. Brain-derived neurotrophic factor as a state-marker of mood episodes in bipolar disorders: A systematic review and meta-regression analysis. J Psychiatr Res 2011; 45:995-1004 252.   Weinstock-Guttman B, Zivadinov R, Tama?o-Blanco M, Abdelrahman N, Badgett D, Durfee J, Hussein S, Feichter J, Patrick K, Benedict R, Ramanathan M. Immune cell BDNF secretion is associated with white matter volume in multiple sclerosis. J Neuroimmunol 2007; 188:167-174 253.   Weber NS, Cowan DN, Millikan AM, Niebuhr DW. Psychiatric and general medical conditions comorbid with schizophrenia in the national hospital discharge survey. Psychiatr Serv 2009; 60:1059-1067 254.   Araya AV, Orellana X, Espinoza J. Evaluation of the effect of caloric restriction on serum BDNF in overweight and obese subjects: Preliminary evidences. Endocrine 2008; 33:300-304 255.   Salehi M, Ferenczi A, Zumoff B. Obesity and cortisol status. Horm Metab Res 2005; 37:193-197 256.   Haltia LT, Viljanen A, Parkkola R, Kemppainen N, Rinne JO, Nuutila P, Kaasinen V. Brain white matter expansion in human obesity and the recovering effect of dieting. J Clin Endocrinol Metab 2007; 92:3278-3284 257.   Faulkner G, Cohn T, Remington G. Interventions to reduce weight gain in schizophrenia. Cochrane Database Syst Rev 2007; (1):CD005148 258.   Maayan L, Vakhrusheva J, Correll CU. Effectiveness of medications used to attenuate antipsychotic-related weight gain and metabolic abnormalities: A systematic review and meta-analysis. Neuropsychopharmacology 2010; 35:1520-1530 172  259.   Sylvia LG, Nierenberg AA, Stange JP, Peckham AD, Deckersbach T. Development of an integrated psychosocial treatment to address the medical burden associated with bipolar disorder. J Psychiatr Pract 2011; 17:224-232 260.   Wu RR, Zhao JP, Jin H, Shao P, Fang MS, Guo XF, He YQ, Liu YJ, Chen JD, Li LH. Lifestyle intervention and metformin for treatment of antipsychotic-induced weight gain: A randomized controlled trial. JAMA 2008; 299:185-193  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0103356/manifest

Comment

Related Items