UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Using linked health data to explore the epidemiology and impact of mental health and health behaviours… McKay, Kyla Anne 2017

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

Item Metadata

Download

Media
24-ubc_2017_september_mckay_kyla.pdf [ 2.32MB ]
Metadata
JSON: 24-1.0348724.json
JSON-LD: 24-1.0348724-ld.json
RDF/XML (Pretty): 24-1.0348724-rdf.xml
RDF/JSON: 24-1.0348724-rdf.json
Turtle: 24-1.0348724-turtle.txt
N-Triples: 24-1.0348724-rdf-ntriples.txt
Original Record: 24-1.0348724-source.json
Full Text
24-1.0348724-fulltext.txt
Citation
24-1.0348724.ris

Full Text

                                                                                                               USING LINKED HEALTH DATA TO EXPLORE THE EPIDEMIOLOGY AND IMPACT OF MENTAL HEALTH AND HEALTH BEHAVIOURS IN MULTIPLE SCLEROSIS by  Kyla Anne McKay  B.Sc., Dalhousie University, 2008  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Experimental Medicine)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  July 2017  © Kyla Anne McKay, 2017 ii  Abstract Few population-based, methodologically rigorous studies have evaluated the association between mental health and health behaviours in MS. The goal of this dissertation is to contribute to the broader understanding of these relationships and their potential impact on MS.   This dissertation was based on two main cohorts: 1) a multi-site clinic-based longitudinal cohort from across Canada; 2) a population-based health administrative and clinical cohort in British Columbia. A large (n=949) sample of MS patients were recruited from four Canadian MS clinics. Participants completed a series of questionnaires at three visits over two years. The prevalence of psychiatric comorbidities (depression [35%] and anxiety [54%]), and adverse health behaviours (smoking [24%] and non-adherence to disease-modifying therapies (DMTs) [22%]) was high. Alcohol dependence was associated with increased odds of anxiety (Odds Ratio (OR):1.84;95% confidence interval (CI):1.32–2.58) and depression (OR:1.53;95%CI:1.05–2.23), as was smoking (anxiety OR:1.29;95% CI:1.02–1.63; depression OR:1.37;95%CI:1.04–1.78). Non-adherence (<80% of expected doses reported as taken in the previous 30 days) was associated with alcohol dependence (OR:2.14;95% CI:1.23–3.75). When compared to adherence rates estimated from prescription dispensation information in pharmacy records in the prior year, self-reported non-adherence was found to be highly specific (0.96), but only moderately sensitive (0.38). Those who self-reported non-adherence were at high risk of non-adherence over the following year. We identified cases of mood or anxiety disorders using a validated algorithm applied to health administrative data. The presence of a mood or anxiety disorder was associated with significantly increased neurologic disability, measured by the Expanded Disability Status Scale (β-iii  coefficient:0.45; p<0.0001) among 1250 incident cases of MS followed for an average of 9 years. Last, we identified incident cases of MS who did not access an MS clinic, using a validated algorithm applied to administrative data. Sex and socioeconomic status distributions were similar, but non-MS clinic users were older (46 vs 41 years, p<0.001), and had a higher comorbidity burden than MS clinic-users (Rate ratio:1.08;95%CI:1.02-1.15).   MS is an unpredictable disease with considerable variability in health outcomes. This dissertation provides evidence that psychiatric conditions and adverse health behaviours are common, and may explain some of this heterogeneity.  iv  Lay Summary  Multiple sclerosis (MS) is a chronic disease that affects the brain and spinal cord. Though the physical symptoms of the disease take precedence in defining its progression, there are also important emotional changes. Using large samples of people with MS from across Canada, we studied the impact of mental health and health behaviours on persons with MS. We found high rates of both mental health disorders and adverse health behaviours (smoking, alcohol dependence, and non-adherence to medication). Smoking and alcohol dependence were associated with depression and anxiety. Alcohol dependence and perceived difficulties in concentration and memory were associated with non-adherence to medication. Depression was associated with a worse MS disease progression over time. There is no cure for MS, and the course of the disease is highly unpredictable. The work from this dissertation provides evidence that mental health and health behaviours may explain some of this variability.   v  Preface This dissertation was conceived, conducted, and written by Kyla McKay, with direction from Dr. Helen Tremlett, Dr. Ruth Ann Marrie, Dr. Sherri Hayden, and Dr. Lorne Kastrukoff. Funding for the operating grant from the Canadian Institutes of Health Research (CIHR CBG 101829) was obtained by Dr. Ruth Ann Marrie (principal investigator) and Dr. Helen Tremlett (UBC co-investigator). Personal funding was provided by two Alistair M. Fraser Master Studentship Awards from the Multiple Sclerosis Society of Canada (2013-2015) and a Frederick Banting and Charles Best Canada Graduate Scholarships Doctoral Award (2016-2019).  All studies were approved by the British Columbia Ministry of Health and were conducted after receipt of ethics approval from the University of British Columbia Clinical Research Ethics Board (Certificate Number H10-00984) and the Vancouver Coastal Health Authority (Certificate Number V10-00984). I completed all data analysis, interpretation and writing of all the published and to-be-published manuscripts in this dissertation. Co-authors assisted with data interpretation, manuscript revision and/or study conceptualization. Copyright permission has been obtained for all publications included in this dissertation. A version of Chapter 2 is published as an original peer-reviewed research article: McKay KA, Tremlett H, Fisk JD, Patten SB, Fiest K, Berrigan L, Marrie RA. Adverse health behaviours are associated with depression and anxiety in multiple sclerosis: A prospective multisite study. Mult Scler 2016; 22(5), 685-693. vi  A version of Chapter 3 is published as an original peer-reviewed research article: McKay KA, Tremlett H, Patten SB, Fisk JD, Evans C, Fiest K, Campbell T, Marrie RA. Determinants of non-adherence to disease-modifying therapies in multiple sclerosis: A cross-Canada prospective study. Mult Scler 2017; 23(4), 588-596. A version of Chapter 4 has been accepted for publication: McKay KA, Evans C, Fisk JD, Patten SB, Fiest K., Marrie RA, Tremlett H. Disease-modifying therapies and adherence in multiple sclerosis: comparing patient self-report with pharmacy records. Neuroepidemiology 2017. A version of Chapter 5 is in preparation for peer-review publication: McKay KA, Tremlett H, Fisk JD, Zhang T, Kastrukoff  L, Marrie, RA. Association between psychiatric comorbidity and disability progression of multiple sclerosis (to be submitted). A version of Chapter 6 is published as an original peer-reviewed research article: McKay, KA, Tremlett H, Zhu F, Kastrukoff  L, Marrie RA, Kingwell E. A population‐based study comparing multiple sclerosis clinic users and non‐users in British Columbia, Canada. Eur J Neurol 2016; 23(6), 1093-1100. Other work which laid the foundation for a broader understanding of MS:  McKay KA, Jahanfar S, Duggan T, Tkachuk S, Tremlett H. Factors associated with onset, relapses or progression in multiple sclerosis: a systematic review. Neurotoxicology 2016; In Press. vii  McKay KA, Kwan V, Duggan T, Tremlett H. Risk factors associated with the onset of relapsing-remitting and primary progressive multiple sclerosis: a systematic review. BioMed Research International 2015; 2015. http://dx.doi.org/10.1155/2015/817238 McKay KA, Tremlett H. The systematic search for risk factors in multiple sclerosis. The Lancet Neurology 2015;14(3):237-238.  viii  Table of Contents Abstract .......................................................................................................................................... ii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ....................................................................................................................... viii List of Tables .............................................................................................................................. xiii List of Figures ............................................................................................................................. xvi List of Abbreviations ................................................................................................................ xvii Acknowledgements .................................................................................................................. xviii Dedication .....................................................................................................................................xx Chapter 1: Introduction ................................................................................................................1 1.1 Multiple Sclerosis ........................................................................................................... 1 1.1.1 Introduction ................................................................................................................. 1 1.1.2 Aetiology and disease activity .................................................................................... 1 1.1.3 Diagnosis and disease course ...................................................................................... 2 1.1.4 Disability progression ................................................................................................. 3 1.2 Mental health and health behaviours .............................................................................. 4 1.2.1 Introduction ................................................................................................................. 4 1.2.2 Mental health and health behaviours in MS ............................................................... 4 1.3 Adherence to medication ................................................................................................ 6 1.3.1 Introduction ................................................................................................................. 6 1.3.2 Adherence to MS disease-modifying therapies .......................................................... 7 1.4 Rationale ......................................................................................................................... 7 ix  1.5 Specific objectives, rationales, and hypotheses .............................................................. 8 1.5.1 Chapter Two: Adverse health behaviours are associated with depression and anxiety in MS………. .......................................................................................................................... 8 1.5.2 Chapter Three: Determinants of non-adherence to disease-modifying therapies in MS……………. ...................................................................................................................... 9 1.5.3 Chapter Four: Disease-modifying therapies and adherence in MS: comparing patient self-report with pharmacy records .......................................................................................... 9 1.5.4 Chapter Five: Association between psychiatric comorbidity and disability progression of MS ................................................................................................................. 11 1.5.5 Chapter Six: A population-based study comparing multiple sclerosis clinic users and non-users in British Columbia, Canada ................................................................................ 11 1.6 Data sources and cohorts .............................................................................................. 12 1.6.1 British Columbia MS Database ................................................................................ 12 1.6.2 Cross-Canada longitudinal cohort............................................................................. 13 1.6.3 Provincial health administrative data ........................................................................ 14 1.7 Approach to analysis ..................................................................................................... 16 Chapter 2: Adverse health behaviours are associated with depression and anxiety in multiple sclerosis ..........................................................................................................................19 2.1 Background and objective ............................................................................................. 19 2.2 Methods......................................................................................................................... 19 2.2.1 Study population ....................................................................................................... 20 2.2.2 Self-reported health behaviours and mental health ................................................... 20 2.2.3 Statistical analysis ..................................................................................................... 21 x  2.3 Results ........................................................................................................................... 22 2.4 Discussion ..................................................................................................................... 29 Chapter 3: Determinants of non-adherence to disease-modifying therapies in multiple sclerosis .........................................................................................................................................35 3.1 Background and objective ............................................................................................. 35 3.2 Methods......................................................................................................................... 35 3.2.1 Study population ....................................................................................................... 36 3.2.2 Self-reported information.......................................................................................... 36 3.2.3 Quantifying adherence .............................................................................................. 37 3.2.4 Statistical analysis ..................................................................................................... 37 3.3 Results ........................................................................................................................... 38 3.4 Discussion ..................................................................................................................... 48 Chapter 4: Disease-modifying therapies and adherence in multiple sclerosis: comparing patient self-report with pharmacy records ................................................................................53 4.1 Background and objectives ........................................................................................... 53 4.2 Methods......................................................................................................................... 54 4.2.1 Study population ....................................................................................................... 54 4.2.2 Self-reported information.......................................................................................... 54 4.2.3 Health administrative (pharmacy records) information ............................................ 54 4.2.4 Definitions of adherence ........................................................................................... 55 4.2.5 Statistical analysis ..................................................................................................... 56 4.3 Results ........................................................................................................................... 57 4.4 Discussion ..................................................................................................................... 61 xi  Chapter 5: Association between psychiatric comorbidity and disability progression of multiple sclerosis ..........................................................................................................................66 5.1 Background and objective ............................................................................................. 66 5.2 Methods......................................................................................................................... 66 5.2.1 Design and setting ..................................................................................................... 66 5.2.2 Study population ....................................................................................................... 67 5.2.3 Defining psychiatric comorbidities ........................................................................... 67 5.2.4 Measuring disability worsening ................................................................................ 68 5.2.5 Statistical analysis ..................................................................................................... 69 5.3 Results ........................................................................................................................... 70 5.4 Discussion ..................................................................................................................... 79 Chapter 6: A population-based study comparing multiple sclerosis clinic users and non-users in British Columbia, Canada ............................................................................................83 6.1 Background and objective ............................................................................................. 83 6.2 Methods......................................................................................................................... 83 6.2.1 Study design .............................................................................................................. 83 6.2.2 Study population ....................................................................................................... 84 6.2.3 Comparisons between clinic and non-clinic cases .................................................... 84 6.2.4 Statistical analysis ..................................................................................................... 86 6.3 Results ........................................................................................................................... 87 6.4 Discussion ..................................................................................................................... 90 Chapter 7: Conclusions ...............................................................................................................95 7.1 Summary of findings..................................................................................................... 95 xii  7.2 Integration of findings................................................................................................... 97 7.3 Strengths and limitations............................................................................................... 99 7.4 Clinical significance and implications ........................................................................ 103 7.5 Knowledge translation ................................................................................................ 105 7.6 Conclusions and future directions ............................................................................... 106 References ...................................................................................................................................108 Appendices ..................................................................................................................................122 Appendix A Supplementary Figures and Tables .................................................................... 122 Appendix B Informed Consent Form ..................................................................................... 127 Appendix C Self-report questionnaire .................................................................................... 132 Appendix D Briefing Note ...................................................................................................... 136  xiii  List of Tables Table 2.1 Demographic and clinical characteristics of study participants at baseline. ................ 23 Table 2.2 Occurrence of mental health and adverse health behaviours over the study period. ... 24 Table 2.3 Frequencies of baseline clinical and demographic variables associated with the four variables of interest: anxiety, depression, alcohol dependence, and current smoking status. ...... 26 Table 2.4 Association between health behaviours and: i) anxiety; ii) depression across all three visits using GEE analysis with an unstructured correlation matrix. ............................................. 27 Table 2.5 Association between baseline health behaviours and incident mental health at year one and two. ......................................................................................................................................... 29 Table 3.1 Baseline clinical and demographic characteristics. ...................................................... 39 Table 3.2 Disease-modifying therapy (DMT) use at baseline, year one, and year two. .............. 40 Table 3.3 Frequency of non-adherence, defined as medication possession ratio < 80% at baseline, year one, and year two by DMT product. ...................................................................... 41 Table 3.4 Univariate and multivariable longitudinal analyses of clinical and demographic variables and their association with non-adherence (medication possession ratio < 80% in previous 30 days). ......................................................................................................................... 44 Table 3.5 Univariate and multivariable longitudinal analyses of clinical and demographic variables and their association with non-adherence (missed ≥ 1 dose in the previous 30 days; ‘yes’ vs ‘no’) ................................................................................................................................. 47 Table 4.1 Demographic and clinical characteristics of the cohort. .............................................. 58 Table 4.2 Agreement between self-reported use of a DMT compared to pharmacy records in the 30 days prior to a MS clinic visit. ................................................................................................. 58 xiv  Table 4.3 Agreement between self-reported missed doses in 30 days and pharmacy records across four time periods in determining rates of non-adherence. ................................................. 60 Table 4.4 Stratified analyses showing the effect of demographic and clinical characteristics on the level of agreement (κ) between self-reported missed doses in 30 days and pharmacy records in the year prior to clinic visit. ...................................................................................................... 61 Table 5.1 Clinical and demographic characteristics of the study cohort. .................................... 72 Table 5.2 Association between all mood or anxiety disorders combined and neurologic disability, as measured by the EDSS. ........................................................................................... 75 Table 5.3 Association between depression, anxiety, and bipolar disorder and neurologic disability, as measured by the EDSS. ........................................................................................... 76 Table 5.4 Association between change in mood or anxiety disorders and neurologic disability, as measured by the EDSS.................................................................................................................. 77 Table 5.5 Association between mood or anxiety disorders and neurologic disability, as measured by the MSSS. ................................................................................................................................ 78 Table 6.1 Characteristics of the MS clinic and the non-clinic cases. ........................................... 88 Table 6.2 Comparison of health services utilization between MS clinic and non-clinic cases. ... 89 Table 6.3 Comparison of comorbidity between MS clinic and non-clinic cases. ........................ 90 Table A.1 Definitions of Expanded Disability Status Scale scores.18 ........................................ 122 Table A.2 Definitions of adherent and non-adherent based on the medication possession ratio of 80% for the four disease-modifying therapy (DMT) types. ....................................................... 124 Table A.3 Diagnosis (ICD-9 and ICD-10) codes and algorithms used to identify psychiatric comorbidities............................................................................................................................... 125 xv  Table A.4 Diagnosis (ICD-9 and ICD-10) codes and algorithms used to identify comorbidities from administrative data. ............................................................................................................ 126 Table A.5 Diagnosis (ICD-9 and ICD-10) codes used to identify MS and demyelinating diseases of the CNS from administrative data. ......................................................................................... 126  xvi  List of Figures Figure 5.1 Selection of incidence-onset MS patients residing in British Columbia, Canada ...... 71 Figure 7.1 Schematic depiction of the potential relationships between MS, mental health, and health behaviours. ......................................................................................................................... 99 Figure A.1 Schematic of data linkages including health administrative and clinical databases.123    xvii  List of Abbreviations  ATC Anatomic Therapeutic Chemical BC British Columbia BCMS  British Columbia Multiple Sclerosis (database) CAGE Cutting down, Annoyance by criticism, Guilty feeling, Eye-openers CI CNS Confidence Interval Central nervous system CIS DAD Clinically isolated syndrome Discharge Abstract Database D-FIS Daily Fatigue Impact Scale DIN Drug Identification Number DMT Disease-modifying therapy DSM Diagnostic and Statistical Manual of Mental Disorders EDSS Expanded Disability Status Scale GEE Generalized estimating equations HADS Hospital Anxiety and Depression Scale HIV Human immunodeficiency disorder HRQoL Health related quality of life HUI Health Utilities Index ICD International Statistical Classification of Diseases and Related Health Problems IFNβ Interferon beta MS Multiple sclerosis MSP NPV Medical Service Plan Negative predictive value PHN Personal Health Number PPMS PPV Primary progressive multiple sclerosis Positive predictive value RRMS Relapsing-remitting multiple sclerosis SES Socioeconomic status SPMS Secondary progressive multiple sclerosis UBC University of British Columbia xviii  Acknowledgements First, I would like to express my sincere gratitude to my supervisor, Dr. Helen Tremlett, for her unwavering support throughout my graduate studies. I could not have done this without her encouragement and guidance, especially through those languid first years. Her unrelenting enthusiasm for science and discovery has been an inspiration to me and many others. She has set an admirable example as a researcher and leader, and I will forever be grateful to have had the opportunity to work with her over the last four years.   My sincere thanks go to Dr. Ruth Ann Marrie, who provided guidance from afar throughout my doctoral studies, and whose mentorship helped shape this dissertation. She has been a constant source of inspiration. I would also like to thank all of my coauthors, as well as my committee members, Dr. Sherri Hayden and Dr. Lorne Kastrukoff, who have provided valuable support from the beginning.   This dissertation was made possible by the patients who volunteered their time to participate in research, and I am deeply appreciative of their generous contributions. I would also like to acknowledge the work of the research assistants who collected the data, and agencies which provided data access and support.  I would like to acknowledge the Multiple Sclerosis Society of Canada for granting me two annual Alistair M. Fraser Master Studentship Awards, the opportunity to attend three endMS Summer School programs, and the endMS Scholar Program for Researchers IN Training (SPRINT) scholarship; the Canadian Institutes of Health Research for granting me the three-year Frederick Banting and Charles Best Canada Graduate Scholarships Doctoral Research Award; xix  the University of British Columbia for the Four-Year Doctoral Fellowship, Faculty of Medicine Graduate Award, the Faculty of Graduate Studies Travel Award, the Experimental Medicine Program Travel Award, and the Publication Award; the Vancouver Coastal Health Research Institute for the Rising Star Award; and the European Committee for Treatment and Research in Multiple Sclerosis for their Young Investigator Travel Grant. This funding has allowed me to broaden my knowledge and develop the expertise necessary to complete this dissertation.   Special thanks are owed to my parents, siblings, and to my partner, Jeffrey, who knew to never ask about my thesis, and who dutifully pretended to read my published work. Thanks to Tom, as well, who made working in a cupboard marginally more interesting.  Jeff, thank you for your enduring optimism and support, and for all of the adventures over the last four years. Last, to my mom, thank you for giving me the freedom to find my own path – your strength and compassion are a continual source of inspiration. xx  Dedication For all the women who came before me. 1  Chapter 1: Introduction 1.1 Multiple Sclerosis 1.1.1 Introduction Multiple sclerosis (MS) is a disabling disease of the central nervous system, estimated to affect over 2.3 million individuals worldwide (2013).1 It typically presents between the ages of 20 and 40 years, and affects twice as many women as men.1 It is the most common non-traumatic cause of disability in young adults.2 The disease can affect virtually every facet of a person’s health and wellbeing. Physical symptoms can include impairments to balance and vision, weakness, ataxia, and pain; most persons with MS will eventually require the use of a walking aid.3 Though the physical symptoms of the disease take precedence in defining its progression, there are also detrimental cognitive and emotional changes. The relatively young onset age and the chronic nature of MS translates into higher societal costs than either stroke or Alzheimer’s disease.4 1.1.2 Aetiology and disease activity MS is widely believed to be an immune-mediated demyelinating disease of the central nervous system (CNS).5 Pathologically, it is characterized by demyelination, axonal loss, and gliosis.6  Though the white matter of the brain is predominantly affected, there is increasing evidence for grey matter involvement as well.7 The aetiology of MS is not fully understood, but is likely multifactorial, combining both genetic and environmental factors. To lay the foundation for the work in this dissertation, we systematically reviewed the literature on the possible risk factors associated with MS disease onset, relapses and progression.8 Best evidence to date indicates that 2  a combination of genetic predisposition, exposure to Epstein-Barr virus (EBV), cigarette smoking, and reduced sunlight exposure/vitamin D levels are involved.8–10 Disease activity is similarly complex and multifaceted. It can be divided into two processes: clinical relapses and longer-term disability progression. MS relapses are defined as acute worsening of function followed by partial or complete recovery,11 while disability progression refers to the gradual worsening of neurological symptoms over time.11 There are some commonalities between the environmental factors associated with the risk of developing MS and those associated with the subsequent disease activity in people with established MS. Low vitamin D serum levels or low sunlight exposure appear to increase the risk of an MS relapse.8 The presence of infections is also correlated with an increased relapse rate; however, other than EBV, the focus has been on common infections, particularly upper respiratory tract infections.8 In terms of modifiable risk factors for the longer-term progression of disability, only cigarette smoking has been consistently implicated as a risk factor for a more rapid progression.8 1.1.3 Diagnosis and disease course The diagnosis of MS remains primarily clinical, with evidence of neurological signs or symptoms disseminated in time and space.12 The diagnosis can be supported by paraclinical evidence (demyelinating lesions on magnetic resonance imaging or oligoclonal bands in the cerebrospinal fluid for primary progressive MS),.12 The diagnosis of MS remains a challenge due to the absence of any single definitive test and the need to exclude all other possible causes,13 but the delay between onset and diagnosis has shortened over time.14 Approximately 85% of patients present with a relapsing-remitting course (RRMS), which is defined by a dynamic course of  relapses and periods of remission.11 RRMS onset typically occurs in early adulthood, and, within 3  around two decades, approximately 50% of persons will go on to develop secondary progressive MS (SPMS).15  SPMS is defined as a steady clinical deterioration, with or without occasional relapses.11 Primary progressive MS (PPMS) affects only 10–15% of the MS population and is associated with the worst prognosis of all MS subtypes.16 PPMS is also associated with an older onset age, and a different sex ratio compared to the more common RRMS.16 In view of the considerable differences in clinical presentation and prognosis, it is possible that these disease courses have distinctive risk factors. During the development of this thesis, we systematically reviewed the literature and found that few studies that explored risk factors for MS considered disease course, such that our current knowledge is predominated by the more common, relapsing-onset MS, and little is known about risk factors for PPMS.17    1.1.4 Disability progression Progression of MS is highly heterogeneous and occurs gradually over many years; the median time from MS onset to requiring a cane is nearly 30 years.3 Consequently, studying MS progression and changes in disability requires large cohorts and long follow-up times. The most widely used measure of MS disability progression is the Expanded Disability Status Scale (EDSS), which is an ordinal scale ranging from 0 = normal to 10 = death due to MS, marked by 0.5 increments.18  It is an expansion of the Disability Status Scale (DSS), which applied the same range, without the 0.5 increments.19 Please see Table A.1 (Appendix A) for a description of the EDSS scale. The multiple sclerosis severity score20 is a related method of describing and quantifying disability  which applies an algorithm that relates EDSS scores to the distribution of disability in patients with comparable disease durations.20 It is a linear scale from 1-10, with a score of 5 representing the average disability for the specified disease duration, and can be 4  applied to cross-sectional data.20 The transition from RRMS to SPMS is also used as a measure of progression. It is typically diagnosed retrospectively by a clinician based on a history of gradual worsening independent of relapses, following a RRMS disease course.11  1.2 Mental health and health behaviours 1.2.1 Introduction Physical and mental health are intricately linked in their causes, comorbidity, and consequences.21 Mental disorders encompass a wide range of mental health conditions characterized by alterations in mood, thinking, and behaviour. Similar to chronic physical illnesses, they have a complex aetiology, resulting from a combination of genetic, environmental, and psychosocial triggers.21 The presence of a chronic medical condition is often associated with higher rates of mood and anxiety disorders.22 The World Health Organization’s ‘World Health Survey’ suggested that depression co-existing with chronic disease leads to significantly greater disease burden and disability compared to having only depression, or only a chronic disease.23  Adverse health behaviours, including cigarette smoking and alcohol misuse, contribute substantially to disability burden and risk of chronic disease24 and commonly occur concomitantly with mental disorders, 25 and chronic disease.26  1.2.2 Mental health and health behaviours in MS The behavioural and psychological changes associated with MS were noted in its earliest descriptions, but went largely overlooked by the research community in the following century.27 In recent years, there has been a resurgence of interest in mental health in MS and the literature 5  is rapidly expanding.28 The focus has predominantly been on depression, with less research devoted to anxiety, alcohol abuse, bipolar disorder, and psychosis.29 Much research has been dedicated to estimating the incidence and prevalence of psychiatric comorbidities in MS, which introduces the challenge of identifying a psychiatric condition at a population-level and in the context of a neurological disorder. There are four main approaches that have been utilized by researchers who study mental health in MS: self-reported psychometric tests (e.g., the Beck Depression Inventory,30 the Beck Fast Screen for Medically Ill Patients,31 and the Hospital Anxiety and Depression Scale32 have been validated for use in MS populations); formal structured clinical interviews; routinely-collected health administrative data; and medical records. Each approach has strengths and limitations, but the gold standard remains the clinical interview based on the diagnostic and statistical manual of mental disorders.33  Regardless of the definition used, depression, anxiety, alcohol abuse, bipolar disorder, and schizophrenia have consistently been reported to occur at elevated frequencies in the MS population, relative to the general population.29,34 A systematic review reported a population-based point prevalence estimate of 24% for depression and 22% for anxiety.29 Both disorders have been associated with numerous negative consequences in MS, including increased risk of suicide, impaired cognition, and a worse quality of life.28 Depression predicts progression and mortality in other chronic immune-mediated diseases, including cancer35 and HIV.36 It’s plausible that a similar pattern may exist in MS. Several cross-sectional studies have examined the correlation between depression and MS disability, with mixed findings.37–39  Both depression and anxiety have consistently been associated with smoking and alcohol use in the general population.25 Tobacco and alcohol can lead to damage of the CNS40,41 which is 6  already compromised in people with MS. Cigarette smoking is an established risk factor for MS onset and progression;42 recent estimates suggest that about half of people with MS identify as ever-smokers, and approximately 1/5 identify as current smokers,.26,43 Alcohol use refers to the consumption of alcohol without the sole purpose of becoming intoxicated. Alcohol use disorders can be defined as abuse or dependence (also known as alcoholism). Alcohol abuse refers to the habitual misuse of alcohol, while alcohol dependence goes beyond misuse to encompass an addiction to alcohol, including tolerance, withdrawal, and loss of control.44 The relationship between alcohol use and MS is less clear – studies have reported increased,45 decreased,46 and no risk47 of MS associated with alcohol consumption. 1.3 Adherence to medication  1.3.1 Introduction Similar to alcohol dependence and smoking, non-adherence to medication can be viewed as a maladaptive health behaviour, potentially indicative of passive coping. Interestingly, adverse behaviours like alcohol use and smoking have been associated with poor adherence in the general population, as has depression.48,49 Adherence has been defined as ‘the extent to which patients take medications as prescribed by their health care providers’.50 The term adherence can refer to a measure of the percentage of doses taken as prescribed (often termed ‘compliance’), and/or the act of continuing treatment for a prescribed duration (often termed ‘persistence’).51 Maintaining adherence to long-term therapies is notoriously challenging; the World Health Organization estimates that only 50% of persons with chronic disease are adherent to their medication.49 Medication non-adherence in chronic conditions such as cardiovascular 7  disease, asthma, and HIV has been associated with clinically significant increased mortality, morbidity, hospitalization rates, and healthcare costs.49   1.3.2 Adherence to MS disease-modifying therapies The first disease-modifying therapies (DMTs) for the treatment of relapsing-remitting multiple sclerosis (MS) were approved in Canada in the 1990s,52 and initially included: the beta-interferons (IFNβ), Betaseron® (IFNβ-1b subcutaneous), Rebif® (high and low dose IFNβ-1a subcutaneous), and Avonex® (IFNβ-1a intramuscular); and glatiramer acetate subcutaneous (Copaxone®). They showed modest efficacy through reductions in relapse rates and MRI lesions in the pivotal clinical trials.53–56   Estimates of adherence in MS vary widely; a recent review suggested that rates of adherence range between 41% and 88%.57  The reasons for poor adherence in MS are complex and varied,58–60 but the outcome is likely universal: inadequate treatment reduces drug effectiveness. Hospitalizations, relapses, and general medical costs have been reported as increased among people with MS who do not adhere well to their prescribed DMT.61,62 1.4 Rationale The broad goal of this doctoral thesis is to contribute to the wider understanding and impact of mental health and health behaviours in persons with MS. The historical neglect of these areas has left many important and complex issues unresolved or unexplored, with many unmet needs. These include issues related to the potential relationships between factors, such as depression, anxiety, alcohol dependence, smoking, adherence to medication, and disability progression. These outcomes do not occur in isolation, but rather are influenced by one another and a range of 8  other clinical and demographic characteristics. Understanding the components that contribute to the total burden of illness and how they relate would be useful in developing strategies to prevent or reduce the severity of these effects. The goals of this dissertation were to bring these topics together in, where possible, a population-based setting and involving a multi-disciplinary team of researchers and clinicians. The objectives and questions had to be of direct relevance to the MS community, clinicians and stakeholders. 1.5 Specific objectives, rationales, and hypotheses 1.5.1 Chapter Two: Adverse health behaviours are associated with depression and anxiety in MS Objective: To evaluate the association between adverse health behaviours and mental health among people with MS in a prospective multi-site longitudinal study.  Rationale: Mental health disorders have been associated with adverse health behaviours, such as smoking and alcohol dependence in the general population;63,64 however, these relationships have not been established in MS.65 We aimed to evaluate the association between adverse health behaviours (cigarette smoking and alcohol dependence) and mental health (depression and anxiety) among people with MS.  Hypothesis: Adverse health behaviours will be associated with higher rates of depression and anxiety in this population.   9  1.5.2 Chapter Three: Determinants of non-adherence to disease-modifying therapies in MS Objective: To estimate adherence rates to the injectable disease modifying therapies in a prospective multi-site longitudinal study and examine clinical and demographic characteristics potentially associated with non-adherence.   Rationale: As recognized by the World Health Organization: ‘Adherence to therapies is a primary determinant of treatment success. Poor adherence attenuates optimum clinical benefits and therefore reduces the overall effectiveness of health systems.’49  In MS, non-adherence has been associated with increases in MS-related hospitalizations and relapse rates.61,62 To maximize adherence, an understanding of potentially modifiable factors that are associated with non-adherence is needed. Depression, anxiety, and cognitive difficulties have been associated with poor adherence to the disease modifying therapies for MS; however, findings have been inconsistent.57,60,66–68 Less is known about the effect of MS symptoms such as pain, fatigue and other comorbidities on drug adherence. If we can establish a demographic pattern of non-adherence, we could potentially develop a more targeted approach to improving adherence.  Hypothesis: Depression, anxiety, and alcohol dependence will be associated with non-adherence to the MS DMTs.  1.5.3 Chapter Four: Disease-modifying therapies and adherence in MS: comparing patient self-report with pharmacy records Objective: To compare self-reported DMT use and adherence against pharmacy records. Specifically, we aimed to assess the level of agreement between a practical, self-report 10  questionnaire and both short and longer-term pharmacy records of prescriptions filled with respect to: a) DMT use (any DMT being currently used); b) specific DMT product (identified by brand name); and c) adherence to a DMT. Rationale: Understanding a patient’s adherence is necessary to facilitate appropriate treatment decisions and disease management. However, non-adherence is challenging for healthcare providers to measure and predict.50 There are multiple methods used to define adherence, but no validated ‘gold standard.’49 Each approach has strengths and limitations;50 and in choosing a method a balance is struck between practicality and precision. Health administrative data, including pharmacy records, offer the opportunity to access ‘objective’ information (actual prescriptions filled), and are being validated and used for research purposes with increasing frequency.69,70 Such data are not always accessible to researchers or prescribing clinicians, however. Patient self-report offers a pragmatic approach, providing results which can be reviewed and used within the same clinical encounter to guide treatment options. Consequently, it is the most widely used adherence measure in the clinical setting.71 Despite the regularity with which self-report and pharmacy records are employed as medication adherence measures, 59,60,69,70 little is known about the agreement between the two.  Hypothesis: There will be high agreement between pharmacy records and self-report in terms of being on a DMT at the time of the clinic visit, and specific DMT product (brand), but lower agreement between sources in estimating adherence. Specifically, estimates of adherence will be higher (more optimistic) based on self-report relative to pharmacy records.  11  1.5.4 Chapter Five: Association between psychiatric comorbidity and disability progression of MS Objective: To investigate the relationship between psychiatric comorbidity and neurologic disability progression in a large multi-site MS clinic population.  Rationale: Depression co-existing with chronic disease leads to significantly greater disease burden and disability compared to having only depression, or only a chronic disease.23 In MS, having a comorbid psychiatric condition has been shown to be correlated with a  worse quality of life;72,73 however, little is known about the extent of its impact on other facets of the disease.65 The course of MS is challenging to predict, and there are few recognized factors that have been associated with progression or worsening of disability.74 Despite the obvious significance of mental health and physical function to persons with MS, little is known about the associations between the two.38,75   Hypothesis: The presence of a psychiatric condition will be associated with worsening in disability progression.  1.5.5 Chapter Six: A population-based study comparing multiple sclerosis clinic users and non-users in British Columbia, Canada Objective: To characterize a clinical cohort in the context of the wider MS population by comparing incident MS cases who were MS clinic users to non-users of the specialty MS clinics in British Columbia, Canada.  12  Rationale: All studies contained within this dissertation involve persons who attend an MS specialty clinic. Further, much of the medical literature and related knowledge surrounding MS is based on information gained from patients who attend MS specialty clinics affiliated with academic centres or tertiary care facilities, whether through observational studies or enrolment in clinical trials.52,74,76,77 A major advantage of studying these cohorts78–83 is the availability of MS-specific clinical information, such as disability scores and disease course as well as confidence in the diagnosis of MS. Whether or not these clinic patients are representative of the wider population of people with MS, however, is largely unknown.83,84 An understanding of the differences and similarities between MS clinic and non-clinic users would be helpful and relevant to clinicians and researchers who recruit from or study clinic-based cohorts as well as healthcare planners and related stakeholders. Hypothesis: MS clinic users and non-users will differ with respect to their demographic and clinical features: specifically age, socioeconomic status, use of health services (hospitalizations and physician visits), and comorbidities. 1.6 Data sources and cohorts 1.6.1 British Columbia MS Database Established in 1980, the British Columbia Multiple Sclerosis (BCMS) database includes clinical information from people who attended one of the four MS clinics in BC. These four clinics, located in Vancouver, Victoria, Kelowna, and Prince George, were the only source of specialty MS care in the province until January 1st 2005, when a fifth clinic opened in Burnaby, which did not contribute to the database.  The BCMS database contains largely prospectively-collected 13  information generated by MS-specialist neurologists on approximately 10,000 patients (as of 2016). The diagnosis of clinically-definite MS, according to the prevailing diagnostic criteria of the period,85–87 the date of MS symptom onset, relapse-specific information, EDSS scores, and disease course (i.e. relapsing-remitting or primary-progressive onset), as well as demographic information such as date of birth and sex are included in the database.    1.6.2 Cross-Canada longitudinal cohort From July 2010 to March 2011, consecutive patients attending a routine visit at one of four participating MS Clinics in British Columbia, Alberta, Manitoba and Nova Scotia were recruited. This study was nested within a larger cross-Canada collaboration funded by a Canadian Institutes of Health Research team grant (The Epidemiology & Impact of Comorbidity on MS in Canada) led by Dr. Ruth Ann Marrie. Participants were approached by a trained research coordinator at each site using a standardized script and invited to participate in a study involving the completion of specific questionnaires relating to comorbidity and quality of life at three time-points over two years. The three time-points (‘baseline’, year one, and year two) coincided with the typical annual visits to the respective MS clinic. For individuals unable to attend their annual clinic visit, follow-up questionnaires were offered via telephone, mail or email to minimize loss to follow-up. Inclusion criteria were: a confirmed diagnosis of definite MS or clinically isolated syndrome (CIS) according to the prevailing diagnostic criteria at the time the participant had been diagnosed; 85–87 age ≥18 years; resident in the province where data collection was occurring; ability and willingness to provide informed consent and to complete the study questionnaires in English. Institutional ethics approval was obtained at all sites. The informed consent form is provided in Appendix B.  14  Clinical and demographic information were captured from medical records using a standardized data abstraction form, including information on sex, date of birth, date of MS symptom onset, clinical course (relapsing-remitting, secondary progressive, primary progressive, clinically isolated syndrome),88 and EDSS (at baseline and years one and two).  Questionnaires were completed at each time point, including: The Hospital Anxiety and Depression Scale (HADS); 32 the Fatigue Impact Scale for Daily Use (D-FIS);89 the CAGE questionnaire (screen for alcohol dependence);90 the Health Utilities Index (HUI, Mark III version); a validated comorbidity questionnaire (capturing physical and psychiatric comorbidities);91 and smoking status (captured as ‘current, past, or never’). The questionnaire is available in Appendix C, excluding the rating scales that are under copyright (HADS, D-FIS, and HUI). Highest education level achieved and race were recorded at baseline.  1.6.3 Provincial health administrative data Population-based, individual-level information was available within several province-wide health administrative databases, covering health care and services, population and vital statistics. Canada has a universal health-care programme, which is publicly funded and administered at the provincial level. All citizens qualify for health coverage, regardless of medical history or income.92   The Medical Service Plan (MSP)93 payment information file provided information on physician visits across the province of British Columbia. The dataset covers all fee-for-service claims, including the date of service and accompanying diagnostic code outlining the primary reason for the visit. Beginning April 1st, 1991, the MSP file was coded per the International Classification of Diseases, Ninth Revision (ICD-9) diagnostic codes. These codes were developed by the 15  United Nations World Health Organization to systematically study disease, and are the global diagnostic standard for clinical and research purposes.94  The Discharge Abstract Database (DAD)95 provided data on all hospital admissions in the province. Visits to emergency rooms/departments are not recorded in the DAD. Similar to the MSP database, the DAD is coded using the ICD Ninth or Tenth Revision diagnostic codes, but with up to 25 distinct ICD codes included on the discharge report, along with the date of admission and discharge.95  PharmaNet,96 BC’s province-wide prescription database, provided data on prescriptions dispensed in the community, coded according to Health Canada’s Drug Identification Number (DIN) classification system.  Drugs used within the hospital setting (i.e., whilst a patient is hospitalized) are not included in PharmaNet, although out-patient hospital prescriptions are. Pharmacists are responsible for inputting information on the drug (using the DIN), date dispensed, quantity dispensed and days supplied.96 We linked the DINs to the more globally-relevant World Health Organization’s Anatomic Therapeutic Class (ATC) classification system.  The BC Ministry of Health’s Consolidation File,97 provided demographic information (birth date and sex) as well as days registered in the province. Together with the BC Vital Statistics Agency database 98 which provided mortality data, we were able to confirm that an individual was alive and resident in BC. Census Geodata provided an area-level measure of socio-economic status (SES), based on postal code and aggregated neighbourhood-level income data from Statistics Canada, which enabled SES to be categorized into quintiles (1= lowest SES; 5=highest SES).99 16  All linkages were facilitated by Population Data BC, a pan-provincial comprehensive data platform (http://www.popdata.bc.ca). Once the data were linked the personal-identifiers were removed, such that these individuals could not be re-identified. . All data were linked and locked prior to the initiation of this dissertation.All inferences, opinions, and conclusions drawn in this dissertation are those of the authors, and do not reflect the opinions or policies of the Data Steward(s). For a schematic outlining the multiple linkages between databases, please see Figure A.1, Appendix A. 1.7 Approach to analysis To evaluate the relationships between multiple predictor variables and a dependent variable, we  employed multivariable regression modelling techniques. Prior to building multivariable models, we explored the characteristics of each variable independently through univariate analyses. Categorical variables were reported as frequencies (percentages), and the distributions of continuous variables were graphed and reported as means (standard deviation) if normally distributed, or medians (interquartile range) if the distribution was skewed.  Second, bivariate analyses were used to analyze the relationship between two variables, including the relationship between the independent variable of interest and the dependent variable, as well as relationships between independent/dependent variables and potential confounders. All studies outlined in the dissertation are observational, and therefore susceptible to confounding. Confounding can be thought of as a ‘confusion of effects.’100 In other words, a confounder is an extraneous variable that can mistakenly make it appear that an observed exposure is associated with an outcome.100 The Chi-squared test or Fisher’s Exact test were used 17  when both variables were categorical.101 The student’s t-test or the Wilcoxon rank sum test were employed to compare continuous variables between groups.101   Third, to address the potential for confounding within the available data, we employed multivariable statistical techniques. Multivariable regression models were built using forward selection, in which a crude model containing only the independent variable of interest and the outcome was evaluated first, and then covariates were added sequentially. Covariates were included based on their established or potential association with the outcome, or because they reached a threshold level of statistical significance (p<0.1). Interactions between predictor variables were checked by creating a product term between the two variables and incorporating it into the model. If an interaction term met significance, using a threshold of p<0.10, it was considered for inclusion in the multivariate analysis. Only interaction terms which remained statistically significant (p<0.05) in the final model were retained. Further, all analyses were stratified by sex. If the risk estimates differed significantly (confidence intervals did not overlap) between men and women, then the stratified analyses were described. If not, the risk estimate for all persons was reported.  Logistic regression models were used for binary outcomes, poisson (or negative binomial if the data were overdispersed) for count data, and linear regression for continuous outcomes.101 Each of these methods requires that all data points are independent. This dissertation was developed using longitudinal datasets of repeated measures for individuals. To account for the within-subject correlations (or clustering) of the repeated measures for each person, we employed generalized estimating equations (GEE). GEE is a semiparametric regression technique which runs a two-step process in which the model is fitted without accounting for the correlations 18  within subjects, and then subsequently ‘re-fitted’ by integrating a working correlation structure to adjust for these correlations.102 The resulting risk estimate is an average of between and within-subject effects, but tends to be dominated by between-subject effects.102 The working correlation structure is established a priori based on the data structure and some conjecture regarding the possible association between repeated measures. In the cross-Canada longitudinal cohort, only three time points were assessed for each individual. As such, the model could process an unstructured correlation structure, in which all correlations are assumed to be different. This is the least restrictive possible structure. An exchangeable structure assumes that the correlations across different time periods are equal for an individual, regardless of the time intervals. This method is analogous to a mixed model analysis, and must be used in instances when there are many repeated measures for an individual, as in Chapter 5.     To test for model fit of the GEE we used the quasi-likelihood under the independence model criterion (QIC) statistic. The QIC is comparable to the Akaike’s information criterion (AIC) statistic, which compares models fitted with likelihood-based methods.102 GEE is not a likelihood-based method, therefore AIC cannot be employed.103 A smaller QIC indicates a better model fit when comparing nested models.103  Analyses unique to each study are outlined in detail in the subsequent chapters, but the general approach is consistent with the process outlined in this subsection.   19  Chapter 2: Adverse health behaviours are associated with depression and anxiety in multiple sclerosis1 2.1 Background and objective The co-occurrence of alcohol dependence and smoking with psychiatric disorders is well-established in the general population,25 as are the neurological consequences of such behaviours.40,41 Smoking leads to deleterious effects on both the immune system and the CNS through increases in nitric oxide levels which contribute to axonal degeneration,41 and augmentation of pro-inflammatory cytokines.104 White matter is particularly vulnerable to damage related to alcoholism, particularly in the corpus callosum, and frontal lobes.40  Despite the high prevalence of depression and anxiety in the MS population, few studies have explored their relationships with smoking or alcohol use.  Objective: To evaluate the association between adverse health behaviours and mental health among people with MS in a prospective multi-site longitudinal study.  2.2 Methods Please refer to Chapter 1 (Subsection 1.6.2) for an outline of the study population. Details of the methodology specific to this study are described below.                                                  1 A version of this chapter has been published as McKay KA, Tremlett H, Fisk JD, Patten SB, Fiest K, Berrigan L, Marrie RA. Adverse health behaviours are associated with depression and anxiety in multiple sclerosis: A prospective multisite study. Mult Scler 2016; 22(5), 685-693.  20  2.2.1 Study population Briefly, from July 2010-March 2011 we recruited consecutive patients attending a routine visit at one of four participating MS Clinics across Canada. Demographic and clinical information were captured from each individual’s medical record.  2.2.2 Self-reported health behaviours and mental health Each participant completed several questionnaires at the three time points (baseline, year one, and year two). The Hospital Anxiety and Depression Scale (HADS) was the primary measure of current mental health status. It is a valid and reliable 14-item questionnaire that assesses symptoms of anxiety and depression separately, and was created for use in the setting of hospital medical outpatient clinics.105 It is robust across different age groups, education levels, and genders.106 Validated cut-off scores of ≥8 on the HADS scale were used to define depression and anxiety.32 This cut-off was estimated to have a sensitivity of 90% and a specificity of 87.3% for major depressive disorder, and a sensitivity of 88.5% and specificity of 80.7% for generalized anxiety disorder, as defined by the Structured Clinical Interview (DSM-IV) among persons with MS.32 To screen for alcohol dependence, we used the CAGE (Cutting down, Annoyance by criticism, Guilty feeling, Eye-openers) questionnaire. A score of ≥ 2 out of a possible 4 suggests alcohol dependence, which was found to be highly sensitive (93%) relative to careful inquiry on alcohol dependence.90 Smoking was captured as ‘current’, ‘past’, or ‘never’ (less than 100 cigarettes smoked over a lifetime).107  21  2.2.3 Statistical analysis The presence of depression, anxiety, smoking, and alcohol dependence was described as frequencies (percentages). The baseline clinical and demographic characteristics (sex, race, age, EDSS, education level, and site) of patients were compared to the two mental health issues (anxiety, depression) and two health behaviours (alcohol dependence, smoking) of interest; initially using the Pearson χ2 test or Fisher’s exact test, then by logistic regression.  We aimed to explore the association between health behaviours and mental health. Given the uncertainties regarding the direction of this relationship, and to maximally use the available data, we initially conducted prevalence analyses, followed by more targeted incidence analyses. We explored the relationship between prevalent health behaviours and prevalent depression and anxiety using logistic generalized estimating (GEE) equations with an unstructured correlation matrix. This method allows for the simultaneous analysis of data from all time points while accounting for correlations between the repeated measures for individuals. Analyses of the odds of having depression or anxiety were measured based on the presence of alcohol dependence or current smoking status, the latter defined as non-smoker or current smoker. Second, to evaluate the incidence of anxiety or depression during the course of the study, we excluded all participants who met criteria for either condition at baseline and then evaluated the risk of developing either condition at year one or two based on baseline health behaviours using logistic regression. Given that the literature in the general population has reported an increased risk of adverse health behaviours secondary to mental health conditions we also evaluated the risk of incident alcohol dependence and smoking.25 Covariates that were considered included current age, sex, concurrent disability status measured by the EDSS (categorized as mild [0-2.5], 22  moderate [3.0-5.5], or severe [6.0+]), baseline education level (categorized as high school or less, any post-secondary or more, and other), and site. Covariates were included either for clinical relevance (e.g. age) or on the basis of their association with the outcome from the baseline analysis (p<0.1). Findings from all regression analyses were expressed as odds ratios (ORs) with corresponding 95% confidence intervals (CI). All analyses were performed using SAS Statistical Software Package 9.4 (SAS Institute Inc., Cary, NC).   2.3 Results Of 1632 patients who visited one of the four MS Clinics between July 2010 and March 2011, 1144 met the inclusion criteria (outlined in Chapter 1.6.2). Of these, 949 (82.6%) consented to participate. Reasons for non-participation cannot be reported as these individuals did not sign informed consent forms.  Over the course of the two follow-up assessments, 58 patients missed their year one assessment (and could not be contacted by phone, mail or email), and 64 missed the year two assessment, for a total of 93.2% (885/949) with complete follow-up. In total, 65 patients had a follow-up assessment by telephone interview. The mean age at baseline was 48.6 years, 75.2% were women and most had a relapsing-remitting course (72.4%). Most participants were white (94.6%), and had achieved a post-secondary education or higher (67.0%) (Table 2.1).       23      Total Cohort Cases of incident anxiety Cases of incident depression Cases of incident alcohol dependence Variable n= 949 n = 138 n= 128 n= 42 Sex, N (%)  Female     Male  714 (75.2) 235 (24.8)  99 (71.7) 39 (28.3)  96 (75.0) 32 (25.0)  25 (59.5) 17 (40.5) Race, N (%)  White  Non-White  810 (94.6) 46 (5.4)  121 (94.5) 7 (5.5)  101 (91.0) 10 (9.0)  33 (97.1) 1 (2.9) Age, mean (SD) 48.6 (11.4) 48.5 (11.2) 48.8 (11.3) 47.0 (11.5) Age (categorized), N (%)           18 – 29           30-39           40-49           50+  58 (6.1) 149 (15.7) 305 (32.2) 436 (46.0)  11 (8.0) 20 (14.5) 42 (30.4) 65 (47.1)  9 (7.0) 17 (13.3) 47 (36.7) 55 (43.0)  4 (9.5) 7 (16.7) 13 (31.0) 18 (42.9) Age of onset, mean (SD) 33.2 (10.0) 33.5 (10.6) 32.2 (9.9) 32.7 (9.0) Disease duration, mean (SD) 15.4 (10.2) 14.9 (9.9) 16.1 (10.0) 14.3 (8.0) EDSS, median (IQR) 2.5 (1.5-5.0) 3.0 (2.0 – 6.0) 3.0 (2.0 – 5.0) 2.5 (1.5 – 4.0) Clinical Course, N (%)  RRMS  SPMS  PPMS  CIS  Unknown  687 (72.4) 193 (20.3) 60 (6.3) 5 (0.5) 4 (0.4)  91 (65.9) 37 (26.8) 8 (5.8) 2 (1.5) 0 (0.0)  92 (71.9) 29 (22.7) 7 (5.5) 0 (0.0) 0 (0.0)  34 (81.0) 7 (16.7) 1 (2.4) 0 (0.0) 0 (0.0) Current Disease Modifying Therapy Use   Yes  No   477 (50.4) 470 (49.6)   64 (45.7) 76 (54.3)   75 (58.6) 53 (41.4)   21 (50.0) 21 (50.0) Education High School or Less Any Post-Secondary or More Other  258 (30.1) 574 (67.1) 24 (2.8)  36 (28.1) 91 (71.1) 1 (0.8)  36 (32.4) 72 (64.9) 3 (2.7)  7 (20.6) 26 (76.5) 1 (2.9) Table 2.1 Demographic and clinical characteristics of study participants at baseline. 24  The occurrence of mental health disorders (anxiety and depression) and adverse health behaviours (alcohol and smoking) over the study period is summarized in Table 2.2. Over the entire study, 53.8% of participants met the HADS-anxiety criterion and 35.1% met the HADS-depression criterion at some point (baseline, year 1, or year 2) (Table 2.2). The percentage of patients who met criteria for alcohol dependence, as measured by a score of ≥2 on the CAGE questionnaire, ranged from 6.0% to 6.6% over the three assessments (Table 2.2). At baseline, 56.7% of respondents had ever smoked (≥ 100 cigarettes in a lifetime), of whom 63.7% no longer smoked. Of the total cohort, 20.6% self-identified as current smokers. Of the past smokers, 52.6% quit after their MS onset and only 4.7% of ever smokers began smoking after their MS onset.  Baseline Year One Year Two ‘Ever met criteria’ (at baseline or year 1 or 2) ‘Met criteria at all three time points’ (at baseline and year 1 and 2)  Mental Health      Anxiety, N (%) 366/932 (39.3) 320/907 (35.3) 309/882 (35.0) 511/949 (53.8) 170/868 (19.6) Depression, N (%) 200/930 (21.5) 207/907(22.8) 197/882 (22.3) 333/949 (35.1) 89/868 (10.3) Adverse health behaviours      Alcohol Dependence, N (%) 60/946 (6.3) 60/906 (6.6) 53/882 (6.0) 102/949 (10.7) 18/868 (2.1) Never Smoker, N (%) 411/947 (43.4) 398/911 (43.7) 382/885 (43.2) 384/949 (40.5) 358/732 (48.9) Past Smoker, N (%) 341/947 (36.0) 330/911 (36.2) 336/885 (37.9) 333/949 (35.1) 272/732 (37.2) Current Smoker, N (%)  195/947 (20.6) 180/911 (19.8) 165/885 (18.7) 232/949 (24.4) 102/732 (13.9) Table 2.2 Occurrence of mental health and adverse health behaviours over the study period.    25  At baseline, anxiety was associated with female sex (OR: 1.55; 95%CI: 1.13-2.13), and reduced odds of severe disability (OR: 0.64; 95%CI: 0.45–0.91), while depression was associated with higher odds of severe disability (OR: 2.71; 95%CI: 1.85–3.99) when compared to mild disability. Both alcohol dependence (1.83; 95%CI: 1.06–3.16) and smoking every day (OR: 1.73; 95%CI: 1.20–2.50) were reported more frequently by men (Table 2.3). Alcohol dependence also was reported more frequently among non-white patients (OR: 3.19; 95%CI: 1.35–7.55) and those with moderate disability (compared to mild, OR: 2.48; 95%CI; 1.39–4.44). Patients who completed a post-secondary education or higher were less likely to currently smoke (OR: 0.50; 95%CI: 0.35–0.73). Age was not associated with any of the mental health or health behaviours of interest (p>0.05). All baseline associations are outlined in Table 2.3. 26  Table 2.3 Frequencies of baseline clinical and demographic variables associated with the four variables of interest: anxiety, depression, alcohol dependence, and current smoking status.   a By Pearson’s chi-squared test. b By Fisher’s Exact Test  Predictor Variable Anxiety  (n= 366) p-value Depression (n=200) p-value Alcohol Dependence (n=60) p-value Current Smoker  (n=195) p-value Sex         Females Males 294/704 (41.8%) 72/228 (31.6%) 0.006a 151/702 (21.5%) 49/228 (21.5%) 0.995 a 38/711 (5.3%) 22/235 (9.4%) 0.029 a 131/714 (18.4%) 64/235 (27.2%) 0.003 a Race         White Non-white 305/793 (38.5%) 21/46 (45.6%) 0.331 a 165/791 (20.9%) 11/46 (23.9%) 0.621 a 43/807 (5.3%) 7/46 (15.2%) 0.014b 170/810 (21.0%) 8/46 (17.4%) 0.559a Age         18 – 29 30-39 40-49 50+ 20/56 (35.7%) 65/144 (45.1%) 123/304 (40.5%) 158/428 (36.9%) 0.314 a 4/54 (7.4%) 29/145 (20.0%) 68/303 (22.4%) 99/428 (23.1%) 0.061 a 7/58 (12.1%) 10/149 (6.7%) 18/305 (5.9%) 25/434 (5.8%) 0.310 a 13/58 (22.4%) 34/149 (15.7%) 70/305 (23.0%) 78/436 (17.9%) 0.314a EDSS         Mild [0-2.5] Moderate [3.0-5.5] Severe [6.0 +] 199/475 (41.9%) 84/210 (40.0%) 64/202 (31.7%) 0.043 a 73/472 (15.5%) 51/210 (24.3%) 67/202 (33.2%) < 0.0001a 25/487 (5.1%) 25/211 (11.9%) 7/202 (3.5%) 0.0006a 99/487 (20.3%) 49/211 (23.2%) 39/204 (19.1%) 0.558a Education Level         High School or Less Any Post-Secondary or More Other 94/254 (37.0%) 221/561 (39.4%) 9/24 (37.5%) 0.805a 61/253 (24.1%) 111/560 (19.8%) 5/24 (20.8%) 0.382a 16/256 (6.3%) 32/573 (5.6%) 2/24 (8.3%) 0.813a 73/258 (28.3%) 98/573 (17.1%) 6/24 (25.0%) 0.0009a Site         Nova Scotia Manitoba Alberta British Columbia 156/392 (39.8%) 42/121 (34.7%) 32/81 (39.5%) 136/338 (40.2%) 0.745a 83/392 (21.2%) 26/119 (21.9%) 23/81 (28.4%) 68/338 (20.1%) 0.441a 13/400 (3.3%) 10/121 (8.3%) 3/84 (3.6%) 34/341 (10.0%) 0.0012a 93/401 (23.2%) 26/121 (21.5%) 17/86 (19.8%) 59/341 (17.3%) 0.260 a 27  In the prevalence analyses, alcohol dependence was associated with increased odds of anxiety (OR:1.88; 95%CI:1.37–2.57). The association persisted after adjusting for age, sex, EDSS, and smoking status (OR:1.84; 95%CI:1.32–2.58, see Table 2.4). Alcohol dependence was also associated with increased odds of depression (OR:1.51; 95%CI:1.07–2.15), which remained statistically significant after adjusting for age, sex, EDSS, and smoking status (OR:1.53; 95%CI:1.05–2.23).  Smoking was assessed in a similar manner to alcohol, with current smokers compared to non-smokers across the study period. Smoking was associated with increased odds of anxiety (unadjusted OR:1.32; 95%CI:1.05–1.65). When adjusted for age, sex, EDSS, and alcohol dependence, the relationship persisted (OR:1.29; 95%CI:1.02–1.63). There was also an association between smoking and depression (unadjusted OR:1.42; 95%CI:1.10–1.83), which was tempered slightly after adjusting for age, sex, EDSS, and alcohol dependence, but remained statistically significant (OR:1.37; 95%CI:1.04–1.78) (see Table 2.4).   Association between anxiety and: Crude OR (95% CI) Adjusted OR* (95%CI) No Alcohol dependence (Reference) 1 1 Alcohol dependence 1.88 (1.37 – 2.57) 1.84 (1.32 – 2.58) Non-Smoker (Reference) 1 1 Current Smoker 1.32 (1.05 – 1.65) 1.29 (1.02 – 1.63) Association between depression and: Crude OR (95% CI) Adjusted OR+ (95%CI) No Alcohol dependence (Reference) 1 1 Alcohol dependence 1.51 (1.07 – 2.15) 1.53 (1.05  – 2.23) Non-Smoker (Reference) 1 1 Current Smoker 1.42 (1.10 – 1.83) 1.37 (1.04 – 1.78) Table 2.4 Association between health behaviours and: i) anxiety; ii) depression across all three visits using GEE analysis with an unstructured correlation matrix.  *Adjusted for age, sex, EDSS, and smoking status. +Adjusted for age, sex, EDSS, and alcohol dependence.   28  For the incidence analysis, those with baseline depression (n=200) and anxiety (n= 366) were excluded. By study end there were 138 cases of incident anxiety and 128 cases of incident depression. There was no increased risk of developing anxiety at year one or two related to alcohol dependence at baseline (OR:0.46; 95%CI:0.14–1.59). However, there was an increased risk of depression (OR:3.19; 95%CI:1.65–6.17). This relationship remained after adjusting for age, sex, EDSS, and smoking status (see Table 2.5). Smoking at baseline was not associated with an altered risk of subsequent anxiety (OR:1.22; 95%CI:0.73–2.05) or depression (OR:1.01; 95% CI:0.60–1.72) during the follow-up period (see Table 2.5).  Because none of the participants began smoking during the study period, we could not examine the risk of smoking. However, 42 patients developed alcohol dependence during the study period.  When adjusted for covariates, we found that baseline depression increased the risk of developing alcohol dependence during follow-up (OR:2.12; 95%CI: 1.04–4.30). Anxiety was not associated with an increased risk of alcohol dependence (Table 2.5).            29  Association between incident anxiety at year one or two and baseline: Crude OR (95% CI) Adjusted OR* (95%CI) No Alcohol dependence (Reference) 1 1 Alcohol dependence 0.46 (0.14 – 1.59) 0.43 (0.13 – 1.50) Never Smoker (Reference) 1 1 Current Smoker 1.22 (0.73 – 2.05) 1.20 (0.71 – 2.04) Association between incident depression at year one or two and baseline: Crude OR (95% CI) Adjusted OR* (95%CI) No Alcohol dependence (Reference) 1 1 Alcohol dependence 3.19 (1.65 – 6.17) 2.97 (1.48 – 5.96) Non-Smoker (Reference) 1 1 Current Smoker 1.01 (0.60 – 1.72) 0.94 (0.54 – 1.64) Association between incident alcohol dependence at year one or two and baseline: Crude OR (95% CI) Adjusted OR* (95%CI) No Anxiety (Reference) 1 1 Anxiety 1.06 (0.56 – 2.02) 1.18 (0.60 – 2.33) No Depression (Reference) 1 1 Depression  1.82 (0.92 – 3.59) 2.12 (1.04 – 4.30) Table 2.5 Association between baseline health behaviours and incident mental health at year one and two.  *Adjusted for age, sex, EDSS, and smoking status or alcohol dependence.  2.4 Discussion We found that alcohol dependence and smoking were associated with both anxiety and depression in this large, representative cohort of clinic-attending MS patients. Smoking at study entry had no measurable influence on the risk of developing subsequent incident mental health symptoms. However, alcohol dependence at study entry was associated with increased risk of incident depression, and depression at baseline increased the risk for incident alcohol dependence during two years of follow-up. Surprisingly few studies have examined the relationships of health behaviors and mental health symptoms in MS despite a general recognition of the importance of these issues and extensive study in the general population.25 Similar to our own findings, the lone prior longitudinal study 30  that examined the association between cigarette smoking and mental health in MS patients also found a higher risk of anxiety and depression in smokers, a link that is well-established in the general population.25 We found three cross-sectional studies examining the relationship between alcohol use and mental health in MS, with mixed findings.108–110 A study of 140 clinic patients in the province of Ontario, Canada reported an association between anxiety, but not depression, and problem drinking as measured using the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders – Fourth Edition or weekly consumption of alcohol (14 drinks or more for males, 9 for females).108 A USA-based study of MS clinic-attending patients (n=157) reported no associations between alcohol dependence (measured using the Alcohol Use Disorders Identification Test Consumption and CAGE) and depression or anxiety.109 Last, a study of a community sample from the USA (n = 739) that utilized a mail-out survey suggested that depression and possible problem drinking were associated (anxiety was not examined).110 The lack of consistency amongst the findings of these studies and our own may be due to the varied assessments of both mental health symptoms and alcohol dependence, as each study employed different scales. The relatively small sample sizes of previous clinic-based studies may also explain at least some of the disparity in results, especially given that the rates of problem drinking were generally low.  However, our finding that the odds of anxiety were nearly two-fold in those with alcohol dependence is in concordance with two large surveys of the general population, which reported odds ratios between 1.5 and 4.0.111    We were unable to find any other prior studies that had evaluated the risk of incident mental health disorders associated with alcohol dependence in people with MS. While we found no altered odds of anxiety associated with alcohol dependence this may have been related to the 31  small number of individuals with alcohol dependence who were at risk for incident anxiety. Studies from the general population that have attempted to elucidate the direction of the relationship between anxiety and alcohol dependence, have had inconsistent results, suggesting that the relationship may be cyclical.111 We found that the risk for incident depression at follow-up was increased among MS patients who met criteria for alcohol dependence at baseline. Moreover, the risk for alcohol dependence was also increased in those who were depressed at baseline, suggesting a bidirectional relationship.112 Given the consequences of depression and alcohol dependence,28,40 this finding is particularly concerning and points to the importance of monitoring of health behaviours and mental health as an integral part of MS care. We found that smoking at baseline did not alter the risk of developing incident anxiety or depression at follow-up.  For some, smoking may be a means of coping with mental health disorders, and therefore would not predict the onset of these disorders. There is evidence from the general population that smoking can lead to both anxiety and depression, but is more likely an indicator of risk than a true causal factor.113 None of the participants enrolled in this study began smoking during the study period, and more than one-third identified as past smokers. The increased prevalence of depression seen among smokers could be related to an inability to quit, due to nicotine dependence, social habituation, or possibly depression itself.114 Despite the belief among many smokers that quitting will lead to worsened mental health, smoking cessation is associated with reduced anxiety and depression.115 Smoking cessation should be a major focus in treating people with MS as it is the only modifiable lifestyle factor that has consistently been associated with worsened long-term disease progression.42  32  Our rates of smoking, alcohol dependence, anxiety, and depression were as expected, based on previous studies in MS. In our cohort, approximately 1/5 were current smokers and most of those who had ever been smokers were past smokers, similar to what has been shown in other MS cohorts.43 Men and participants with less time spent in formal education were more likely to identify themselves as smokers, similar to findings in the general population.116 Few participants began smoking after their MS onset, and most of those who identified as past-smokers quit after disease onset, suggesting that a diagnosis of MS may influence smoking patterns.  Alcohol dependence was reported in a minority of participants - approximately 6%, which is in accordance with previous studies of alcohol dependence in MS that also employed the CAGE questionnaire.109,110 Since a large Canadian general population-based survey reported similar results (5.8%),117 people with MS do not appear at an increased risk of alcohol dependence overall. Importantly however, the rates of alcohol dependence in our sample were correlated with level of disability; those with moderate disability appeared to be at the highest risk of alcohol dependence relative to mild or severe disability.  Patients with severe disability have a greater probability of experiencing bladder dysfunction,118 which may compel them to abstain. Mobility limitations or the need for full-time care may also reduce their access to alcohol. A recent systematic review reported a summary estimate of 23.7% (95%CI: 17.4 – 30.0%) for the prevalence of depression in people with MS, though rates ranged widely between studies, likely due to varied follow-up times and measures of depression.29 Anxiety has been studied less frequently, but prevalence estimates range from 1.2 – 44.6%.29 It is not clear whether these mental health disorders are caused by an endogenous mechanism related to the pathology of MS,119 or a psychosocial response to a chronic illness,120 or, most likely, a combination of both. We found higher rates of anxiety among women, while an equal proportion of men and women 33  met criteria for depression. Sex differences in mental health are well-established, but the reasons for this disparity are not. Social roles, coping style, and biological factors likely contribute to the higher incidence of depression and anxiety among females in the general population.121 In MS, men face a disproportionately higher rate of depression and anxiety than women, relative to the general population.122  We found no evidence for the modification of an effect of health behaviours on mental health by sex.  Strengths of this study included the large multi-site cohort, longitudinal design, high completion rates (>90%) over the three assessments and the use of validated questionnaires that have previously been employed in an MS population. The consecutive nature of recruitment and high proportion of individuals who agreed to participate suggests that our cohort is representative of individuals attending an MS clinic. Further, two of the clinics deliver the only MS care in their province (Manitoba and Nova Scotia) which also suggests these findings are generalizable to the wider MS population. A limitation of this study was the reliance on self-reported adverse health behaviours, which are susceptible to reporting bias.123 Removing all persons with anxiety at baseline (366/949; 38.6%) in the incidence analyses reduced the power to detect an association and may have resulted in a Type II error. Although this is the most common approach used to assess such behaviours and we used validated instruments, it remains possible that people have varied perspectives on what constitutes problem drinking, and some people may have been misclassified for this reason.   We examined the burden of comorbidity between adverse health behaviours and mental health, and the risk for developing mental health symptoms, and alcohol dependence. Alcohol dependence and smoking were associated with both anxiety and depression. Alcohol dependence 34  was also associated with an increased risk of developing subsequent depression, and depression was associated with incident alcohol dependence. Prevention or reduction of the severity of these mental health problems is of considerable importance in MS care. Aside from the general benefits associated with the promotion of healthy behaviours our findings suggest that promoting cessation of smoking and alcohol use could potentially contribute to improving mental health in MS. Physician advice to reduce alcohol use is effective,124 as are pharmacological treatments aimed at smoking cessation.125 Increased awareness of the effects of adverse health behaviours on mental health in MS might help target appropriate counselling and support for those ‘at risk’.  35  Chapter 3: Determinants of non-adherence to disease-modifying therapies in multiple sclerosis2  3.1 Background and objective Perhaps similar to alcohol dependence and smoking, adherence to medication may be considered a maladaptive behaviour, or reflect a process of passive coping.126  Therefore, it is conceivable that alcohol use and smoking might be associated with a higher risk of non-adherence. Depression and anxiety have also been associated with poor drug adherence in MS; however, findings have been inconsistent.57,60,66–68 In MS, non-adherence leads to increased MS-related hospitalizations and relapse rates.61,62 Identifying modifiable factors that are associated with non-adherence may be useful in order to target ‘at risk’ patients.  Objective: To estimate adherence rates to the injectable DMTs in a multi-site longitudinal study and examine clinical and demographic characteristics associated with non-adherence.   3.2 Methods Please refer to Chapter 1 (Subsection 1.6.2) for an outline of the study population. Details of the methodology specific to this study are described below.                                                  2 A version of this chapter has been published as McKay KA, Tremlett H, Patten SB, Fisk JD, Evans C, Fiest K, Campbell T, Marrie RA. Determinants of non-adherence to disease-modifying therapies in multiple sclerosis: A cross-Canada prospective study. Mult Scler 2017; 23(4), 588-596.  36  3.2.1 Study population Briefly, consecutive patients attending a routine visit at one of four MS Clinics across Canada were recruited and followed at three time-points. Specifically for this study, participants were included if use of an injectable DMT (beta-interferon or glatiramer acetate) was reported at least once during the study period (at baseline, year one, or year two). These injectable drugs were the predominant DMT for MS at the time of data collection, and the questionnaire specifically asked about the number of injections missed in the previous 30 days.  3.2.2 Self-reported information Participants completed questionnaires at the baseline, one and two-year follow-up visits. The questionnaire captured DMT use (by brand name) and number of missed injections in the previous 30 days. Comorbidities were recorded via a validated questionnaire;91 the total number of physical comorbidities for each participant was calculated and categorized as 0, 1, or ≥2. The Hospital Anxiety and Depression Scale (HADS) measured current symptoms of depression and anxiety.105 Validated cut-off scores of ≥8 on the HADS scale were used to define the presence of both.32 The CAGE (Cutting down, Annoyance by criticism, Guilty feeling, Eye-openers) questionnaire was used as a screening tool, with a score of ≥2 out of a possible 4 suggesting alcohol dependence.90 Smoking status was captured as either ‘current’ or ‘non’ smoker. Fatigue was measured using the Daily Fatigue Impact Scale, a 36-item scale, dichotomized as ‘no fatigue’ (<5) and ‘any fatigue’ (≥5).127 Health-related quality of life (HRQOL) was measured using the Health Utilities Index Mark III version (HUI-3),128,129 a 15-item measure that assesses health state with respect to eight single-attribute scores: vision, hearing, speech, mobility, dexterity, emotion, cognition, and pain. These single attribute scores were combined into an 37  overall score which can range from 0 (equivalent to death) to 1 (perfect health), which was categorized as “no to moderate disability” (>0.70) and “severe disability” (≤0.70).130  Individual attribute scales for pain and cognition were also examined separately as both are associated with MS, and with drug adherence in other patient populations.59,131 These were categorized as: pain that disrupts normal activities (yes vs. no; ≤0.77 vs. >0.77) and moderate to severe cognitive difficulties (yes vs. no; ≤0.70 vs. >0.70).  3.2.3 Quantifying adherence The medication possession ratio (MPR)57 was used to estimate adherence, calculated as the number of doses taken divided by the number of expected doses, expressed as a percentage at each study visit. A MPR of ≥80% was considered adherent, and <80% as non-adherent66,132–134 (see Table A.2, Appendix A for details). This definition was chosen as it is a common binary cutoff  in MS DMT adherence literature,57 thereby lending itself to comparisons with prior research. Further, non-adherence defined as <80% has been shown to be associated with important clinical outcomes in MS, including increased relapse rates, inpatient visits, and overall medical costs.61As a supplementary analysis, non-adherence was defined as any missed dose during the 30-day period (yes vs. no).  3.2.4 Statistical analysis MS-specific clinical features and demographic characteristics of the DMT users were described as frequencies (percentages), mean and standard deviation (SD), or median and interquartile range (IQR).  We employed generalized estimating equations (GEE) with an unstructured correlation matrix to examine characteristics associated with non-adherence at all three time 38  points. This approach allowed for the simultaneous analysis of both patient characteristics and adherence measures from all three time points, accounting for correlations between the repeated measures for individuals. If a person was on a DMT at only one of the three visits, only this visit would be included in the analysis; if they were on a DMT at all three visits, all three would be included. Sex, race, and education were captured at baseline only and included as constant variables, while the remaining predictor variables were included as time-varying covariates over the three visits. We employed univariate logistic regression using GEE, followed by a multivariable logistic regression based on the significance (p>0.1) of characteristics from the univariate analysis. To estimate the predictive effect of baseline non-adherence on non-adherence at follow-up (year one and two) we used logistic regression modeling, adjusted for confounders (all measured at baseline). Findings were reported as odds ratios (ORs) with 95% confidence intervals (CI). Analyses were performed using the Statistical Analysis System (SAS) Software Package 9.4 (SAS Institute Inc., Cary, NC). 3.3 Results Of 1632 patients who visited one of the four MS clinics, 949 consented and participated in the primary study,135 of which, 485 reported use of an injectable DMT during the study period and were included in the analyses (Table 3.1). Nine participants missed their year one visit, and 21 missed the year two visit, for a retention rate of 95.7%. Females and those of younger age and shorter disease duration were more likely to be on a DMT. The average age of DMT users was 45.5 years, average disease duration was 12.6 years, and most had RRMS (90%) (Table 3.1).  39  Baseline characteristics Exposed to an injectable DMT during study period Unexposed to an injectable DMT during study period p-value  n= 485 n=464  Sex, N (%)  Female     Male  383 (79.0) 102 (21.0)  331 (71.3) 133 (38.7)  0.007a Race, N (%) (36 missing)  White  Non-White  427 (94.3) 26 (5.7)  383 (95.0) 20 (5.0)  0.615a Age, mean (SD) Age Range (years) 45.5 (10.2) 19  – 71 52.1 (11.4) 19 – 80 <0.0001b Age of symptom onset, mean (SD) 32.9 (9.1) 33.7 (10.3) 0.240b Disease duration, mean (SD) 12.6 (8.7) 18.4 (9.0) <0.0001b EDSS, median (IQR) 2.0 (1.5-3.5) 3.5 (2.0 – 6.0) <0.0001c Clinical Course, N (%)  RRMS  SPMS             PPMS             CIS  RRMS at onset, but unknown if    reached SPMS  435 (89.7) 47 (9.7) 0 (0.0) 0 (0.0) 3 (0.6)  252 (54.4) 146 (31.5) 60 (13.0) 5 (1.1) 0 (0.0)  <0.0001d Education  High School or Less  Any Post-Secondary or More  Other  135 (29.6) 311 (68.2) 10 (2.2)  123 (30.8) 263 (65.8) 14 (3.5)  0.453a Table 3.1 Baseline clinical and demographic characteristics.  Patients who reported taking a disease-modifying drug at some point during the study period (baseline, year one, or year two) compared to participants who did not report taking an injectable therapy during the study period. a Pearson’s chi-squared test; b Student’s t-test; c Wilcoxon rank sum test; d Fisher’s exact test. RRMS = relapsing-remitting MS; SPMS = secondary progressive MS; PPMS = primary progressive MS; CIS = clinically isolated syndrome.   At baseline, 46% (435/949) of the participants were on a first-line DMT, which remained relatively steady over the follow-up period.  The frequency of use for each DMT also remained largely stable over the follow-up period, although use of the second-line DMTs, fingolimod and natalizumab, increased with time (Table 3.2).    40   Disease-modifying therapy (route and frequency) Baseline Year One Year Two None‡  47 (9.7) 62 (13.1) 78 (17.0) Interferon β-1a (intramuscular, weekly) 93 (19.2) 79 (16.6) 66 (14.4) Interferon β-1b (subcutaneous, every other day) 63 (13.0) 52 (11.0) 43 (9.4) Interferon β-1a (subcutaneous, three times per week) 146 (30.2) 132 (27.8) 111 (24.1) Glatiramer acetate (subcutaneous, daily) 133 (27.5) 136 (28.6) 127 (27.6) Natalizumab (intravenous, every 28 days) 1 (0.2) 8 (1.7) 14 (3.0) Fingolimod (oral, daily) 0 (0.0) 2 (0.4) 14 (3.0) Other 2 (0.2) 4 (0.8) 7 (1.5) Table 3.2 Disease-modifying therapy (DMT) use at baseline, year one, and year two. ‘Other’ includes clinical trials participants, mitoxantrone users, or unknown.  ‡all individuals were exposed to an injectable DMT at some point during follow-up, but not necessarily at every time point.  At baseline, 11% (48/426) of participants reported non-adherence (MPR <80%), 13% (50/386) at year one, and 14% (48/341) at year two. The denominators reflect the number of responders to the question of missed doses. Non-responders at each time-point totaled 9, 13 and 6, respectively (Table 3.3).     41   Disease-modifying therapy (route and frequency) Baseline N (%) Year One Year Two Glatiramer acetate (subcutaneous, daily)         Number (%) 8/131 (6.1) 11/130 (8.5) 7/126 (5.6) Interferon β-1a (intramuscular, weekly)  Number (%) 10/90 (11.1) 14/76 (18.4) 12/65 (18.5) Interferon β-1b (subcutaneous, every other day) Number (%) 12/62 (19.4) 8/51 (15.7) 8/43 (18.6) Interferon β-1a (subcutaneous, three times per week) Number (%) 18/143 (12.6) 17/129 (13.2) 21/107 (19.6) Table 3.3 Frequency of non-adherence, defined as medication possession ratio < 80% at baseline, year one, and year two by DMT product. Denominator represents the total number of participants who were on the specified DMT at each visit and who responded to the question “how many doses did you miss in the previous 30 days”. Non-responders to this question at each time-point totaled 9, 13 and 6, respectively. Numerator represents the number of people who were not adherent to that DMT. Percentages are shown in brackets.   During the entire study, 22.1% (107/485) of participants were estimated to be non-adherent at least once. Over half (51%; 255/485) of participants reported missing at least one dose of their DMT in the previous thirty days over the study period. Findings from the longitudinal, uni- and multivariable analysis are shown in Table 3.4. After adjusting for potential confounders, alcohol dependence, EDSS, disease duration, DMT product, and perceived functional cognitive difficulties were associated with non-adherence (Table 3.4). Those who met criteria for alcohol dependence had more than twice the odds of non-adherence compared to those who did not. Relative to participants with moderate disability (EDSS 3.0-5.5), those with mild disability (EDSS 0-2.5) were more likely to be non-adherent. Longer disease 42  duration (≥ 5 years vs < 5 years) was associated with increased odds of non-adherence. Glatiramer acetate users were more likely to be adherent relative to all three types of beta-interferon (Table 3.4).               43  Variables Univariate odds ratio (95% CI) Multivariable odds ratio (95%CI) Age (continuous) 0.98 (0.96 – 1.00) 0.98 (0.96 – 1.01) Sex  Female (reference) 1.00 1.00 Male 1.33 (0.84 – 2.10) 1.32 (0.84 – 2.08) Race  White (reference) 1.00 N/A Non-White 1.36 (0.54 – 3.40) Education High school or less 1.00 N/A Post-secondary or higher 1.05 (0.64 – 1.73) Site British Columbia (reference) 1.00 N/A Alberta 0.67 (0.31 – 1.47) Manitoba 0.64 (0.31 – 1.31) Nova Scotia 0.66 (0.41 – 1.07) EDSS EDSS mild (0-2.5)   1.76 (1.06 –2.92)  1.08 (1.06 –3.04) EDSS moderate (3.0-5.5) (reference) 1.00 1.00 EDSS severe (6.0+)  1.27 (0.64 – 2.55)  1.30 (0.63 – 2.66)  Disease Course Relapsing-remitting (reference) 1.00 N/A Secondary Progressive 0.83 (0.43– 1.62) Disease Duration   < 5 years (reference) 1.00 1.00 ≥ 5 years 1.78 (0.91 – 3.38) 2.23 (1.10 – 4.52) Disease-modifying therapy (route and frequency) Glatiramer acetate (subcutaneous, daily)         (reference) 1.00 1.00  Interferon β-1a (intramuscular, weekly)  3.21 (1.65 – 6.24) 2.83 (1.43 – 5.62) Interferon β-1b (subcutaneous, every other day) 3.49 (1.60 – 7.60) 3.23 (1.49 – 6.98) Interferon β-1a (subcutaneous, three times per week) 2.85 (1.51 – 5.37) 2.74 (1.48 – 5.12) 44  Number of physical comorbidities  0  (reference) 1.00 N/A 1  1.07 (0.72 – 1.60) ≥2 1.10 (0.71 – 1.70) Health Utilities Index (health-related quality of life) None to moderate disability (HUI score > 0.70) [reference]) 1.00 N/A Severe disability (HUI score ≤ 0.70) 0.79 (0.55 - 1.15) Health behaviours, mental health, and symptoms of MS No alcohol dependence (reference) 1.00 1.00 Alcohol dependence 2.28 (1.29 – 4.05) 2.14 (1.23 – 3.75) Non-smoker (reference) 1.00 N/A Current Smoker  1.21 (0.70 – 2.09) No depression (reference) 1.00 Depression 1.42 (0.93 – 2.17) No anxiety (reference) 1.00 Anxiety 1.10 (0.76 – 1.59) No fatigue (reference) 1.00 Fatigue  0.93 (0.63 – 1.38) No pain (reference) 1.00 Pain  1.00 (0.68 – 1.46) None to mild perceived functional cognitive difficulties (reference) 1.00 1.00 Moderate to severe perceived functional cognitive difficulties  1.32 (0.94 – 1.86) 1.55 (1.08 – 2.22) Table 3.4 Univariate and multivariable longitudinal analyses of clinical and demographic variables and their association with non-adherence (medication possession ratio < 80% in previous 30 days). Variables were measured at baseline, year one, and year two and included in the analysis as time-varying, with the exception of sex, race, education, and site which were collected at baseline only. Odds ratio of > 1 indicates a higher odds of non-adherence. Multivariable model was adjusted for age, sex, EDSS, disease duration, DMT product, alcohol dependence, and perceived functional cognitive difficulties. Bold indicates a significant association. N/A = Not applicable.   The odds of missing any doses were assessed in the supplementary analysis (Tables 3.5). When adjusting for potential confounders in the multivariable model (Table 3.5), similar factors 45  emerged as significant as in the primary adjusted analysis, including: alcohol dependence, cognition and DMT product. However, with this alternative method of assessing adherence, the odds of non-adherence were greater for the glatiramer acetate users (subcutaneous, daily) relative to the IFNβ-1a users (weekly intramuscular or subcutaneous three times per week). In addition, younger age, and the presence of multiple physical comorbidities (≥ 2 relative to none) were associated with higher odds of non-adherence.                 46  Variable Univariate odds ratio (95%CI) Multivariable odds ratio (95 % CI) Age (continuous) 0.99 (0.98 – 1.01) 0.98 (0.96 – 0.99) Sex  Female (reference) 1.00 1.00 Male 1.53 (1.06 – 2.21) 1.31 (0.89 – 1.92) Race  White (reference) 1.00 N/A Non-White 1.01(0.49 – 2.10) Education High school or less 1.00 N/A Post-secondary or higher 1.09 (0.76 – 1.54) Site British Columbia (reference) 1.00 N/A Alberta 0.78 (0.46 – 1.32) Manitoba 0.98 (0.58 – 1.67) Nova Scotia 0.73 (0.52 – 1.04) EDSS EDSS mild (0-2.5)  1.04 (0.78 – 1.39) N/A EDSS moderate (3.0-5.5) (reference) 1.00 EDSS severe (6.0+)   0.82 (0.53 – 1.29) Disease Course Relapsing-remitting (reference) 1.00 N/A Secondary Progressive 0.76 (0.49 – 1.17) Disease Duration   < 5 years (reference) 1.00 N/A ≥ 5 years 1.35 (0.92 – 1.98) Disease-modifying therapy (route and frequency) Glatiramer acetate (subcutaneous, daily) (reference) 1.00 1.00  Interferon β-1a (intramuscular, weekly)  0.23 (0.14 – 0.39) 0.22 (0.13 – 0.37) Interferon β-1b  (subcutaneous, every other day) 1.21 (0.75 – 1.95) 1.26 (0.78 – 2.03) Interferon β-1a (subcutaneous, three times per week) 0.59 (0.75 – 0.85) 0.60 (0.41 – 0.86) 47  Number of physical comorbidities  0  (reference) 1.00 1.00 1  1.14 (0.83 – 1.57) 1.33 (0.96 – 1.84) ≥ 2 1.39 (1.00 – 1.94) 1.54 (1.07 – 2.21) Health Utilities Index (health-related quality of life) None to moderate disability (HUI score > 0.70) [reference]) 1.00 N/A Severe disability (HUI score ≤ 0.70) 1.32 (1.02 – 1.72) Health behaviours, mental health, and symptoms of MS No alcohol dependence (reference) 1.00 1.00 Alcohol dependence 2.39 (1.51 – 3.78) 2.61 (1.57 – 4.34) Non-smoker (reference) 1.00 N/A Current Smoker  1.13 (0.77 – 1.66) No depression (reference) 1.00 Depression 1.09 (0.81 – 1.48) No anxiety (reference) 1.00 Anxiety 1.14 (0.88 – 1.46) No fatigue (reference) 1.00 Fatigue  1.23 (0.93 – 1.64) No pain (reference) 1.00 Pain 1.13 (0.88 – 1.44) None to mild perceived functional cognitive difficulties (reference) 1.00 1.00 Moderate to severe perceived functional cognitive difficulties  1.44 (1.13 – 1.84) 1.38 (1.05 – 1.80) Table 3.5 Univariate and multivariable longitudinal analyses of clinical and demographic variables and their association with non-adherence (missed ≥ 1 dose in the previous 30 days; ‘yes’ vs ‘no’) Variables were measured at baseline, year one, and year two and included in the analysis as time-varying, with the exception of sex, race, education, and site which were collected at baseline only. Odds ratio of > 1 indicates a higher odds of non-adherence. Multivariable model was adjusted for age, sex, DMT product, physical comorbidity count, alcohol dependence, and perceived functional cognitive difficulties. Bold indicates a significant association. N/A = Not applicable.    48  Finally, previous non-adherence (determined at baseline using the MPR) was associated with over 4 times the odds of future non-adherence at year one or two (OR: 4.42; 95%CI: 2.23 – 8.75), adjusting for sex, and baseline age, disease duration, EDSS, and alcohol dependence. 3.4 Discussion Over one in five participants in this large, cross-Canada cohort reported missing more than 20% of their doses and less than half were fully adherent in the last 30 days. Our study identified specific characteristics that influenced the likelihood of adherence, including some modifiable attributes. These characteristics may put an individual at risk of non-adherence or be useful markers of future non-adherence for the treating health professional. Of the patient-related characteristics explored, previous non-adherence, alcohol dependence, perceived functional cognitive difficulties, longer disease duration, and mild disability (EDSS), emerged as factors associated with non-adherence.  Rates of adherence were stable over time, and within the range of estimates from previous studies;57 however, variability in the definition of adherence makes direct comparisons between studies challenging.57 Authors that also employed a MPR cut-off of 80% reported estimates between 70 and 85%132,133, consistent with our overall adherence rate of 78%. Baseline non-adherence was the most significant predictor of future non-adherence in our study, with over 4 times the odds, suggesting that poor adherence is an enduring pattern of behaviour for some individuals. This may also serve as a useful early marker of poor future adherence. These results are in concordance with another MS study60 and in chronic diseases in general,131 and may provide an opportunity for early identification of individuals who may benefit from additional support.  49   Alcohol dependence was associated with twice the risk of non-adherence. We are aware of only one prior study examining this risk factor in MS, which found that alcohol was the strongest predictor of missed doses among people with MS living in Tasmania, Australia.60 These findings may be of particular concern given the broader negative effects of alcohol dependence,135 and given that individuals with MS have been reported as having high rates of alcohol dependence.29 Smoking has not been extensively studied in relation to adherence;  a single study that had examined the association in MS found no relationship, similar to our findings.60  There was no association between mental health (anxiety or depression, measured by the HADS) and missed drug doses. This is consistent with another study using similar methods.60 However, other studies that used different measures for both mental health (e.g. Beck’s Depression Inventory), and adherence have shown an association between depression and missed doses in MS.66,136 Neither of these studies considered the effect of alcohol use on adherence. In fact, one study excluded patients who abused alcohol,66 which could account for the conflicting results, as there is a complex and bidirectional relationship between alcohol dependence and depression.135 Depression can increase the likelihood of treatment discontinuation in MS.137 Thus, further study of this issue is warranted; particularly given the high prevalence of mental health disorders in those with MS.34  The risk of poor adherence increased with increasing disease duration, consistent with a previous study.59 However, the odds of non-adherence modestly decreased with age, as observed in other disease states.49 Interestingly, persons with mild disability (EDSS) were less likely to be adherent. These individuals may perceive themselves as having less serious disease and hence, 50  are less motivated to take their medication. They may benefit from an open dialogue as to the rationale for taking drug and expectations related to drug treatment. Overall health-related quality of life and pain, as measured by the HUI-3, were not associated with adherence. However, self-reported cognitive difficulties as reported on the HUI-3 were consistently an important determinant of non-adherence in our study and in some other chronic diseases.131 We found one other MS study in which the relationship between cognition and adherence was assessed.11 Using an extensive battery of cognitive tests in 55 MS individuals, who were primarily taking glatiramer acetate, associations with adherence were found.11  The HUI cognition scale, as used in our study, provides a more pragmatic option, being of lower burden to patients and feasible to implement in routine clinical practice. It specifically addresses forgetfulness, which might be the key element related to missed doses or poor adherence.59  In previous studies, persons with multiple comorbid diseases have expressed challenges with adhering to their medication regimens, especially when on multiple medications. ,138 Our findings suggested that MS patients with multiple physical comorbidities had increased odds of missing at least one dose of DMT;  however, no relationship was found between the total number of physical comorbidities as categorized by 0, 1, or ≥2 and adherence, when adherence was defined by the MPR.   Interestingly, the relationship between the different DMT products and adherence varied considerably, depending on the definition of adherence. Glatiramer acetate had the most frequent (daily) dosing schedule of the injectable DMTs. It was associated with better adherence relative to beta-interferon, when adherence was defined as ≥80% of expected doses taken, but not when using ‘any missed dose.’ Previous research has suggested that people do not adhere well to 51  glatiramer acetate relative to the beta-interferons.57 Together, these highlight the substantial impact that the definition and method for measuring adherence may have on findings. Clarity on this issue is important for future studies examining the clinical implications of poor adherence and to consider when comparing study findings.   Our study included a large multi-site sample, had a high retention, and recruited from at least two sites which served as the only source of MS care in their regions, all suggesting that this was a representative sample of clinic-attending MS patients. A potential limitation was the use of self-report, such that our non-adherence rates might be considered conservative. However this study used a specific recall period of 30 days, and collected the information via survey to reduce desirability bias as recommended elsewhere.139 Reporting bias is a possibility; those that are more likely to report adverse health behaviours, such as alcohol dependence, may also be more likely to report their missed doses. There is also a possibility of recall bias, in that some participants, especially those with cognitive difficulties, may not remember how many doses they missed in the last 30 days.  Since the implementation of this study, several oral and other parenterally-administered therapies have become available. A single study has reported better adherence among fingolimod users relative to users of the injectable therapies,140 but determinants of non-adherence to the oral therapies have not been established. It is conceivable the characteristics that were recognized in this study may also contribute to missed doses of oral therapies. Future studies should address this important question. Adherence to medication is central in the treatment of chronic disease to help derive the maximum possible clinical benefit. In the wider medical literature, poor adherence has been 52  linked to worsening morbidity, death, and increased health care costs.50  In MS, poor adherence has been associated with an increased risk of MS relapse, and MS-related hospitalization.61 In this study, nearly one-quarter of participants reported poor adherence to their DMT at least once during the study period, and over half were not fully adherent. Healthcare professionals should be aware of the greater potential for poor adherence among patients with low levels of disability, longer disease duration, and a history of poor adherence. Alcohol dependence and perceived functional cognitive difficulties were also important markers of non-adherence. Improving adherence is an ongoing process that involves patients, health care providers, and health systems. Enhancing communication between health professionals (including neurologists, general practitioners, nurses and pharmacists), patients, and their families; implementing educational interventions for those at risk; and addressing modifiable risk factors such as alcohol dependence could effectively improve health outcomes in individuals with MS and ultimately reduce costs to health systems.50        53  Chapter 4: Disease-modifying therapies and adherence in multiple sclerosis: comparing patient self-report with pharmacy records3 4.1 Background and objectives The study outlined in Chapter 3 measured medication adherence based on self-reported missed doses. This is the most commonly employed method in the clinical setting, but its precision and validity have been questioned.50 Health administrative data, or pharmacy records, are being validated and used for research purposes with increasing frequency;70,141,142 however, little is known about how these two sources compare.  Objectives: To compare self-reported DMT use and adherence against pharmacy records. Specifically, we aimed to assess the level of agreement between a practical, self-report questionnaire and both short and longer-term pharmacy records of prescriptions filled with respect to: a) DMT use (any DMT being currently used); b) specific DMT product (identified by brand name); and c) adherence to a DMT.                                                  3 3 A version of this chapter has been accepted for publication as McKay KA, Evans C, Fisk JD, Patten SB, Fiest K., Marrie RA, Tremlett H. Disease-modifying therapies and adherence in multiple sclerosis: comparing patient self-report with pharmacy records. Neuroepidemiology 2017.   54  4.2 Methods 4.2.1 Study population Consecutive MS patients were recruited from the University of British Columbia MS Clinic, Vancouver, Canada, between 2010 and 2011 and followed up until 2013. Details of the inclusion criteria and data collection are described in Chapter 1, Subsection 1.6.2.  4.2.2 Self-reported information A questionnaire captured the current DMT used by asking each participant, “If you are currently taking a disease-modifying therapy please mark which therapy you are taking with a√”, followed by the choices: Avonex, Betaseron, Rebif, Copaxone, or Tysabri. These represented the only available (approved) DMTs for MS in Canada when the questionnaire was administered. To measure adherence to the injectable DMTs (ie, a beta-interferon [Avonex®, Betaseron®, Rebif®] or glatiramer acetate [Copaxone®]), participants were then asked: “If you are currently taking a disease-modifying therapy please indicate how many injections you missed in the last 30 days.”  The self-report questionnaire was completed at the time of the MS clinic visit. As our focus was on adherence to the first-line, injectable DMTs, participants who reported use of the second-line and intravenously administered natalizumab [Tysabri®] at baseline or who switched to this therapy at a later point in the study period were excluded from the analyses.  4.2.3 Health administrative (pharmacy records) information Clinical and questionnaire data were linked to the province-wide prescription database (PharmaNet)96 via each participant’s unique personal health number (PHN). PharmaNet provides 55  data on all prescriptions filled (dispensed) in the out-patient setting for the province of British Columbia. Information included the drug dispensed, date of dispensation, quantity dispensed, and days supply provided. Availability of an injectable DMT in the 30 days prior to a clinic visit was determined by combining the date dispensed and days supply. The DIN at this most recent dispensation was used to identify the DMT product. Data were linked to the BC Ministry of Health's Registration and Premium Billing Files97 to ensure that all participants resided in the province. Pregnancy-related physician and hospital encounters were obtained by linkage with the Medical Service Plan (MSP)93 and the Discharge Abstract Database (DAD);95 these data were used to ensure that any women who became pregnant during the study period were not misclassified as being DMT ‘non-adherent.’   4.2.4 Definitions of adherence Adherence was estimated for self-report and pharmacy records using the medication possession ratio (MPR). A MPR of ≥80% was considered adherent, and <80% as non-adherent.  This definition has been frequently used in assessing MS DMT adherence,57 with non-adherence (MPR<80%) associated with higher relapse rates and increased medical costs.61,70 Based on the 30 day self-report, the MPR was calculated as the number of doses taken divided by the number of expected doses, per the product monograph. This practical, short-term estimate of adherence was compared to longer-term adherence, assessed using pharmacy records over two distinct and clinically meaningful time periods: one year pre- and post-clinic visit. From pharmacy records, the MPR was calculated using the total number of days supplied over the one-year period divided by 365 days. Only individuals who initiated drug at least one year prior to the clinic visit were included in the pre-year analysis. For the post-year analysis, only individuals on a DMT at 56  the time of the clinic visit and with ≥1 dispensation over the subsequent year were included. This method allows comparisons between studies61,62,143 and makes best use of the prescription refill process (which can be on an-up-to 90 days cycle). Participants who switched to a second-line therapy (natalizumab or fingolimod) were excluded from the analyses. As a supplementary analysis, adherence over 6 months (183 days) pre- and post-clinic visit was estimated from pharmacy records and compared with the 30-day self-report using the same approach.  To determine if participant characteristics influenced agreement between sources, we stratified the analyses by demographic and clinical characteristics. Age, sex, disease course, and disability status were collected from the MS Clinic charts. Participants self-reported their highest education level achieved and completed a series of questionnaires including the Hospital Anxiety and Depression Scale (HADS) and the Health Utilities Index Mark III version (HUI-3). Validated cut-off scores of ⩾8 on the HADS scale were used to define depression and anxiety. Perceived functional cognitive impairment was measured using the individual attribute scale for cognition on the HUI-3, categorized as none-to-mild versus moderate-to-severe cognitive difficulties (⩽0.70 vs >0.70). 4.2.5 Statistical analysis Categorical variables were described as frequency (percent), and continuous variables as mean (standard deviation) or median (interquartile range), as appropriate.   Self-report and pharmacy records were compared and the sensitivity, specificity, positive and negative predictive value (PPV, NPV) of each component (i.e., DMT use, product and adherence) were estimated with pharmacy data as the reference standard.  We also calculated 57  kappa statistics to measure agreement between the two data sources, where neither was considered the reference standard. Kappas of 0–0.20 indicated slight, 0.21–0.40 fair, 0.41–0.60  moderate, 0.61–0.80 substantial, and 0.81–<1.0 indicated almost perfect agreement.144 We calculated Cohen’s kappa as well as prevalence and bias adjusted kappa (PABAK) scores.145  All linkages were facilitated by Population Data BC. BC Ministry of Health and Data Stewardship Committee approved access to administrative health data. A requirement of data access necessitated suppression of cell sizes <5 to protect patient confidentiality. 4.3 Results Of 340 participants in British Columbia who agreed to participate and were linked to their health administrative data, 326 were eligible for the present study. The majority of the participants were female (74.5% [243/326]) and 73.6% (240/326) had a relapsing-remitting course. The mean age at assessment was 48.7 years and the mean disease duration was 15 years (Table 4.1). A total of 135 participants reported use of an injectable DMT at their clinic visit. Of these, 128/135 (94.8%) had drug available according to pharmacy records in the 30 days before their clinic visit. Of those who reported no DMT use, 99.5% (190/191) had no supply of an injectable DMT available during this period according to the pharmacy records.  There was near perfect agreement between self-report and pharmacy records in determining current use of injectable DMTs (κ = 0.95; 95%CI:0.91-0.98; Table 4.2).     58  Characteristics n = 326 Sex, N (%)  Female     Male  243 (74.5) 83 (25.5) Race, N (%) (86 missing)  White  Non-White  208 (86.6) 32 (13.4) Age, mean (SD) Age Range (years) 48.7 (11.9) 19-80 Age of symptom onset, mean (SD) 33.4 (10.3) Disease duration, mean (SD) 15.3 (9.9) EDSS, median (IQR) 2.5 (2.0 – 4.0) Clinical Course, N (%) RRMS  SPMS PPMS  243 (74.5) 64(19.6) 19 (5.8) Education (83 missing) High School or Less Post-Secondary or higher Other  72 (29.7) 165 (67.8) 6 (2.5) Self-reported DMT Product, N (%) None Interferon-β-1a (intramuscular) Interferon- β-1b (subcutaneous) Interferon- β-1a (subcutaneous) Glatiramer acetate (subcutaneous)  191 (58.5) 19 (5.8) 25 (7.7) 60 (18.4) 31 (9.5) Table 4.1 Demographic and clinical characteristics of the cohort.  Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI) Kappa (95% CI) 0.99 (0.96 – 1.00) 0.96 (0.93 – 0.99) 0.95 (0.90 – 0.98) 0.99 (0.97 – 1.00) 0.95 (0.91 – 0.98) Table 4.2 Agreement between self-reported use of a DMT compared to pharmacy records in the 30 days prior to a MS clinic visit.  59  We found perfect agreement between pharmacy records and self-report in identifying the DMT product among the 128 participants who reported use of an injectable DMT and who had drug available in the 30 days prior according to their pharmacy records (κ = 1.00; 95%CI: 1.00-1.00).  From the self-report data, 13.2% (17/128) of participants were non-adherent (MPR <80%) in the previous 30 days. According to pharmacy records, 113 participants were receiving DMT one year prior to the clinic visit; 34 (30.0%) of these individuals were non-adherent. In the year after the clinic visit, 43.1% (53/123) of the included participants were non-adherent. The proportion of non-adherent participants in the 6-month period before the clinic visit was 26.1% (31/119), and 37.9% (47/124) in the 6 months after the clinic visit.  Agreement between the MPR based on self-reported missed doses in 30 days and pharmacy records in the one year (κ = 0.41; 95% CI: 0.22-0.59) or 6 months (κ=0.46; 95% CI: 0.27-0.64) prior to the clinic visit was moderate. Fair agreement was found between the one year (κ = 0.22; 95%CI: 0.09-0.36) and 6 months (κ = 0.23; 95%CI: 0.08-0.38) following the clinic visit, and the MPR based on self-reported missed doses (Table 4.3). The prevalence and bias-adjusted kappa (PABAK) scores did not change interpretation of findings; while the actual scores increased marginally, the level or category of agreement remained the same. In the year prior and post the clinic visit, agreement was moderate (PABAK =0.57) and fair (PABAK =0.30), respectively. Sensitivity estimates were modest, ranging from 0.25 – 0.42, while the specificity was high (0.95-0.97).      60   Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI) Kappa (95% CI) One year prior to clinic visit (n=113)a 0.38 (0.22 – 0.56) 0.96 (0.89 – 0.99) 0.81 (0.54 – 0.96) 0.78 (0.69 – 0.86) 0.41 (0.22 – 0.59) One-year post- clinic visit (n=123)b  0.25 (0.14 – 0.38) 0.96 (0.88 – 0.99) 0.81 (0.54 – 0.96) 0.63 (0.53 – 0.72) 0.22 (0.09 – 0.36) Six Months prior to clinic visit (n=119)a 0.42 (0.25 – 0.61) 0.97 (0.90 – 0.99) 0.81 (0.54 – 0.96) 0.83 (0.74 – 0.89) 0.46 (0.27 – 0.64) Six Months Post- clinic visit   (n=124) b 0.26 (0.14– 0.40) 0.95 (0.87 – 0.99) 0.75 (0.48 – 0.93) 0.68 (0.58 – 0.76) 0.23 (0.08 – 0.38) Table 4.3 Agreement between self-reported missed doses in 30 days and pharmacy records across four time periods in determining rates of non-adherence. Non-adherence was defined as a medication possession ratio of <80%. a The total number of participants in the pre-clinic visit periods reflects the number of participants who had DMT available at the beginning of that period. b The total number of participants in the post-clinic visit periods reflect those that had at least 1 prescription dispensed following their clinic visit and who did not switch to a second-line therapy in that period.  None of the clinical and demographic characteristics that were examined influenced the agreement between self-report and administrative definitions of adherence (Table 4.4). Kappa scores were notably higher among men (vs women), relapsing-remitting (vs secondary progressive), and non-depressed (vs depressed) individuals, but none of these differences met statistical significance.     61  Characteristic (N) Kappa (κ) 95% Confidence Interval Sex Male (28) Female (85)  0.59 0.31  0.30 – 0.89 0.09 – 0.53 Age <45 years (63) ≥45 years (50)  0.47 0.31  0.24 – 0.69 0.01 – 0.60 Education (21 missing) ≤ High School (25) > High School (67)  0.25 0.37  -0.26 – 0.76 0.14 – 0.60 Disease Course  RRMS (102) SPMS (11)  0.44 0.12  0.25 – 0.63 -0.43 – 0.67 Disability Status (10 missing) Mild (EDSS 0-2.5) (69) Moderate – Severe (EDSS 3.0+) (34)  0.44 0.27  0.22 – 0.67 -0.10 – 0.64 Depression – HADS-D score ≥ 8  Yes (16) No (97)  0.21 0.44  -0.26 – 0.68 0.24 – 0.64 Anxiety – HADS-A score ≥ 8 Yes (44) No (69)  0.43 0.44  0.08 – 0.60 0.18 – 0.70 Perceived functional cognitive impairment – HUI-3 cognition score ≤0.70 Yes (50) No (63)   0.34 0.47   0.07 – 0.60 0.22 – 0.72 Table 4.4 Stratified analyses showing the effect of demographic and clinical characteristics on the level of agreement (κ) between self-reported missed doses in 30 days and pharmacy records in the year prior to clinic visit.   4.4 Discussion We evaluated agreement between self-report and pharmacy records in estimating current DMT use, product, and adherence in MS. We found near-perfect to perfect agreement for current DMT use and product, and fair to moderate agreement for DMT adherence between sources. These findings have practical implications, potentially informing clinical care and the design of future studies. Regardless of source, adherence to the injectable DMTs for MS was suboptimal; 62  more than 1 in 10 (13%) participants were non-adherent based on the 30-days self-report and over 4 in 10 (43%) were non-adherent over the subsequent year based on pharmacy records.  To our knowledge, this is the first study to compare DMT use, product, and adherence between self-report and pharmacy records in MS. Previous research on antidepressant146 and general medication use147 found substantial agreement (kappa of 0.69 and 0.64, respectively) between self-report and claims records in terms of drug use. We found near-perfect agreement between sources, suggesting that self-report truly reflects the actual DMT received. It also suggests that self-report may be a pragmatic means of collating this information when other sources are not readily accessible.  When short-term self-report was compared to longer-term pharmacy records in the assessment of adherence, the level of agreement was more modest, ranging from moderate to fair. Eight of ten original studies outlined in a review on concordance between adherence sources (MS not included) also reported lower rates of adherence measured by objective methods relative to self-report.148 We found no significant differences in agreement between subgroups of our population, suggesting that our findings were not unduly influenced by a particular characteristic. Thus, this self-report question seems appropriate for most persons with MS. Other groups have also failed to find clinical or demographic characteristics associated with inaccurate reporting of adherence. 149,150 Both patient self-report and pharmacy records suggested suboptimal rates of adherence to the injectable DMTs for MS. Two studies employed similar definitions of self-reported adherence and their estimates of non-adherence ranged from 12 to 15% for the injectable DMTs, comparable to our finding of 13% .132,134  Two additional studies were found using pharmacy 63  records. Their estimates were similar to ours, ranging from 40% over one year to 32% over two years in insured USA-based MS populations.61,151   We observed high specificity between self-reported non-adherence and pharmacy records, indicating that patients who reported non-adherence over 30 days were very likely to also exhibit non-adherence over the longer time periods. Specificity refers to the proportion of negatives correctly identified as such. This implies that self-reported missed doses recalled over a recent and limited period could be a useful surrogate in estimating non-adherence over a longer, more clinically meaningful period.  The sensitivity of self-reported missed doses was low, meaning that individuals who reported high adherence (i.e., few missed doses) may still be ‘at risk’ for non-adherence. Sensitivity measures the proportion of positives correctly identified as such. None of the demographic or clinical characteristics which we explored could explain this disparity. Non-adherence should be considered as a possible explanation for suboptimal clinical outcomes in these individuals.  This low sensitivity of self-reported non-adherence may reflect several factors. First, assuming both sources are ‘true,’ it is possible that a high rate of adherence as found by self-report in the 30 days before a MS clinic visit coexists with a poorer adherence rate over the full-year (from pharmacy records). This ‘white-coat compliance’ reflects a recognized improvement in health behaviours exhibited immediately before a physician visit.152 Second, the different time periods explored in self-report and pharmacy records (30 days versus 1 year) likely contributed to the fair to moderate agreement between the two sources. Rates of non-adherence in MS increase with greater follow-up time.69 Last, we assessed adherence using a fixed medication possession ratio, which takes into account not only compliance, but also persistence (continuing 64  treatment for the prescribed duration). Combining these two principles can be important because if a person stops taking their medication they can no longer derive any clinical benefit from it. Self-reported missed doses only captures compliance, as persons who were not persistent (ie not on drug) were not included in the analysis. These different approaches may contribute to the low estimated sensitivity in the year following the clinic visit in which some patients stopped having their DMT prescriptions dispensed entirely.   Direct methods of measuring adherence, such as a third party observing the administration of a medication, offer increased accuracy, but are burdensome to patients and clinicians and impractical to implement in a clinical setting.50 Pharmacy records and self-report represent pragmatic options, but they cannot directly assess if a person administered the medication, as prescribed. Pharmacy records do not capture the reasons for starting or stopping a prescription. True non-adherence remains a challenge to measure in the absence of a readily available biomarker which captures drug levels. Self-report is susceptible to social desirability bias, in which persons under-report ‘undesirable’ behaviours,148 such as missed doses. The use of a questionnaire provided by a research coordinator, as opposed to interview by the prescriber, should minimize this effect.139 Recall bias may also affect self-report estimates, although limiting the time period to 30 days should minimize this issue.139 Finally, we do not know whether findings are applicable to the newer oral or infusion-administered DMTs for MS. Future studies could explore the utility of third party observations of adherence, which may be less susceptible to bias.  Our study has practical implications relevant to the clinic setting. The high PPV indicated that participants reporting missed doses in the last 30 days were at high risk for future long-term non-adherence. This implies that asking an individual about recent missed doses may help 65  identify those patients who could benefit most from educational interventions, communication regarding medication tolerance, or medication changes in order to prevent future chronic non-adherence.50   In summary, we found high agreement between pharmacy records and self-report in determining current DMT use and product, and fair to moderate agreement in estimating prior and future DMT adherence. In the absence of a gold standard for measuring adherence, our findings may help guide study design, facilitating selection of the most appropriate and pragmatic method which best suits the research question and resources.         66  Chapter 5: Association between psychiatric comorbidity and disability progression of multiple sclerosis 5.1 Background and objective As findings from the previous Chapters have shown, psychiatric symptoms are very common in MS. Over two years of follow-up more than half of our cohort met criteria for anxiety, and 36% met criteria for depression. The effect of psychiatric comorbidity on disability progression is not understood, despite the fact that ambulation and mental health have both been cited as important utilities to persons with MS.153,154    Objective: To investigate the relationship between psychiatric comorbidity and neurologic disability, and potential sex differences in these relationships, in a longitudinal multi-site MS clinic study using linked health administrative data in British Columbia, Canada.   5.2 Methods 5.2.1 Design and setting This long-term cohort study employed patient information collated prospectively from the BCMS database and population-based health administrative databases, details of which are outlined in Chapter 1.6. These patients were linked via their unique personal health number to their individual-level information from the BC Registration and Premium Billing Files (residency in the province);97 BC Vital Statistics Agency (mortality data);98 Census Geodata (neighbourhood-level socioeconomic status - median value for each individual was calculated using all available values over the study);99Medical Service Plan payment information file 67  (physician visits);93 Discharge Abstract Database (hospital visits); 95 and PharmaNet (prescriptions dispensed).96 Data were available from all sources until December 31st, 2008.  5.2.2 Study population The study cohort included adults (≥18 years at MS onset) with clinically definite MS, as determined by the patient’s neurologist according to the prevailing diagnostic criteria of the time.85–87 Date of MS onset and registration with a BCMS clinic must have occurred between April 1st, 1993 (allowing two full years of both hospital and physician health administrative data) and December 31st, 2004 (the last year in which the participating clinics were the only MS clinics in BC, and to allow sufficient time to elapse before the study end of December 31st 2008). The cohort entry (baseline) was the date of MS symptom onset. MS symptom onset is collected based on an interview with the patient by their neurologist to determine the first date of clinical signs of MS. Residency in the province for each of the two years (≥270 days in each year) prior to baseline was required to determine psychiatric comorbidity status. Patients were followed until their last recorded EDSS prior to study end, emigration (>90 consecutive days not registered in the province), or death, whichever came first. Individuals with less than 2 EDSS measurements during the study period were excluded.  5.2.3 Defining psychiatric comorbidities Psychiatric comorbidities were identified using validated algorithms combining hospital and physician diagnostic codes (coded according to the International Classificiation of Disease [ICD] 9/10), and prescription information (coding according to the World Health Organization Anatomic Therapeutic Chemical (WHO-ATC)155 classification system).156 First, an omnibus 68  approach was employed in which all ‘mood or anxiety disorders’ were considered, including depression, anxiety, and bipolar disorder. To meet the omnibus definition a person had to have ≥ 1 hospitalization or ≥ 5 physician visits or (≥ 1 physician visit and ≥4 filled prescriptions) for a mood or anxiety disorder within a two-year period (see Appendix A; Table A.3 for specific ICD and ATC codes). This allowed us to capture comorbidity present in the two years prior to the onset of MS, as well as developing thereafter. The individual effects of depression, anxiety, and bipolar disorder were also explored as for the omnibus definition.   For all definitions, the first relevant ICD code (or prescription filled) for a psychiatric comorbidity was considered the date of diagnosis. This date was used to determine whether the comorbidity was present or absent. Due to the recurrent or chronic nature of these disorders, once the definition was reached, it was considered present for the remainder of follow-up in all models. 5.2.4 Measuring disability worsening To maximize use of available EDSS assessments, we included all scores for each patient, treated as continuous values to determine the association between the presence of psychiatric comorbidity and disability status.  EDSS scores were collected prospectively and measured by the neurologist at a patient’s clinic visit. Complementary approaches to assessing disability included: the change in EDSS score between visits and use of the Multiple Sclerosis Severity Score (MSSS).20  The MSSS is a continuous variable created using an algorithm that relates EDSS scores to the distribution of disability in patients with comparable disease durations (possible range: 1-10).20   69  5.2.5 Statistical analysis The association between psychiatric comorbidity status and disability was examined using multivariable linear regression techniques, fitted using an identity link with generalized estimating equations (GEE) with an exchangeable working correlation structure. Findings were reported as beta-coefficients with 95% confidence intervals (CIs). The GEE model generates risk estimates that are population averages of within- and between-subject effects, but are generally dominated by between-subject effects. To better estimate within-person effects over time and the effect of incident comorbidity on disability, we modeled change in EDSS relative to change in status of the covariates, using similar GEE methodology.102  Covariates were included in the model based on their established or potential association with disability progression or because they reached a threshold level of statistical significance (p<0.1); those considered were: age at MS symptom onset, sex, year of MS onset, disease course [relapsing-onset versus primary progressive], socioeconomic status, physical comorbidities, and disease-modifying therapy [DMT] use. Time-varying covariates were estimated at the time of each EDSS assessment date, and included: disease duration, current DMT use (yes versus no), and a count of physical comorbidities. Current DMT exposure was estimated using the dispensation date and days supply from the prescription database (PharmaNet). The physical comorbidities including in the count included diabetes, epilepsy, heart disease, hyperlipidemia, hypertension, and chronic lung disease. These were chosen based on their potential impact on disability progression157 combined with availability of a validated algorithm to identify them using administrative data.141,158,159 Individuals were categorized as having 0, 1, 2, or 3+ comorbidities based on the presence of these comorbidities at the time of each EDSS assessment. 70  The remaining covariates, sex, MS disease course, and median SES, were considered constant throughout the study period. Analyses were stratified by sex.  All analyses were performed using SAS Statistical Software Package 9.4 (SAS Institute Inc., Cary, NC, USA). All linkages were facilitated by Population Data BC. BC Ministry of Health and Data Stewardship Committee approved access to administrative health data. Cell sizes <5 were suppressed to protect the confidentiality of the participants. 5.3 Results There were 1250 incident cases of MS identified between 1993 and 2004 who met criteria for inclusion (Figure 5.1). The majority were women (75.0%), with a relapsing-onset disease course (94.0%), and an average age of MS symptom onset of 36.9 years (Table 5.1). The median first EDSS score was 2.0 (interquartile range: 1.5-3.0), and the average follow-up time, from two years pre-MS onset until the last available EDSS, was 9.1 years (range 2.1 – 17.0).   71   Figure 5.1 Selection of incidence-onset MS patients residing in British Columbia, Canada        1973 individuals registered with a BCMS clinic with MS onset between April 1, 1993 and Dec 31, 2004  245 excluded 151 became resident in BC after MS symptom onset 94 did not have ≥ 270 days of registration in the province in each of the 2 years prior to their onset  1728 individuals resident in BC between April 1, 1993 and December 31, 2004  445 excluded as they had <2 EDSS scores available             1695 individuals remained in the study cohort  33 excluded as aged < 18 years at their MS onset       1250 individuals included in the analysis  72  Characteristics Total cohort n= 1250 Sex, N (%)   Women 938 (75.0)  Men 312 (25.0) Age at MS symptom onset (categorized), N (%)  18 – 29 346 (27.7) 30-39 444 (35.5) 40-49 341 (27.3) 50+ 119 (9.5) Age at MS symptom onset, mean (SD) 36.9 (9.7) Disease Course  Relapsing-onset  Primary Progressive  1176 (94.0) 74 (6.0) Disease duration at first EDSS in years, mean (SD) 2.6 (2.2) Follow-up time in years, mean (SD) 9.1 (3.3) First EDSS score (median, interquartile range) 2.0 (1.5 – 3.0) Ever exposed to disease-modifying therapy during follow-up  Yes No   649 (51.9) 601 (48.1) Median socioeconomic status over follow-up  1 (Low) 2 3 4 5 (High)   156 (12.5) 263 (21.0) 299 (24.0) 313 (25.0) 218 (17.5) Table 5.1 Clinical and demographic characteristics of the study cohort.  Over the follow-up period, 597 (47.8%) individuals met the omnibus case definition for a mood or anxiety disorder. When stratified by sex, 51.8% (486/938) of women and 35.6% (111/312) of men met criteria. While many individuals had a prevalent mood or anxiety disorder, meeting the case definition prior to their MS onset (276/597; 46.2%), the majority were incident, occurring after MS onset (321/597; 53.8%). Depression was by far the most common disorder, with 35.8% of the cohort ever affected, followed by anxiety (23.4%) and bipolar disorder (4.9%).  Significant 73  differences in sex were noted in depression (39.8% of women vs 23.7% of men, p<0.0001) and anxiety (26.8% of women vs 13.5% of men, p <0.0001), but not bipolar disorder (5.1% of women vs 4.2% of men, p = 0.500). Using the omnibus definition, the prior presence of a mood or anxiety disorder was associated with a higher EDSS score (β-coefficient = 0.45, p<0.0001, adjusted for disease duration). This remained statistically significant after adjustment for age at MS onset, sex, disease course and socioeconomic status (β-coefficient = 0.47, p<0.0001). Including a count of physical comorbidities and DMT exposure in the model did not affect findings (β-coefficient = 0.45, p<0.0001) (Table 2). In this full model, the average EDSS was 0.40 points higher among men compared to women (p=0.0002). Following stratification by sex, the relationship between mood and anxiety disorders and EDSS remained statistically significant among women (β-coefficient 0.52, p<0.0001), but not men (β-coefficient 0.22, p=0.219), Table 5.2.   Examining mood or anxiety disorders independently suggested that the association was driven predominantly by the effect of depression, (β-coefficient=0.44, p <0.0001). Following adjustment for confounders, there was no significant relationship between anxiety (β-coefficient=0.19, p=0.091, nor bipolar disorder (β-coefficient= 0.22, p=0.358) and disability score (Table 5.3).  The complementary approaches showed similar results. Using the omnibus definition of a mood or anxiety disorder and EDSS modeled as the change in score between assessments or the MSSS as the disability outcome did not change the direction of findings. However, the former did not reach statistical significance (β-coefficient= 0.19, p=0.124, Table 5.4), while the latter did (e.g., 74  the presence of a mood or anxiety disorder was associated with a higher MSSS, by 0.82 points, p<0.0001, Table 5.5).                    75   Crude* Adjusted** Adjusted*** Variable β SE p-value β SE p-value β SE p-value Intercept 1.73 0.07 <0.0001 0.05 0.22 0.8373 0.06 0.22 0.7873 No mood or anxiety disorder (reference category)          Mood or anxiety disorder 0.45 0.09 <0.0001 0.47 0.09 <0.0001 0.45 0.09 <0.0001 Women          Intercept 1.61 0.08 <0.0001 0.27 0.25 0.2939 0.29 0.25 0.2490 No mood or anxiety disorder (reference category)          Mood or anxiety disorder 0.54 0.10 <0.0001 0.52 0.10 <0.0001 0.52 0.10 <0.0001 Men          Intercept 2.03 0.11 <0.0001 -0.11 0.42 0.7835 -0.15 0.43 0.7311 No mood or anxiety disorder (reference category)           Mood or anxiety disorder  0.26 0.18 0.1442 0.23 0.17 0.1902 0.22 0.17 0.2187 Table 5.2 Association between all mood or anxiety disorders combined and neurologic disability, as measured by the EDSS. *Disease duration (time) included in model; **Adjusted for disease duration, age at MS onset, sex, socioeconomic status, disease course; ***Adjusted for disease duration, age at MS onset, sex, socioeconomic status, disease course, disease-modifying therapy use, and physical comorbidity count.      76    Crude* Adjusted** Adjusted*** Variable β SE p-value β SE p-value β SE p-value Depression           Intercept 1.80 0.06 <0.0001 0.08 0.22 0.7044 0.10 0.22 0.6643 No depression (reference category)          Depression 0.41 0.09 <0.0001 0.45 0.09 <0.0001 0.44 0.09 <0.0001 Anxiety          Intercept 1.86 0.06 <0.0001 0.15 0.22 0.4802 0.17 0.22 0.4507 Anxiety disorder (reference category)          Anxiety disorder 0.14 0.11 0.2281 0.20 0.11 0.0764 0.19 0.11 0.0909 Bipolar Disorder          Intercept 1.88 0.06 <0.0001 0.19 0.22 0.3870 0.20 0.23 0.3620 No bipolar disorder (reference category)          Bipolar disorder  0.33 0.25 0.1875 0.26 0.24 0.2893 0.22 0.24 0.3577 Table 5.3 Association between depression, anxiety, and bipolar disorder and neurologic disability, as measured by the EDSS. *Disease duration (time) included in model; **Adjusted for disease duration, age at MS onset, sex, socioeconomic status, disease course; ***Adjusted for disease duration, age at MS onset, sex, socioeconomic status, disease course, disease-modifying therapy use, and physical comorbidity count.     77    Crude* Adjusted** Adjusted*** Variable Β SE p-value β SE p-value β SE p-value Intercept 0.05 0.03 0.0680 0.05 0.06 0.4068 0.04 0.06 0.5189 No mood or anxiety disorder (ref)          Mood or anxiety disorder 0.21 0.12 0.1019 0.19 0.12 0.1179 0.19 0.12 0.1237 Women          Intercept 0.06 0.04 0.1379 0.12 0.08 0.1339 0.11 0.08 0.1417 No mood or anxiety disorder (ref)          Mood or anxiety disorder 0.26 0.15 0.0822 0.25 0.15 0.0969 0.25 0.15 0.0971 Men          Intercept 0.05 0.04 0.2339 0.07 0.12 0.5214 0.03 0.11 0.7923 No mood or anxiety disorder (ref)          Mood or anxiety disorder  0.07 0.21 0.7224 0.07 0.21 0.7255 0.05 0.21 0.8133 Table 5.4 Association between change in mood or anxiety disorders and neurologic disability, as measured by the EDSS. *Time difference between EDSS assessments included in model; **Adjusted for time difference between EDSS assessment, baseline EDSS score, age at MS onset, sex, socioeconomic status, disease course; *** Adjusted for time difference between EDSS assessments, baseline EDSS score, age at MS onset, sex, socioeconomic status, disease course, change in disease-modifying therapy use between EDSS assessments, change in physical comorbidity count.    78   Crude* Adjusted** Adjusted*** Variable β SE p-value β SE p-value β SE p-value Intercept 4.37 0.12 <0.0001 1.77 0.34 <0.0001 1.79 0.34 <0.0001 No mood or anxiety disorder (ref)          Mood or anxiety disorder 0.81 0.15 <0.0001 0.85 0.15 <0.0001 0.82 0.15 <0.0001 Women          Intercept 4.09 0.14 <0.0001 2.00 0.39 <0.0001 2.05 0.40 <0.0001 No mood or anxiety disorder (ref)          Mood or anxiety disorder 0.99 0.18 <0.0001 0.97 0.17 <0.0001 0.95 0.17 <0.0001 Men          Intercept 5.09 0.20 <0.0001 1.89 0.63 0.0025 1.80 0.63 0.0042 No mood or anxiety disorder (ref)          Mood or anxiety disorder  0.31 0.24 0.2052 0.34 0.24 0.1573 0.31 0.24 0.1932 Table 5.5 Association between mood or anxiety disorders and neurologic disability, as measured by the MSSS.  * Disease duration (time) included in model; **Adjusted for disease duration, age at MS onset, sex, socioeconomic status, disease course; ***Adjusted for disease duration, age at MS onset, sex, socioeconomic status, disease course, disease-modifying therapy use, and physical comorbidity count.   79  5.4 Discussion Psychiatric comorbidities were common in our new (incident) onset MS cohort. Presence of these comorbid conditions significantly increased the subsequent disability over a nine year period. Overall, nearly half of the 1250 incident MS cases were found to have a mood or anxiety disorder. The association between these disorders and disability remained even after accounting for sex, age at MS onset, disease course, socioeconomic status, disease-modifying therapy use, and physical comorbidities. The effect was significant among women, but not men (though the direction of the relationship was the same for both). When exploring the individual effect of each mood or anxiety disorder, depression arose as the only significant moderator of disability. Both anxiety and bipolar disorder were associated with a higher EDSS score, but neither met statistical significance.  Despite the high proportion of individuals who have comorbid psychiatric disorders and MS, the literature exploring the longer-term effects of these conditions is limited. As such, the consequences of having a comorbid psychiatric disorder are not well-understood. The vast majority of previous work has focused on depression in MS, and has shown that it is associated with a worse quality of life,72,73 impaired cognition,160,161 and adverse lifestyle factors, including smoking, and obesity.135,162  Depression is associated with increased progression and risk of mortality in other chronic diseases, including cancer35 and HIV.36 A similar relationship has been suggested in MS mortality risk;163 however, the current knowledge regarding mental health and MS disability has largely been derived from cross-sectional studies. We found nine studies in total exploring this relationship; six reported a positive correlation between mental health and disability status, 80  75,161,164–167 while three found no association.38,39,168 Five of the positive findings were cross-sectional, all relied on self-reported psychiatric conditions or psychometric scales.161,164–167 Two cross-sectional studies reported no relationship.38,39 Conflicting results were found among the two longitudinal studies, one of which used a patient reported disability outcome,75 the other the onset of SPMS.168 The study of 269 RRMS patients who completed the Hospital and Anxiety Depression Scale and the Multiple Sclerosis Walking Scale-12 at three time points over two years reported a significant reciprocal relationship between depressive symptoms and walking impairment.75 Baseline score on the Center for Epidemiological Studies Depression Scale did not predict risk of SPMS among 149 patients who were followed for ten years.168 Our study was the first to attempt to elucidate a temporal relationship between psychiatric disorders and MS. While our findings were highly suggestive that the presence of a psychiatric disorder contributed to a subsequent increase in EDSS, we cannot fully rule out the possibility of a bidirectional relationship. It remains possible that for some individuals, a psychiatric condition may either develop in response to worsening disability or be more readily recognized (and diagnosed) in these individuals. Further, when we assessed the change in psychiatric comorbidity status and the subsequent change in EDSS, findings did not reach significance, suggesting that the results were largely driven by between-person effects, as opposed to within-person effects.  The high prevalence of psychiatric disorders in MS and their association with disability may be explained by both biological and psychosocial factors. First, it may reflect a shared underlying pathophysiologic process. For instance, MS patients with major depression had more T2-weighted lesions, a proxy for disease burden, than patients without.169 Inflammatory dysregulation has been implicated in depressive disorders,170 and bipolar disorder171 and is a 81  central component of MS disease pathology. The presence of depression could be a direct (biological) reaction to increased inflammation,170 which in turn leads to increased neurodegeneration and disability progression.172 Second, psychiatric comorbidities may contribute to maladaptive coping strategies, and poor health behaviours, which could alter the course of MS. For instance, depressed and anxious individuals are more likely to smoke,135 a known risk factor for MS disability worsening.42   Depression and anxiety were more common in women, and the relationship between mood or anxiety disorders and disability was only statistically significant among women. It is possible that because a mental health condition had to be medically recognized, that our findings are, in part, driven by sex differences either in mental health help-seeking behaviours or readiness of practitioners to diagnose mental health in women.173 Second, because on average, men progress faster in their disease than women (as shown in our study and others)74, it is possible that combined with the lower number of men, the effect size was diminished among men.    Strengths of this study include the large, representative cohort of clinic-attending MS patients. All data were prospectively collected, thereby eliminating the potential for recall bias. Selection bias is minimized by the broad criteria for inclusion, and the lack of an a priori research question when the data were collected. We used a highly specific case definition for mood or anxiety disorders, but the trade-off was lower sensitivity, meaning we may have underestimated the number of affected persons, such that the true risk of disability with a psychiatric condition could be even larger. Another limitation of this study is the operationalization of the ordinal EDSS as a linear scale. A 0.5 change on the scale can have clinically different meanings depending on where on the scale it occurs, and the probability of progression is not necessarily evenly 82  distributed along the scale.174 For these reasons we conducted supplementary analyses, using the linear MSSS scale, with the interpretation of findings remaining the same. This provides additional confidence in the results. The EDSS has other limitations, including imperfect intra- and inter-rater reliability.174 Future studies could examine the impact of psychiatric comorbidity on alternative clinical outcome measures, such as the Multiple Sclerosis Functional Composite or the Timed 25-Foot Walk (T25-FW) alone.    Psychiatric comorbidities were very common in this cohort and adversely associated with disability progression. Women with comorbid mood or anxiety disorders, in particular, were at a significantly increased risk of worsened disability. Future research should try to clarify if there are differences in progression between those with treated psychiatric conditions (pharmacological or non-pharmacological) and untreated individuals. MS disease progression is highly variable, and these results suggest that psychiatric comorbidities may explain some of the heterogeneity between individuals. Last, these findings pose the intriguing possibility of whether a well-designed, sufficiently powered randomized controlled trial could show whether effective treatment of a psychiatric disorders could reduce disability worsening.       83  Chapter 6: A population-based study comparing multiple sclerosis clinic users and non-users in British Columbia, Canada4 6.1 Background and objective The previous Chapters of this dissertation all focus on individuals with MS who attended an MS specialty clinic.  Outside of this work, much of what we know about MS comes from patients who access MS clinics.52,74,76,77 However, not all persons with MS visit an MS clinic.84 Using a validated algorithm to identify persons with MS outside of the clinic setting, we can investigate whether measurable differences existed between individuals attending an MS clinic to persons with MS who do not, which would be of interest to researchers and clinicians who study and treat these cohorts.   Objective: To characterize the clinical cohort in the context of the wider MS population by comparing incident MS cases who were MS clinic users to non-users of the specialty MS clinics in British Columbia, Canada.  6.2 Methods 6.2.1 Study design This was a record-linkage cohort study based on prospectively-collected patient data from the BCMS database and province-wide health administrative databases, details of which are outlined in Chapter 1, Subsection 1.6. Data for all patients who first visited a BC MS clinic by the end of 2004 were linked via their personal health number to their individual-level information contained within                                                  4 A version of this chapter has been published as McKay, KA, Tremlett H, Zhu F, Kastrukoff  L, Marrie RA, Kingwell E. A population‐based study comparing multiple sclerosis clinic users and non‐users in British Columbia, Canada. Eur J Neurol 2016; 23(6), 1093-1100.  84  several province-wide health administrative databases. The Medical Service Plan93 payment information file provided information on physician visits and the Discharge Abstract Database95 provided data on hospital visits, PharmaNet, provided data on prescriptions dispensed. The BC Vital Statistics Agency database98 provided mortality data, and together with the BC Ministry of Health’s Registration and Premium Billing Files97, enabled confirmation that an individual was alive and resident in BC. Finally, Census Geodata provided an area-level measure of socio-economic status (SES).99  6.2.2 Study population Incident cases of MS in the BC general population were identified using a validated algorithm of hospital and physician diagnostic codes.175,176 The MS case definition was >7 hospital or physician claims specifically for MS (ICD 9 code 340 or ICD 10 code G35) for people who were resident in BC for more than 3 years following their first demyelinating disease claim [See Appendix A; Table A.5 for the relevant ICD codes]), and >3 MS claims for those with 3 years of residency or less.176 To meet the incident case definition, each case had to be resident in BC for at least 5 years before their first demyelinating disease claim; this first claim date was considered the index date for both the non-BCMS clinic and the BCMS clinic cases. Incident cases between 1996 and 2004 (the last year of registration as a confirmed MS case in the BCMS cohort for this study) were examined.  6.2.3 Comparisons between clinic and non-clinic cases The incident MS cases that were seen at a BCMS clinic (‘clinic’) were compared with incident cases that had never attended a BCMS clinic (‘non-clinic’) during the study period. Demographic comparisons were made by sex, year, age and quintile of neighbourhood SES at the index date.  Dispensation of at least one prescription for a DMT, including IFNβ-1b (Betaseron®), IFNβ-1a  85  (Avonex® and Rebif®), glatiramer acetate (Copaxone®), or natalizumab (Tysabri®), at any time during follow-up was compared, as was the average time to reach the case definition (3rd claim for patients with < 3 years of follow-up, or 7th claim for those with > 3 years of follow-up), the number of distinct all-cause hospitalizations and the number of physician visits during follow-up. We performed an additional analysis in which pregnancy-related hospitalization claims were excluded on the basis that they are not representative of a medical illness, unlike other hospitalizations. Hospitalizations specifically for MS (as the primary diagnosis, or listed anywhere on the discharge report) were also compared. We examined the presence of specific comorbidities in the 8 years surrounding the index date (4 years prior and 4 years post). Comorbidities were selected based on their high prevalence among people with MS and the availability of validated algorithms to identify them using health administrative data in MS populations.141,156,159 The following comorbidities were identified by validated algorithms141,156,159 based on ICD-9/10 diagnostic codes from physician billings and hospital admissions (see Appendix A; Table A.4) and compared between the clinic and non-clinic cases: hypertension, hyperlipidemia, diabetes, chronic lung disease, migraine, and mood or anxiety disorders. Finally, the total number of distinct prescription medication classes dispensed in the year following the index date was compared between the clinic and non-clinic cases as a complementary and more global measure of comorbidity. A similar measure was shown to be the best predictor of future physician visits, and health care expenditures relative to 5 other measures of comorbidity in a cohort of older adults in BC.177 The DINs were used to group medications according to the World Health Organization Anatomic Therapeutic Chemical (WHO-ATC)155 classification system; the second level (main therapeutic group) was used to define unique drug classes.    86  6.2.4 Statistical analysis Comparisons between clinic cases and the non-clinic cases were assessed using the Pearson’s chi-squared test for categorical variables, and the Student’s t-test or Wilcoxon rank sum test for continuous variables. The number of hospitalizations and physician visits during follow-up were analyzed using negative binomial regression with findings reported as incidence rate ratios with 95% confidence intervals. To account for the differences in follow-up time, the logarithm of the follow-up time was included as an offset. The presence of each of the comorbidities of interest during the 8 years surrounding the index date (4 years before and 4 years after) was compared between clinic and non-clinic cases using logistic regression. All cases had full data for the four years prior to the index date, but those with less than 4 years of follow-up after the index date were excluded from the comorbidity comparisons. As a sensitivity analysis, the potential influence of excluding these cases with insufficient follow-up was assessed by restricting the comparison of comorbidities to the 4 year time period prior to the index date.  Findings were reported as odds ratios (ORs) with 95% confidence intervals. The count of distinct medication classes dispensed in the year following the index date was compared between the two groups using Poisson regression, with findings reported as rate ratios. All models were adjusted for age (continuous), sex, and index year (continuous). Analyses were performed using SAS Statistical Software Package 9.4 (SAS Institute Inc., Cary, NC).   All linkages were facilitated by Population Data BC. BC Ministry of Health and Data Stewardship Committee approved access to administrative health data. Cell sizes <5 were suppressed to protect the confidentiality of the participants.  87  6.3 Results There were 2,841 incident MS cases in British Columbia between 1996 and 2004; these included 1,648 clinic cases (58%) and 1,193 non-clinic cases (42%). The clinic cases represented 58% to 66% of the total for each year between 1996 and 2003, dropping to 38% in the final year (2004). The proportion of men and women, and distribution across the SES quintiles did not differ between the clinic and non-clinic groups. Although the clinic incident cases reached the administrative case criterion within a similar time period compared with the non-clinic incident cases, the clinic cases were approximately 5 years younger at the time of their incident claim (41 v. 46 years old). Nearly half of the clinic cases had received a prescription for a DMT (51%) at some point during follow-up; not unexpectedly, this proportion was significantly greater than that seen with the non-clinic cases (1%). The characteristics of the clinic and non-clinic cases are shown in Table 6.1.                88    MS cases  (n = 2,841)  Incident cases (1996-2004) Clinic cases n=1,648 Non-clinic cases n=1,193  p-value Sex                                     Females                                                                               Males 1,242 (75%)406 (25%) 878 (74%)      315 (26%) a0.31 Prescription filled for an MS disease-modifying drug at any time during follow-up 847 (51%)   13 (1%) a<0.001 SES quintilec                      1 (Low)                                                                  2                                                 3                                                4                                       5 (High) 277 (17%)   286 (18%)   347 (22%)  342 (21%)  359 (22%) 209 (18%) 226 (20%) 257 (22%) 236 (20%) 226 (20%) a0.38 Age at incident claim in years Median (1st, 3rd quartile)  41.1          (33.4, 47.9)  45.8           (37.7, 54.9)  b<0.001 Index year                               1996 1997 1998 1999 2000 2001 2002 2003 2004 169 (10%)   212 (13%)  184 (11%)  196 (12%)  205 (13%)  207 (12%)  191 (12%)  175 (11%)  109   (7%) 116 (10%) 111 (9%) 131 (11%) 147 (12%) 126 (11%) 134 (11%) 122 (10%) 121 (10%) 185 (16%) a<0.001 Time to meet case definition  (yrs)d Median (1st quartile; 3rd quartile) 1.55             (0.73; 2.99) 1.65            (0.62; 3.75) b0.28 Available follow-up time from index date Median (1st quartile; 3rd quartile)  8.3 (6.3; 8.4)  7.3 (4.7; 9.9)  b<0.001 Table 6.1 Characteristics of the MS clinic and the non-clinic cases. a Pearson’s Chi-squared test; bWilcoxon rank sum test; c SES quintile; as measured at the index date; missing for 78 cases;d Beginning of time period to meet case definition measured at index date.   The non-clinic cases had higher rates of hospitalizations (all-cause, either with or without pregnancy-related admissions); however, when the analysis was limited to hospitalizations in which the primary diagnosis was specifically for MS, the non-clinic group had lower rates (see Table 6.2).  89  The same pattern emerged for physician visits, in which non-clinic cases had higher rates of physician services use in general, but significantly lower rates specifically coded as MS (see Table 6.2).    Incidence Rate Ratio (95%CI) aAdjusted Incidence Rate Ratio (95%CI) Hospitalizations (all cause; excluding pregnancy)  2.05 (1.85 – 2.28) 1.73 (1.55 – 1.92) Hospitalizations (all-cause;  including pregnancy) 1.97 (1.78 – 2.17) 1.72 (1.56– 1.90) Hospitalizations (MS- as primary diagnosis) 0.67 (0.52 – 0.89) 0.65 (0.49 – 0.86) Hospitalizations (MS-reported anywhere on the discharge report) 0.90 (0.76 – 1.08) 0.83 (0.67 – 0.96) Physician Visits (all cause)  1.24 (1.18 – 1.30) 1.14 (1.08 – 1.20) Physician Visits (MS-specific) 0.49 (0.46 – 0.52) 0.43 (0.41 – 0.46) Table 6.2 Comparison of health services utilization between MS clinic and non-clinic cases. Reference group is the MS clinic group; rate ratios > 1 indicate a higher rate for non-clinic cases relative to clinic cases. aAdjusted for sex, age and index year.  We also found differences in comorbidities among the 2,650 patients (1,600 clinic cases and 1,050 non-clinic cases) with sufficient follow-up for this comparison (i.e. four years prior to and four years following their index date). After adjustment for age, sex, and index year, the non-clinic cases had higher odds for meeting each of the definitions of hypertension, chronic lung disease, diabetes, and mood or anxiety disorder at some point during the 8 years surrounding the index date (see Table 6.3), but the odds of hyperlipidemia did not differ significantly between the two groups (see Table 6.3). These findings were no different when the period for comorbidity measurement was restricted to the 4 years prior to the index date, with inclusion of all cases (data not shown).  The non-clinic group had higher rates of comorbidity in general, as measured by the number of distinct prescription medication classes dispensed in the year following the index date (see Table 6.3).    90  Specific Comorbidity Odds Ratio (95% CI) aAdjusted Odds Ratio (95% CI) Chronic Lung Disease 1.70 (1.30 – 2.21) 1.66 (1.27 – 2.19) Hyperlipidemia 1.45 (1.03– 2.03) 0.98 (0.68 – 1.40) Hypertension 1.95 (1.57 – 2.44) 1.41 (1.11  - 1.78) Diabetes 1.82 (1.39 – 2.39) 1.58 (1.19 – 2.10) Migraine 1.21 (0.99 – 1.50) 1.34 (1.08 – 1.67) Mood or Anxiety Disorder 1.24 (1.06 – 1.45) 1.25 (1.06 – 1.48) General Comorbidity Rate Ratio (95% CI) aAdjusted Rate Ratio (95% CI) Number of prescription classes dispensed in year following index date 1.17 (1.10 – 1.24) 1.08 (1.02– 1.15) Table 6.3 Comparison of comorbidity between MS clinic and non-clinic cases. Reference group is the MS clinic group; odds ratio or rate ratio > 1 indicate higher odds for the non-clinic cases relative to the clinic cases. aAdjusted for sex, age and index year.   6.4 Discussion We compared the characteristics of individuals with definite MS who registered at an MS Clinic and individuals who met an administrative definition of MS and had not registered at an MS Clinic in British Columbia, Canada and identified several differences. People with MS who did not register at an MS clinic were older, accessed health services more frequently, and had a higher risk of comorbidity than those registered with an MS clinic. In addition, while the groups were comparable in their sex distribution and socioeconomic status, only 1% of the non-clinic patients had filled a prescription for an MS-specific DMT.  Our findings have several important implications: i) even within a publicly funded healthcare system, a high number of individuals with MS may not access an MS specialty clinic and may be managed in the community; ii) the needs of MS patients managed in the community may differ from those referred to an MS clinic, including management of comorbidity; iii) depending on the research question, studies that use clinic-based MS cohorts may not be generalizable to those that do  91  not seek MS specialist care; iv) access to total population-based health administrative data offers the opportunity to gain a broader understanding of MS.  Using the validated administrative algorithm we estimate that approximately 60% of people who were newly diagnosed with MS (i.e. incident cases) and living in BC between 1996 and 2004 were registered at one of the four specialty MS clinics in the province. In BC, MS clinics are part of the universal health program, and offer comprehensive care to people living with MS, including MS specialist neurologists and nurses, neuro-ophthalmologists, physiotherapists, psychiatrists, and social workers. To attend a specialty MS clinic one must be referred by a physician. Importantly, these clinics are the only resource for DMT prescriptions under the BC government’s reimbursement scheme; consequently, only 1% of the non-clinic cases were prescribed a DMT during the study period. While prescribing patterns and drug reimbursement policies vary between jurisdictions, our findings may indicate that the population-based rates of DMT use, as reported in other studies (i.e. outside of BC), might be lower than previously thought.1,178,179  The only other study, to our knowledge, that has compared a group of MS cases that had attended an MS clinic to a group that had not, was based in Lorraine, France and focused on demographic and disease characteristics.84 Similar to our findings, they found no differences in the sex distribution, but an older age at onset in the non-clinic cases. We could not capture age at MS symptom onset from the administrative data, but the age at first demyelinating claim was 5 years older on average in the non-clinic group. It is possible that the non-clinic cases were older at MS symptom onset, but it is also possible that the findings indicate a delay in the medical recognition of MS.14,166,175,180 As the non-clinic cases were more likely to have comorbidity, it is conceivable that their early MS symptoms were not noticed or misattributed to a comorbid condition, thereby contributing to a delayed diagnosis.166 It is not possible to tell whether these individuals were less  92  likely to be referred to an MS clinic or if they actively chose not to attend, however it may represent a missed opportunity to offer treatment (pharmacological or non-pharmacological). It is also conceivable that having benign disease which did not require treatment, or progressive disease for which there were no treatments, may have precluded one from attending an MS clinic.  This is the first study that we are aware of to compare comorbidities and health services utilization between clinic and non-clinic MS cases. The non-clinic MS cases were burdened with more comorbidities than the MS clinic cases, even after accounting for the age difference. They were more likely to have hypertension, diabetes, chronic lung disease, migraine, or a mood or anxiety disorder around the index date as well as a higher burden of global comorbidity, based on prescriptions filled. The importance of comorbidities in chronic diseases, such as MS, has been recently highlighted.141,159,181 Emerging work indicates that  comorbidity is associated with a delayed MS diagnosis, and a higher level of disability at diagnosis.166 Our results suggest that patients who accessed a specialty MS clinic had less comorbidity; thus, estimates from a clinic-based sample may underestimate the true burden of comorbidity in MS. This emphasizes the importance of broader population-based estimates141 that are derived from both MS clinic and non-clinic users.   Health services utilizations in the non-clinic group for all-cause hospitalizations and physician visits were higher than in the clinic group, and these were independent of age. In contrast to this, we observed that the clinic patients had a higher rate of physician visits and hospitalizations specifically for MS. While the former is not unexpected in a group of patients accessing an MS clinic, the reason for their higher rate of hospitalizations for MS is less obvious. It is possible that the MS clinic cases generally have more active disease (e.g. more frequent relapses)11 for which they were more likely to seek care and perhaps require hospitalization.   93  Study limitations included an inability to access and compare MS-specific clinical features across the MS clinic and non-clinic patients, such as disease course (i.e. relapsing-onset or primary-progressive) or relapse frequency. We did not have sufficient information on a person’s place of residence in order to estimate how far they reside from an MS clinic, which likely influences whether they attend. Study strengths included a large cohort of 2,841 cases, comprising all incident MS patients in the province of British Columbia over a 9 year incidence period. Our study also captured longitudinal data, with 18 years of follow-up, and access to extensive population-based administrative health data linked to a province-wide MS clinical database. Furthermore, the incident MS cases and the specific comorbidities were identified using algorithms that were tested and validated in Canadian MS cohorts.175,176 The MS algorithm was estimated to have a sensitivity of 88% and a specificity of 68% (with a confirmed diagnosis of MS by a MS specialist neurologist as the gold standard) among all individuals in Nova Scotia, Canada with at least one demyelinating disease claim.176 Specificity (or avoidance of false positives) would naturally increase considerably among the general population, for whom the vast majority have never had a demyelinating claim. For example, using a similar 7 claim definition in the Canadian province of Ontario yielded an estimated specificity of 100%.182 Nonetheless, it is possible, that inclusion of a small number of false positive cases in the non-clinic group may have contributed to some of the differences observed, although this would unlikely be high enough to influence interpretation of findings.  No false positive MS cases would have been included in the MS clinic group because all included cases had been diagnosed with definite MS by a specialist MS neurologist using the most current internationally recognised criteria; 1,648 of 1,735 clinically confirmed MS cases were correctly identified by the algorithm which indicates that the algorithm has very high sensitivity in the BC MS population. The 87 cases that were not identified  94  by the algorithm were more likely to have had their incident claim in 2004, and therefore had less follow-up time, or opportunity to meet the administrative case definition, and, not unexpectedly, had fewer claims within the 4 years after their index date compared with people that met the definition.   The BC clinical database was broadly representative of new cases of MS in the province of BC, in that it captured the majority of incident MS cases, and the sex distribution and socioeconomic status were comparable to the wider MS population. Importantly, nearly all of the patients that received a DMT were captured by the BCMS database, which allows for comprehensive monitoring of the long-term safety and effectiveness of these drugs. Our results indicate that people who attended MS clinics were younger at their first MS-related claim, suggesting an earlier age at first medical recognition of their MS. Studies of clinic populations have been enormously valuable in developing therapies, understanding the natural history of MS, and generally advancing our knowledge of this complex disease; however, the MS community should remain mindful that a broader MS population exists and may differ in both subtle and important ways to those assessed in MS specialty clinics.     95  Chapter 7: Conclusions 7.1 Summary of findings The long-term goal of this dissertation is to contribute to the broader understanding of the impact of mental health and health behaviours in MS. Much of the previous research in this area has focused on estimating the rates of depression in the MS population, with less dedicated to the study of other psychiatric conditions or health behaviours.28 Even fewer studies have examined the relationships between mental health, health behaviors, and MS. Those that did tended to focus on single factors, and were typically cross-sectional in design, creating challenges when attempting to establish a broader understanding of these relationships.39,108,167,183 This dissertation represents one of the first multifactorial, longitudinal assessments of these associations.  Studies presented in Chapters 2, 3, and 5 support previous findings of a high prevalence of psychiatric disorders in MS. Using a self-reported psychometric scale, more than half of the longitudinal cross-Canada cohort of prevalent MS patients met criteria for anxiety, while over a third met criteria for depression over two years (Chapters 2 & 3).135,184 Nearly half of incident MS cases (followed for an average of nine years) met the case definition for depression, and nearly a quarter met the definition for anxiety using administrative data (Chapter 5). Individuals seen outside of the MS clinic setting had even higher odds of having a mood or anxiety disorder, relative to clinic users (Chapter 6).185  Alcohol dependence and smoking were associated with both anxiety and depression (Chapter 2); however, smoking had no measurable association with the subsequent risk of incident depression or anxiety. A history of alcohol dependence was associated with a higher risk of developing depression, and pre-existing depression increased the risk for incident alcohol dependence,  96  suggesting a bidirectional relationship (Chapter 2).135 Nearly one-quarter of participants in this cohort were not adherent to their injectable DMT (defined as taking less than 80% of their expected doses) and over half missed at least one dose during the study period (Chapter 3). Contrary to our hypothesis, we found no altered risk of non-adherence among persons with depression or anxiety. Alcohol dependence, perceived functional cognitive difficulties, longer disease duration, and mild disability status were associated with non-adherence.184 To estimate the validity of the self-reported adherence measure used in Chapter 3, we compared these reports to pharmacy records of prescriptions dispensed in British Columbia. Patients were able to accurately report whether they were on a DMT, and which brand of medication they were taking, as expected (Chapter 4). Those who identified as non-adherent were highly likely to also exhibit non-adherence over the previous year, and be at risk of future non-adherence (Chapter 4). The sensitivity of self-reported missed doses was lower than hypothesized. Assuming that the pharmacy records are a more precise measure than self-report, this suggests that individuals who reported high adherence, may still be ‘at risk’ for non-adherence. Based on both definitions, adherence rates were suboptimal; more than 1 in 10 (13%) participants were non-adherent based on the 30-days self-report and over 4 in 10 (43%) were non-adherent over the subsequent year based on pharmacy records.  Using linked clinical and province-wide health administrative data, we found that the presence of a mood or anxiety disorder significantly increased subsequent neurologic disability over an average of nine years (Chapter 5). Depression was the strongest predictor of a worsened progression, and the association was stronger in women than men.  Last, we estimated how clinic patients, who were the focus of this dissertation, compared to the wider population of persons with MS, who do not attend clinic. We found that over 40% of incident onset individuals with MS in British Columbia did not attend an MS clinic. Sex and socioeconomic  97  status distributions were similar, but non-users of the MS clinics were older and had higher rates of comorbidity than the clinic population (Chapter 6).185   7.2 Integration of findings The relationships between MS, mental health, health behaviours, adherence to medications, and disability are represented in Figure 7.1. This framework characterizes the potential roles of mental health and health behaviours in MS based on evidence contained in this dissertation, and findings from previous observational research.  The rate of disability worsening varies substantially from person to person, but it is considered a persistent, inevitable process for most persons with MS. Few factors have been established as mediators of long-term disability progression.8 The study reported in Chapter 5 suggested that mental health may lead to disability worsening, but we cannot rule out the possibility of a bidirectional relationship (3↔1, Figure 7.1). While psychosocial factors which lead to maladaptive coping strategies may be responsible, in part, for this relationship,186 a biological connection is also possible. Patterns of CNS lesions and cerebral atrophy have been associated with depression in MS patients.187,188  The relationship between smoking and disability worsening has been systematically reviewed outside of this dissertation with best evidence suggesting that one leads to the other in MS (2→1, Figure 7.1).42 A number of  biological mechanisms have been postulated to explain how smoking increases MS risk and disability, most of which have focused on the detrimental effects of smoking on the immune response104 and the CNS.41 Alcohol may have a protective effect against long-term disability, but this cross-sectional study requires replication.183  The influence of adherence to MS DMTs on disability has typically focused on relapses or hospitalizations, both of which are increased following non-adherence.62 We found that persons  98  with mild disability were more likely to be non-adherent than those with moderate disability, but cannot be certain of the direction of this relationship (1↔4, Figure 7.1). It’s conceivable that people who are experiencing less serious disease might feel less motivated to take their medication. Findings outlined in Chapter 3 did not support the relationship between mental health and adherence to DMTs; however, previous research has suggested an association between depression and missed doses in MS (3→4, Figure 7.1).66,136 Depression appears to increase the likelihood of treatment discontinuation,137 which could, in part, explain our null findings. Depressed patients may have discontinued therapy prior to our study start date, and therefore have been excluded from the analysis. Similar to previous research,60 we found that alcohol dependence, but not smoking, was associated with a significantly increased risk of non-adherence to MS DMTs (2→4, Figure 7.1). To the best of my knowledge, specific mechanisms that explain this link have not been identified. A potential pathway is that each of these behaviours reflects a general process of maladaptive coping.  Mental health and health behaviours were found to influence one another in a bidirectional fashion. Depression increased the risk of alcohol dependence, and alcohol dependence increased the risk of depression (2↔3, Figure 7.1). While smoking was associated with both depression and anxiety, it did not increase the risk of incident mental health conditions (2↔3, Figure 7.1). Smoking and alcohol use may be a means of self-medication, which produce a short-term alleviation of psychiatric symptoms, but then induce longer-term increases in depression and anxiety in a cyclical fashion.  These relationships were based on an MS clinic population. Chapter 6 suggested that the wider MS population has a similar sex and socioeconomic status distribution, but that these patients were older at first medical recognition of MS, and may have even higher rates of mood and anxiety disorders than the MS clinic population. Whether the relationships revealed in Chapters 2 and 5, which  99  focused on mood and anxiety disorders, would be applicable to the wider MS population is challenging to estimate without detailed information from these individuals. Chapters 3 and 4 focused exclusively on DMT users, virtually all of whom (98%) were in the BCMS database, so we expect that these relationships are applicable to DMT users in general.   Figure 7.1 Schematic depiction of the potential relationships between MS, mental health, and health behaviours. Red arrows represent relationships inferred from the studies in this dissertation. Black arrows represent associations that were based on evidence from the wider literature. The dotted arrow indicates areas of uncertainty, with study findings differing across available publications. Numbers are used to reference this diagram within the text of section 7.2.   7.3 Strengths and limitations  Methodological strengths of this dissertation include the longitudinal design, large samples, use of comprehensive, population-based data sources, and the focus on clinically-relevant issues. Advantages of using linked health administrative databases include: the minimal burden on patients;  100  confidentiality of patient information; and the longitudinal, population-based nature of the data.189 Further, the definitions used to ascertain cases of MS and comorbid conditions were validated for use in Canada.156,159,175,176  Three major sources of bias in epidemiological studies, broadly categorized as selection bias, information bias, and confounding, have been mitigated where possible. Selection bias can be thought of as a misrepresentation of effects resulting from a biased selection of study subjects.100 The outcomes were not established during data collection in the studies included in Chapters 5 and 6, thereby minimizing this potential bias. Consecutive sampling was employed in the cross-Canada longitudinal cohort study (Chapters 2 – 4), which is considered to have the best validity of the non-random sampling techniques.190 The response rate was high: of all eligible participants, 83% agreed to participate. Multiple outcomes were collected through this study, and participants were informed that they would be asked to complete a series of questionnaires related to comorbidity and quality of life. It is therefore unlikely that participation status would be influenced by the outcomes contained in this dissertation. Selection bias can occur during the implementation of a study as well, in terms of attrition (withdrawal or loss to follow-up) rates which disproportionately affect one group of participants. By the end of two years, 93% of this cohort had complete follow-up. A threshold of 60% has been reported as an acceptable rate of follow-up to minimize bias.191   Information bias arises due to measurement error, or misclassification. All participants included in Chapters 2 through 5 had a neurologist-confirmed diagnosis of MS, thereby eliminating the possibility of misclassification of MS. Cases of MS were identified in Chapter 6 by a validated algorithm of hospital and physician diagnostic codes specific for MS.175,176 The MS algorithm was estimated to have a sensitivity of 88% and a specificity of 68% (with a confirmed diagnosis of MS by a MS specialist neurologist as the gold standard) among all individuals in Nova Scotia, Canada  101  with at least one demyelinating disease claim.176 Within our cohort, 95% of clinically confirmed MS cases were correctly identified by the algorithm, which indicates that it has a high sensitivity in the BC MS population. Nevertheless, it is possible that false positive cases may have been included in the non-clinic group, which may have contributed to some of the differences observed. We do not expect that this number would have been high enough to influence interpretation of findings.  The algorithms used to identify comorbidities were also validated in the MS population, and tended to have modest sensitivity and high specificity. The imperfect sensitivity may have led to the misclassification of comorbidity in Chapters 5 and 6. Clinic and non-clinic users would have been equally affected by this misclassification in Chapter 6, thus creating a non-differential misclassification. This form of misclassification typically results in a bias toward the null.100 The algorithm used to define the presence of a mood or anxiety disorder was estimated to have a sensitivity of 63% and a specificity of 87%, when compared to medical chart review in Manitoba, Canada.156 This suboptimal sensitivity suggests that we likely misclassified some individuals who had a mood or anxiety disorder as being free of such disorders. As such, we may have underestimated the true number of affected persons, and our estimate of the relationship between psychiatric comorbidity and disability progression may have been diluted. Further, we only identified medically-recognized psychiatric comorbidities. Not all persons with MS and comorbid psychiatric conditions receive medical attention for these conditions. In fact, in the longitudinal cross-Canada cohort, one-third of individuals with symptoms of depression and two-thirds with symptoms of anxiety had not received a diagnosis for these conditions.192 We minimized misclassification by using validated questionnaires when possible. The thresholds used in the Hospital and Anxiety Depression Scale had a sensitivity of 90% and a specificity of 87.3% as compared to major depression defined by the Structured Clinical Interview in the DSM-IV  102  in a Canadian clinic setting.32 The CAGE questionnaire is a validated screening tool for alcohol dependence,193 as were the questions regarding smoking status in the self-reported questionnaires completed by the cross-Canada longitudinal cohort.107   Data gathered from clinical and administrative sources was done so prospectively, and without specific knowledge of our a priori research question, so were not susceptible to recall or reporting bias. There is a potential for both of these forms of bias in the questionnaire-based studies (Chapters 2 and 3). Questions regarding smoking history, alcohol dependence, and adherence to medication all rely on a patient’s accurate recollection of these behaviours. Smoking status has been studied over time in MS and found to be reliable, especially in terms of identifying as a current, ever- or never-smoker, as we did in our study.194  The CAGE questionnaire addresses alcohol dependence, as opposed to the amount of alcohol consumed over a specific timeframe. Considering the severity of the syndrome, it is unlikely that a person would not recall experiencing alcohol dependence. Finally, when assessing self-reported missed doses, the recall period of 30 days used has been shown to reduce the ceiling effects relative to shorter time periods, and minimize recall bias relative to longer periods.139 Reporting bias refers to the selective revealing or suppression of information by research participants. Self-reported alcohol use123 and adherence to medication148 are particularly susceptible to this form of bias. To mitigate this, participants were approached by a research assistant, as opposed to their treating neurologist (i.e., the primary prescriber of MS DMTs). It is possible that people have differing perspectives on what constitutes problem drinking, and some people may have been misclassified for this reason. Self-reported medication adherence is particularly vulnerable to social desirability bias. Our findings in Chapter 4 suggest that people were optimistic and potentially over-reported their ability to adhere to medication, relative to the more object  103  estimates from pharmacy records. We found that persons who reported alcohol dependence were more likely to be non-adherent. This finding may relate, in part, to a reporting bias in which individuals that are more likely to ‘honestly’ report alcohol dependence, may also be more likely to report missing doses of their medication. Self-reported smoking status is potentially less susceptible to these biases; previous work has shown a high agreement between self-reported smoking and serum cotinine levels, an objective measure of smoking exposure.195 A similar biomarker is not readily suitable for alcohol use.  Confounding occurs when a non-causal association between two variables arises as a result of a third variable. We controlled for known confounders when possible, but residual confounding may have biased our results. This could have resulted from the unavailability of information on potential confounders in our dataset (for instance, we did not have information on smoking status for Chapters 5 and 6), or imprecise measures of potential confounding factors, such as socioeconomic status, which was measured as neighbourhood-level income.  Findings from Chapter 6 emphasized that studies which include MS clinic-based patients may not be generalizable to individuals that do not seek MS specialist care. An important finding was that the odds of having a mood or anxiety disorder were 25% higher among non-users of the MS clinics. At a population-level, our estimates may have underestimated the true burden of this comorbidity, based on clinic samples (Chapters 2 through 5).   7.4 Clinical significance and implications Our findings have several important clinical implications. First, the proportion of MS patients who experience depression and anxiety was high, even at early stages in the disease. Neurologists and primary care providers should be aware of the high risk of depression and anxiety among MS  104  patients. Having a psychiatric comorbidity was associated with alcohol dependence, smoking, and worsening in MS disability outcomes.  A high proportion of patients were not adherent to their MS disease-modifying therapy. The first step toward improving adherence involves identifying whether patients are adherent to their medication or not. We found that a simple question regarding missed doses over 30 days was highly specific when compared to a more objective measure of longer-term adherence. The sensitivity of this question was suboptimal, such that persons who self-reported ‘good’ adherence, may still be at risk of longer-term non-adherence (based on the more objective measure of dispensed prescriptions).  No single intervention can improve adherence for all patients. Healthcare professionals may wish to work closely with patients to tailor an individualized approach which addresses specific barriers to adherence. Based on our findings, clinicians should be particularly aware of the possibility of non-adherence among patients with a longer disease duration, mild MS disability, alcohol dependence, and cognitive difficulties. Non-adherence was a persistent pattern in our cohort – the strongest risk factor for future non-adherence was a history of non-adherence, which highlights the importance of continually assessing prior ability to adhere to a drug.  Instead of advocating for a paternalistic approach, I would argue that the healthcare provider should endeavour to provide education and open communication, allowing for patient engagement in their own health strategies.50 Patients have better health outcomes when they are actively involved in their own healthcare.196 The high proportion of persons affected by and consequences of psychiatric comorbidity noted in this dissertation support the use of a screening tool for such conditions, which could be incorporated  105  into routine care within the MS clinic, which has been recommended by the Goldman Consensus Group.65 A simple two-question screen has been validated in the MS population, with a sensitivity of 99% and specificity of 87%, when compared to major depressive disorder defined by the Structured Clinical Interview for the DSM-IV.31 The utility of screening for depression or anxiety depends on the availability and effectiveness of appropriate treatment for such disorders. Until clinics can support persons with comorbid psychiatric disorders in an effective manner, then screening for such disorders would be impractical. It is perhaps in part for this reason, that empirical studies of the efficacy of screening for depression (in primary care) have not convincingly shown that it improves health outcomes.197   A systematic review reported that pharmacological and psychological treatments for depression are effective in the MS population.198 In the absence of treatment, symptoms of depression are likely to worsen in MS,199 suggesting that proactive treatment is an appropriate strategy. 7.5 Knowledge translation The theme of this dissertation arose from observations I made as a research coordinator for the University of British Columbia MS Clinical Trials group (2009-2011). I noted the pervasive nature of psychiatric disorders; the significant gaps in our understanding of the wider impact of these disorders; and the limited mental health services available to patients seen both in the MS clinic and wider community.  Through a series of quantitative epidemiological studies, I have provided empirical evidence of the impact of mental health and health behaviours. I felt that in order to effect change, this work should be disseminated beyond the traditional reporting in scientific journals. As such, I have composed a briefing note to be distributed to policymakers and key stakeholders within the British Columbia,  106  Manitoba, and Nova Scotia Ministries of Health, and the MS Society of Canada. This note summarizes findings from this dissertation, as well as broader findings related to mental health and health behaviours from the cross-Canada collaboration, the Epidemiology and Impact of Comorbidities in MS, led by Dr. Ruth Ann Marrie (Appendix D). It underscores the consequences of psychiatric comorbidity in MS, which include reduced quality of life, increased pain and fatigue, disability worsening, and mortality. Ultimately, we hope that it will be used to support the case for improved mental health services for persons living with MS in Canada and for more resources to be dedicated to research into this important area.  7.6 Conclusions and future directions This dissertation contributes to the wider discourse on the potential consequences of mental health and health behaviours in persons with MS. MS is an unpredictable and complex disease with considerable variability in outcomes. Through a series of province and nation-wide studies, we found that psychiatric conditions and adverse health behaviours were common in MS, and may explain some of the heterogeneity in health outcomes. Psychiatric comorbidities were associated with alcohol dependence, smoking, and worsening in MS disability. Adherence rates to medication were suboptimal, and associated with alcohol dependence and a history of poor adherence. A substantial portion of persons with MS do not access MS specialty clinics; these patients were older and had more comorbidities than the MS clinic users.  This dissertation has laid the foundation for further research into these areas. Firstly, confirmation of these associations using alternative data sources or definitions should be attempted. For instance, future studies should confirm and expand the findings surrounding the association between psychiatric comorbidities and disability worsening. Improvements could include a more sensitive  107  definition of psychiatric comorbidity. Second, we focused predominantly on disability worsening, but there are other important outcomes to persons with MS, including cognition, educational achievements, and employment. The influence of mental health on these areas could be explored. Third, future studies should determine whether interventions which effectively treat or prevent psychiatric comorbidities could positively impact other health outcomes in MS. Ideally, effective prevention or treatment of psychiatric conditions and reduction of adverse health behaviours should be integrated into standard MS care.     108  References 1.  Multiple Sclerosis International Federation. Atlas of MS 2013: Mapping multiple sclerosis around the world. 2013. 2.  Leary SM, Porter B, Thompson A. Multiple sclerosis: diagnosis and the management of acute relapses. Postgrad Med J 2005; 81: 302–308. 3.  Tremlett H, Paty D, Devonshire V. Disability progression in multiple sclerosis is slower than previously reported. Neurology 2006; 66: 172–177. 4.  Pugliatti M, Rosati G, Carton H, et al. The epidemiology of multiple sclerosis in Europe. Eur J Neurol 2006; 13: 700–722. 5.  Dendrou CA, Fugger L, Friese MA. Immunopathology of multiple sclerosis. Nat Rev Immunol 2015; 15: 545–558. 6.  Hemmer B, Kerschensteiner M, Korn T. Role of the innate and adaptive immune responses in the course of multiple sclerosis. Lancet Neurol 2015; 14: 406–419. 7.  Calabrese M, Magliozzi R, Ciccarelli O, et al. Exploring the origins of grey matter damage in multiple sclerosis. Nat Rev Neurosci 2015; 16: 147–158. 8.  McKay KA, Jahanfar S, Duggan T, et al. Factors associated with onset, relapses or progression in multiple sclerosis: a systematic review. Neurotoxicology; In Press. 9.  Belbasis L, Bellou V, Evangelou E, et al. Environmental risk factors and multiple sclerosis: an umbrella review of systematic reviews and meta-analyses. Lancet Neurol 2015; 14: 263–273. 10.  Sawcer S, Franklin RJM, Ban M. Multiple sclerosis genetics. Lancet Neurol 2014; 13: 700–9. 11.  Lublin FD, Reingold SC, Cohen J a, et al. Defining the clinical course of multiple sclerosis: The 2013 revisions. Neurology 2014; 83: 278–286. 12.  Polman CH, Reingold SC, Banwell B, et al. Diagnostic criteria for multiple sclerosis : 2010 revisions to the McDonald Criteria. Ann Neurol 2011; 69: 292–302. 13.  Miller DH, Weinshenker BG, Filippi M, et al. Differential diagnosis of suspected multiple sclerosis: a consensus approach. Mult Scler 2008; 14: 1157–74. 14.  Kingwell E, Leung AL, Roger E, et al. Factors associated with delay to medical recognition in two Canadian multiple sclerosis cohorts. J Neurol Sci 2010; 292: 57–62. 15.  Koch M, Kingwell E, Rieckmann P, et al. The natural history of secondary progressive multiple sclerosis. J Neurol Neurosurg Psychiatry 2010; 81: 1039–43. 16.  Koch M, Kingwell E, Rieckmann P, et al. The natural history of primary progressive multiple  109  sclerosis. Neurology 2009; 73: 1996–2002. 17.  McKay KA, Kwan V, Duggan T, et al. Risk factors associated with the onset of relapsing-remitting and primary progressive multiple sclerosis: a systematic review. Biomed Res Int 2015; 2015: http://dx.doi.org/10.1155/2015/817238. 18.  Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983; 33: 1444–1452. 19.  Kurtzke JF. A new scale for evaluating disability in multiple sclerosis. Neurology 1955; 5: 580. 20.  Roxburgh RHSR, Seaman SR, Masterman T, et al. Multiple Sclerosis Severity Score: Using disability and disease duration to rate disease severity. Neurology 2005; 64: 1144–1151. 21.  World Health Organization and Calouste Gulbenkian Foundation. Integrating the response to mental disorders and other chronic diseases in health care systems.  http://www.who.int/mental_health/publications/gulbenkian_paper_integrating_mental_disorders/en/ (2014, accessed 15 February 2017) 22.  Patten SB, Beck CA, Kassam A, et al. Long-term medical conditions and major depression: Strength of association for specific conditions in the general population. Can J Psychiatry 2005; 50: 195–202. 23.  Moussavi S, Chatterji S, Verdes E, et al. Depression, chronic diseases, and decrements in health: results from the World Health Surveys. Lancet 2007; 370: 851–58. 24.  Fine LJ, Philogene GS, Gramling R, et al. Prevalence of multiple chronic disease risk factors: 2001 National Health Interview Survey. Am J Prev Med 2004; 27: 18–24. 25.  Jané-Llopis E, Matytsina I. Mental health and alcohol, drugs and tobacco: a review of the comorbidity between mental disorders and the use of alcohol, tobacco and illicit drugs. Drug Alcohol Rev 2006; 25: 515–536. 26.  Marrie RA, Horwitz R, Cutter G, et al. High frequency of adverse health behaviors in multiple sclerosis. Mult Scler 2009; 15: 105–113. 27.  Siegert RJ, Abernethy DA. Depression in multiple sclerosis: a review. J Neurol Neurosurg Psychiatry 2005; 76: 469–475. 28.  Feinstein A, Magalhaes S, Richard J-F, et al. The link between multiple sclerosis and depression. Nat Rev Neurol 2014; 10: 507–517. 29.  Marrie RA, Reingold S, Cohen J, et al. The incidence and prevalence of psychiatric disorders in multiple sclerosis: A systematic review. Mult Scler 2015; 21: 305–317. 30.  Sullivan MJL, Weinshenker B, Mikail S, et al. Screening for major depression in the early stages of multiple sclerosis. Can J Neurol Sci 1995; 22: 228–231.  110  31.  Mohr DC, Hart SL, Julian L, et al. Screening for depression among patients with multiple sclerosis: two questions may be enough. Mult Scler 2007; 215–219. 32.  Honarmand K, Feinstein A. Validation of the Hospital Anxiety and Depression Scale for use with multiple sclerosis patients. Mult Scler 2009; 15: 1518–1524. 33.  American Psychiatric Association. Diagnostic and statistical manual of mental disorders (DSM-5®). American Psychiatric Pub., 2013. 34.  Marrie RA, Fisk JD, Tremlett H, et al. Differences in the burden of psychiatric comorbidity in MS vs the general population. Neurology 2015; 85: 1972–1979. 35.  Spiegel D, Giese-Davis J. Depression and cancer: mechanisms and disease progression. Biol Psychiatry 2003; 54: 269–282. 36.  Leserman J. HIV disease progression: Depression, stress, and possible mechanisms. Biol Psychiatry 2003; 54: 295–306. 37.  Mattioli F, Bellomi F, Stampatori C, et al. Depression, disability and cognitive impairment in multiple sclerosis: a cross sectional Italian study. Neurol Sci 2011; 32: 825–32. 38.  Möller A, Wiedemann G, Rohde U, et al. Correlates of cognitive impairment and depressive mood disorder in multiple sclerosis. Acta Psychiatr Scand 1994; 89: 117–121. 39.  Koch M, Mostert J, Heerings M, et al. Fatigue, depression and disability accumulation in multiple sclerosis: a cross-sectional study. Eur J Neurol 2009; 16: 348–52. 40.  Sullivan E V., Pfefferbaum A. Neurocircuitry in alcoholism: a substrate of disruption and repair. Psychopharmacology (Berl) 2005; 180: 583–594. 41.  Smith KJ, Kapoor R, Hall SM, et al. Electrically active axons degenerate when exposed to nitric oxide. Ann Neurol 2001; 49: 470–476. 42.  Handel AE, Williamson AJ, Disanto G, et al. Smoking and multiple sclerosis: an updated meta-analysis. PLoS One 2011; 6: e16149. 43.  Manouchehrinia A, Tench CR, Maxted J, et al. Tobacco smoking and disability progression in multiple sclerosis: United Kingdom cohort study. Brain 2013; 136: 2298–2304. 44.  Grant BF, Harford TC. Comorbidity between DSM-IV alcohol use disorders and major depression: results of a national survey. Drug Alcohol Depend 1995; 39: 197–206. 45.  Pakpoor J, Goldacre R, Disanto G, et al. Alcohol misuse disorders and multiple sclerosis risk. JAMA Neurol 2014; 71: 1088–1189. 46.  Hedström AK, Hillert J, Olsson T, et al. Alcohol as a modifiable lifestyle factor affecting multiple sclerosis risk. JAMA Neurol 2014; 71: 300–5. 47.  Massa J, O’Reilly EJ, Munger KL, et al. Caffeine and alcohol intakes have no association  111  with risk of multiple sclerosis. Mult Scler 2012; 19: 53–58. 48.  Dimatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of anxiety and depression on patient adherence. Arch Intern Med 2000; 160: 2101–2107. 49.  World Health Organization. Adherence to long-term therapies: evidence for action. http://www.who.int/chp/knowledge/publications/adherence_report/en/ (2003). 50.  Osterberg L, Blaschke T. Adherence to medication. N Engl J Med 2005; 353: 487–497. 51.  Cramer JA, Roy A, Burrell A, et al. Medication compliance and persistence: Terminology and definitions. Value Heal 2008; 11: 44–47. 52.  Filippini G, Del Giovane C, Vacchi L, et al. Immunomodulators and immunosuppressants for multiple sclerosis: a network meta-analysis. Cochrane Libr 2013; 6: 1–134. 53.  IFNβ Multiple Sclerosis Study Group I. Interferon beta-1b is effective in relapsing-remitting multiple sclerosis I. Clinical results of a multicenter, randomized, double-blind, placebo-controlled trial. Neurology 1993; 43: 655–55. 54.  Ebers GC, PRISMS (Prevention of Relapses and Disability by Interferon beta-1a Subcutaneously in Multiple Sclerosis) Study Group. Randomised double-blind placebo-controlled study of interferon beta-1a in relapsing-remitting multiple sclerosis. Lancet 1998; 352: 1498–1504. 55.  Jacobs L, Beck R, Simon J, et al. Intramuscular interferon beta-1a therapy initiated during a first demyelinating event in multiple sclerosis. N Engl J Med 2000; 343: 898–904. 56.  Johnson KP, Brooks BR, Cohen JA, et al. Copolymer 1 reduces relapse rate and improves disability in relapsing-remitting multiple sclerosis. Neurology 1995; 45: 1268–1276. 57.  Menzin J, Caon C, Nichols C, et al. Narrative review of the literature on adherence to disease-modifying therapies among patients with multiple sclerosis. J Manag Care Pharm 2013; 19: S24–S40. 58.  McKay KA, Tremlett H, Patten SB, et al. Determinants of non-adherence to disease-modifying therapies in multiple sclerosis: A cross-Canada prospective study. Mult Scler 2017; 23: 588–596. 59.  Devonshire V, Lapierre Y, Macdonell R, et al. The Global Adherence Project (GAP): A multicenter observational study on adherence to disease-modifying therapies in patients with relapsing-remitting multiple sclerosis. Eur J Neurol 2011; 18: 69–77. 60.  Tremlett H, Van der Mei I, Pittas F, et al. Adherence to the immunomodulatory drugs for multiple sclerosis: contrasting factors affect stopping drug and missing doses. Pharmacoepidemiol Drug Saf 2008; 17: 565–576. 61.  Tan H, Cai Q, Stephenson JJ, et al. Impact of adherence to disease-modifying therapies on  112  clinical and economic outcomes among patients with multiple sclerosis. Adv Ther 2011; 28: 51–61. 62.  Steinberg SC, Faris RJ, Chang CF, et al. Impact of adherence to interferons in the treatment of multiple sclerosis: A non-experimental, retrospective, cohort study. Clin Drug Investig 2010; 30: 89–100. 63.  Wang J, Patten SB. Alcohol consumption and major depression: findings from a follow-up study. Can J Psychiatry 2001; 46: 632–638. 64.  Morrell HER, Cohen LM. Cigarette Smoking, Anxiety, and Depression. J Psychopathol Behav Assess 2006; 28: 281–295. 65.  Goldman Consensus Group. The Goldman Consensus statement on depression in multiple sclerosis. Mult Scler 2005; 11: 328–337. 66.  Bruce JM, Hancock LM, Arnett P, et al. Treatment adherence in multiple sclerosis: association with emotional status, personality, and cognition. J Behav Med 2010; 33: 219–227. 67.  Turner AP, Williams RM, Sloan AP, et al. Injection anxiety remains a long-term barrier to medication adherence in multiple sclerosis. Rehabil Psychol 2009; 54: 116–121. 68.  Mohr DC, Boudewyn AC, Likosky W, et al. Injectable medication for the treatment of multiple sclerosis: the influence of self-efficacy expectations and injection anxiety on adherence and ability to self-inject. Ann Behav Med 2001; 23: 125–132. 69.  Evans C, Marrie RA, Zhu F, et al. Adherence and persistence to drug therapies for multiple sclerosis: A population-based study. Mult Scler Relat Disord 2016; 8: 78–85. 70.  Ivanova JI, Bergman RE, Birnbaum HG, et al. Impact of medication adherence to disease-modifying drugs on severe relapse, and direct and indirect costs among employees with multiple sclerosis in the US. J Med Econ 2012; 15: 601–609. 71.  Bosworth HB, Oddone EZ, Weinberger M (eds). Patient treatment adherence: concepts, interventions, and measurement. In: Medication Treatment Adherence. Mahwah, NJ: Psychology Press, 2006, pp. 147–194. 72.  Amato MP, Ponziani G, Rossi F, et al. Quality of life in multiple sclerosis: the impact of depression, fatigue and disability. Mult Scler 2001; 7: 340–344. 73.  Berrigan LI, Fisk JD, Patten SB, et al. Health-related quality of life in multiple sclerosis Direct and indirect effects of comorbidity. Neurology 2016; 86: 1–8. 74.  Tremlett H, Zhao Y, Rieckmann P, et al. New perspectives in the natural history of multiple sclerosis. Neurology 2010; 74: 2004–15. 75.  Ensari I, Adamson BC, Motl RW. Longitudinal association between depressive symptoms and walking impairment in people with relapsing-remitting multiple sclerosis. J Health  113  Psychol 2016; 21: 2732–2741. 76.  D’Netto MJ, Ward H, Morrison KM, et al. Risk alleles for multiple sclerosis in multiplex families. Neurology 2009; 72: 1984–1988. 77.  Confavreux C, Vukusic S, Moreau T, et al. Relapses and progression of disability in multiple sclerosis. N Engl J Med 2000; 343: 1430–1438. 78.  Hillert J, Stawiarz L. The Swedish MS registry - Clinical support tool and scientific resource. Acta Neurol Scand 2015; 132: 11–19. 79.  Leray E, Morrissey SP, Yaouanq J, et al. Long-term survival of patients with multiple sclerosis in West France. Mult Scler 2007; 13: 865–874. 80.  Myhr KM, Grytten N, Aarseth JH, et al. The Norwegian Multiple Sclerosis National Competence Centre and National Multiple Sclerosis registry - A resource for clinical practice and research. Acta Neurol Scand 2006; 113: 37–40. 81.  Sayao A-L, Devonshire V, Tremlett H. Longitudinal follow-up of benign multiple sclerosis at 20 years. Neurology 2007; 68: 496–500. 82.  Kalincik T, Cutter G, Spelman T, et al. Defining reliable disability outcomes in multiple sclerosis. Brain 2015; 1–12. 83.  Hurwitz BJ. Registry studies of long-term multiple sclerosis outcomes: Description of key registries. Neurology 2011; 76: S3–S6. 84.  Debouverie M, Laforest L, Van Ganse E, et al. Earlier disability of the patients followed in Multiple Sclerosis centers compared to outpatients. Mult Scler 2009; 15: 251–257. 85.  Poser CM, Paty DW, Scheinberg L, et al. New diagnostic criteria for multiple sclerosis: Guidelines for research protocols. Ann Neurol 1983; 13: 227–231. 86.  McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann Neurol 2001; 50: 121–7. 87.  Polman CH, Reingold SC, Edan G, et al. Diagnostic Criteria for Multiple Sclerosis: 2005 Revisions to the ‘McDonald Criteria’. Ann Neurol 2005; 58: 840–6. 88.  Lublin FD, Reingold SC. Defining the clinical course of multiple sclerosis: Results of an international survey. Neurology 1996; 46: 907–11. 89.  Benito-León J, Martínez-Martín P, Frades B, et al. Impact of fatigue in multiple sclerosis: the Fatigue Impact Scale for Daily Use (D-FIS). Mult Scler J 2007; 13: 645–51. 90.  Ewing JA. Detecting Alcoholism: The CAGE Questionnaire. J Am Med Assoc 1984; 252: 1905–7.  114  91.  Horton M, Rudick R a., Hara-Cleaver C, et al. Validation of a self-report comorbidity questionnaire for multiple sclerosis. Neuroepidemiology 2010; 35: 83–90. 92.  Government of Canada. Canada’s health care system. Canada. https://www.canada.ca/en/health-canada/services/canada-health-care-system.html (2016, accessed 15 February 2017). 93.  British Columbia Ministry of Health [creator] (2013). Medical Services Plan (MSP) Payment Information File. V2. Population Data BC [publisher]. Data Extract. MOH (2013) http://www.popdata.bc.ca/data. 94.  World Health Organization. International Classification of Diseases (ICD). World Health Organization. www.who.int/classifications/icd/en/ (2016, accessed 15 February 2017). 95.  Canadian Institute for Health Information [creator] (2013). Discharge Abstract Database (Hospital Separations). V2. Population Data BC. Data Extract. MOH (2013). http://www.popdata.bc.ca/data. 96.  British Columbia Ministry of Health [creator] (2013). PharmaNet. V2. BC Ministry of Health [publisher]. Data Extract. PSBC (2013). http://www.popdata.bc.ca/data). 97.  British Columbia Ministry of Health [creator] (2013). Consolidation File (MSP Registration & Premium Billing). V2. Population Data BC [publisher]; Data Extract. MOH (2013). www.popdatabc.ca/data. 98.  BC Vital Statistics Agency [creator]. Vital Statistic Deaths.V2. Population Data BC [publisher]. Data Extract. BC Vital Statistics Agency (2013) www.popdatabc.ca/data. 99.  Wilkins R. PCCF+ Version 5E User’s Guide. Automated Geographic Coding Based on the Statistics Canada Postal Code Conversion Files, Including Postal Codes through March 2009. Catalogue 82F0086-XDB. Statistics Canada. 100.  Rothman KJ, Greenland S, Lash TL. Validity in epidemiologic studies. In: Rothman KJ, Greenland S, Lash TL (eds) Modern Epidemiology. Philadelphia, PA US, 2008, pp. 128–147. 101.  Pagano M, Gauvreau K. Principles of biostatistics. 7th ed. Pacific Grove, CA: Duxbury, 2000. 102.  Twisk J. Applied Longitudinal Data Analysis for Epidemiology: A Practical Guide. New York, NY: Cambridge University Press, 2003. 103.  Hardin J, Hilbe J. Generalized estimating equations. New York, NY: Chapman & Hall/CRC, 2003. 104.  Arnson Y, Shoenfeld Y, Amital H. Effects of tobacco smoke on immunity, inflammation and autoimmunity. J Autoimmun 2010; 34: J258–J265. 105.  Zigmond A, Snaith R. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983; 67.  115  106.  Mykletun A, Stordal E, Dahl AA. Hospital Anxiety and Depression (HAD) scale: factor structure, item analyses and internal consistency in a large population. Br J Psychiatry 2001; 179: 540–544. 107.  WHO. Guidelines for Controlling and Monitoring the Tobacco Epidemic. Geneva: World Health Organization, 1996. 108.  Quesnel S, Feinstein A. Multiple sclerosis and alcohol: a study of problem drinking. Mult Scler 2004; 10: 197–201. 109.  Beier M, D’Orio V, Spat J, et al. Alcohol and substance use in multiple sclerosis. J Neurol Sci 2014; 338: 122–127. 110.  Bombardier CH, Blake KD, Ehde DM, et al. Alcohol and drug abuse among persons with multiple sclerosis. Mult Scler 2004; 10: 35–40. 111.  Kushner MG, Abrams K, Borchardt C. The relationship between anxiety disorders and alcohol use disorders: A review of major perspectives and findings. Clin Psychol Rev 2000; 20: 149–171. 112.  Bulloch A, Lavorato D, Williams J, et al. Alcohol consumption and major depression in the general population: The critical importance of dependence. Depress Anxiety 2012; 29: 1058–1064. 113.  Kessler RC. The epidemiology of dual diagnosis. Biol Psychiatry 2004; 56: 730–737. 114.  Haukkala A, Uutela A, Vartiainen E, et al. Depression and smoking cessation: The role of motivation and self-efficacy. Addict Behav 2000; 25: 311–316. 115.  Taylor G, McNeill A, Girling A, et al. Change in mental health after smoking cessation: systematic review and meta-analysis. Br Med J 2014; 348: 1–22. 116.  Agaku IT, King BA, Dube SR. Current Cigarette Smoking Among Adults — United States , 2005 – 2012. Morb Mortal Wkly Rep 2014; 63: 29–46. 117.  Poulin C, Webster I, Single E. Alcohol disorders in Canada as indicated by the CAGE questionnaire. Can Med Assoc J 1997; 157: 1529–1535. 118.  Giannantoni A, Scivoletto G, Di Stasi SM, et al. Lower urinary tract dysfunction and disability status in patients with multiple sclerosis. Arch Phys Med Rehabil 1999; 80: 437–441. 119.  Gold SM, Irwin MR. Depression and immunity: inflammation and depressive symptoms in multiple sclerosis. Immunol Allergy Alinics North Am 2009; 29: 309–20. 120.  Patten SB, Metz LM, Reimer M a. Biopsychosocial correlates of lifetime major depression in a multiple sclerosis population. Mult Scler 2000; 6: 115–120. 121.  Piccinelli M, Wilkinson G. Gender differences in depression: Critical review. Br J Psychiatry  116  2000; 177: 486–492. 122.  Marrie RA, Patten SB, Tremlett H, et al. Sex differences in comorbidity at diagnosis of multiple sclerosis: A population-based study. Neurology 2016; 86: 1279–1286. 123.  Davis CG, Thake J, Vilhena N. Social desirability biases in self-reported alcohol consumption and harms. Addict Behav 2010; 35: 302–311. 124.  Fleming M, Barry K, Manwell L, et al. Brief physician advice for problem drinkers. J Am Med Assoc 1997; 278: 1059–1060. 125.  Hughes J, Stead L, Cahill K, et al. Antidepressants for smoking cessation (Review). Cochrane Libr 2014; 1–145. 126.  Carver C, Scheier M, Weintraub J. Assesing coping strategies: a theoretically based approach. J Pers Soc Psychol 1989; 56: 267–283. 127.  Fisk JD, Doble SE. Construction and validation of a fatigue impact scale for daily administration ( D-FIS ). Qual Life Res 2002; 11: 263–272. 128.  Horsman J, Furlong W, Feeny D, et al. The Health Utilities Index (HUI ®): concepts, measurement properties and applications. Health Qual Life Outcomes 2003; 1: 54–67. 129.  Fisk JD, Brown MG, Sketris IS, et al. A comparison of health utility measures for the evaluation of multiple sclerosis treatments. J Neurol Neurosurg Psychiatry 2005; 76: 58–63. 130.  Feng Y, Bernier J, Mcintosh C, et al. Validation of disability categories derived from Health Utilities Index Mark 3 scores. Heal Reports 2009; 20: 43–50. 131.  Dunbar-Jacob J, Mortimer-Stephens MK. Treatment adherence in chronic disease. J Clin Epidemiol 2001; 54: S57–S60. 132.  Siegel SD, Turner AP, Haselkorn JK. Adherence to disease-modifying therapies in multiple sclerosis: Does caregiver social support matter? Rehabil Psychol 2008; 53: 73–79. 133.  Halpern R, Agarwal S, Dembek C, et al. Comparison of adherence and persistence among multiple sclerosis patients treated with disease-modifying therapies: A retrospective administrative claims analysis. Patient Prefer Adherence 2011; 5: 73–84. 134.  Turner AP, Kivlahan DR, Sloan AP, et al. Predicting ongoing adherence to disease modifying therapies in multiple sclerosis: utility of the health beliefs model. Mult Scler 2007; 13: 1146–1152. 135.  McKay KA, Tremlett H, Fisk JD, et al. Adverse health behaviours are associated with depression and anxiety in multiple sclerosis: A prospective multisite study. Mult Scler J 2016; 22: 685–693. 136.  Treadaway K, Cutter G, Salter A, et al. Factors that influence adherence with disease-modifying therapy in MS. J Neurol 2009; 256: 568–576.  117  137.  Mohr DC, Goodkin DE, Likosky W, et al. Therapeutic expectations of patients with multiple sclerosis upon initiating interferon beta-1b: Relationship to adherence in treatment. Mult Scler 1996; 2: 222–226. 138.  Bayliss EA, Steiner JF, Fernald DH, et al. Descriptions of barriers to self-care by persons with comorbid chronic diseases. Ann Fam Med 2003; 1: 15–21. 139.  Stirratt MJ, Dunbar-Jacob J, Crane HM, et al. Self-report measures of medication adherence behavior: recommendations on optimal use. Transl Behav Med 2015; 5: 470–482. 140.  Bergvall N, Petrilla A a, Karkare SU, et al. Persistence with and adherence to fingolimod compared with other disease-modifying therapies for the treatment of multiple sclerosis: a retrospective US claims database analysis. J Med Econ 2014; 17: 696–707. 141.  Marrie RA, Yu BN, Leung S, et al. The utility of administrative data for surveillance of comorbidity in multiple sclerosis: a validation study. Neuroepidemiology 2013; 40: 85–92. 142.  Evans C, Zhu F, Kingwell E, et al. Association between beta-interferon exposure and hospital events in multiple sclerosis. Pharmacoepidemiol Drug Saf 2014; 12–15. 143.  Kleinman NL, Beren IA, Rajagopalan K, et al. Medication adherence with disease modifying treatments for multiple sclerosis among US employees. J Med Econ 2010; 13: 633–640. 144.  Landis JR, Koch GG. The Measurement of Observer Agreement for Categorical Data. Biometrics 1977; 33: 159–174. 145.  Byrt T, Bishop J, Carlin JB. Bias, prevalence and kappa. J Clin Epidemiol 1993; 46: 423–429. 146.  Kwon A, Bungay KM, Pei Y, et al. Antidepressant use: concordance between self-report and claims records. Med Care 2003; 41: 368–374. 147.  Wu C-S, Lai M-S, Gau SS-F, et al. Concordance between patient self-reports and claims data on clinical diagnoses, medication use, and health system utilization in Taiwan. PLoS One 2014; 9: e112257. 148.  Adams AS, Soumerai SB, Lomas J, et al. Evidence of self-report bias in assessing adherence to guidelines. Int J Qual Heal Care 1999; 11: 187–192. 149.  Wallihan D, Stump T, Callahan C. Accuracy of self-reported health services use and patterns of care among urban older adults. Med Care 1999; 37: 662–670. 150.  Wang PS, Benner JS, Glynn RJ, et al. How well do patients report noncompliance with antihypertensive medications?: A comparison of self-report versus filled prescriptions. Pharmacoepidemiol Drug Saf 2004; 13: 11–19. 151.  Lafata JE, Cerghet M, Dobie E, et al. Measuring adherence and persistence to disease-modifying agents among patients with relapsing remitting multiple sclerosis. J Am Pharm Assoc 2008; 48: 752–757.  118  152.  Feinstein AR. On white-coat effects and the electronic monitoring of compliance. Arch Intern Med 1990; 150: 1377–1388. 153.  Heesen C, Böhm J, Reich C, et al. Patient perception of bodily functions in multiple sclerosis: gait and visual function are the most valuable. Mult Scler J 2008; 14: 988–991. 154.  Rothwell PM, McDowell Z, Wong CK, et al. Doctors and patients don’t agree: cross-sectional study of patients’ and doctors’ perceptions and assessments of disability in multiple sclerosis. BMJ 1997; 314: 1580. 155.  World Health Organization Collaborating Centre for Drug Statistics Methodology. ATC Classification Index with DDDs. https://www.whocc.no/atc_ddd_index/ (2017, accessed 15 Feb 2017). 156.  Marrie RA, Fisk JD, Yu BN, et al. Mental comorbidity and multiple sclerosis: validating administrative data to support population-based surveillance. BMC Neurol 2013; 13: 1–8. 157.  Marrie RA, Rudick R, Horwitz R, et al. Vascular comorbidity is associated with more rapid disability progression in multiple sclerosis. Neurology 2010; 74: 1041–1047. 158.  Marrie RA, Yu BN, Leung S, et al. Prevalence and incidence of ischemic heart disease in multiple sclerosis: A population-based validation study. Mult Scler Relat Disord 2013; 2: 355–361. 159.  Marrie RA, Yu BN, Leung S, et al. Rising prevalence of vascular comorbidities in multiple sclerosis: validation of administrative definitions for diabetes, hypertension, and hyperlipidemia. Mult Scler 2012; 18: 1310–1319. 160.  Gilchrist AC, Creed FH. Depression, cognitive impairment and social stress in multiple sclerosis. J Psychosom Res 1994; 38: 193–201. 161.  Benito-León J, Morales JM, Rivera-Navarro J. Health-related quality of life and its relationship to cognitive and emotional functioning in multiple sclerosis patients. Eur J Neurol 2002; 9: 497–502. 162.  Taylor KL, Hadgkiss EJ, Jelinek GA, et al. Lifestyle factors, demographics and medications associated with depression risk in an international sample of people with multiple sclerosis. BMC Psychiatry 2014; 14: 1–12. 163.  Marrie RA, Elliott L, Marriott J, et al. Effect of comorbidity on mortality in multiple sclerosis. Neurology 2015; 85: 240–247. 164.  Chwastiak L, Ehde DM, Gibbons LE, et al. Depressive symptoms and severity of illness in multiple sclerosis: epidemiologic study of a large community sample. Am J Psychiatry 2002; 159: 1862–8. 165.  Haase CG, Lienemann M, Faustmann PM. Neuropsychological deficits but not coping strategies are related to physical disability in multiple sclerosis. Eur Arch Psychiatry Clin Neurosci 2008; 258: 35–39.  119  166.  Marrie RA, Horwitz R, Cutter G, et al. Comorbidity delays diagnosis and increases disability at diagnosis in MS. Neurology 2009; 72: 117–24. 167.  Patten SB, Williams JVA, Lavorato DH, et al. Disability in a community population with MS with and without mental disorders. Int J Psychiatry Med 2012; 43: 51–65. 168.  Koch M, Uyttenboogaart M, van Harten  a, et al. Fatigue, depression and progression in multiple sclerosis. Mult Scler 2008; 14: 815–22. 169.  Feinstein A, Roy P, Lobaugh N, et al. Structural brain abnormalities in multiple sclerosis patients with major depression. Neurology 2004; 62: 586–90. 170.  Dantzer R, O’Connor JC, Freund GG, et al. From inflammation to sickness and depression: when the immune system subjugates the brain. Nat Rev Neurosci 2008; 9: 46–56. 171.  Altamura AC, Buoli M, Pozzoli S. Role of immunological factors in the pathophysiology and diagnosis of bipolar disorder: Comparison with schizophrenia. Psychiatry Clin Neurosci 2014; 68: 21–36. 172.  Frischer JM, Bramow S, Dal-Bianco A, et al. The relation between inflammation and neurodegeneration in multiple sclerosis brains. Brain 2009; 132: 1175–1189. 173.  Oliver MI, Pearson N, Coe N, et al. Help-seeking behaviour in men and women with common mental health problems: Cross-sectional study. Br J Psychiatry 2005; 186: 297–301. 174.  Amato MP, Portaccio E. Clinical outcome measures in multiple sclerosis. J Neurol Sci 2007; 259: 118–122. 175.  Marrie RA, Yu N, Blanchard J, et al. The rising prevalence and changing age distribution of multiple sclerosis in Manitoba. Neurology 2010; 74: 465–471. 176.  Marrie RA, Fisk JD, Stadnyk KJ, et al. The incidence and prevalence of multiple sclerosis in Nova Scotia, Canada. Can J Neurol Sci 2013; 40: 824–31. 177.  Schneeweiss S, Seeger JD, Maclure M, et al. Performance of comorbidity scores to control for confounding in epidemiologic studies using claims data. Am J Epidemiol 2001; 154: 854–864. 178.  Sketris I, Hick V, Brown M, et al. Multiple sclerosis disease-modifying drug utilization and cost patterns following introduction of a provincially funded program in Nova Scotia Canada 1998-2004. J Appl Ther Res 2011; 8: 65–78. 179.  Kobelt G, Lindgren P, Parkin D, et al. Costs and quality of life in multiple sclerosis: A cross-sectional observational study in the UK. SSE/EFI Work Pap Ser Econ Financ; 398. 180.  Esbjerg S, Keiding N, Koch-Henriksen N. Reporting delay and corrected incidence of multiple sclerosis. Stat Med 1999; 18: 1691–1706. 181.  Alexandra P-T, Calderon-Larranaga A, Hancco-Saavedra J, et al. Multimorbidity patterns: a  120  systematic review. J Clin Epidemiol 2014; 67: 254–266. 182.  Widdifield J, Ivers NM, Young J, et al. Development and validation of an administrative data algorithm to estimate the disease burden and epidemiology of multiple sclerosis in Ontario, Canada. Mult Scler 2015; 21: 1045–1054. 183.  D’hooghe MB, Haentjens P, Nagels G, et al. Alcohol, coffee, fish, smoking and disease progression in multiple sclerosis. Eur J Neurol 2012; 19: 616–624. 184.  McKay KA, Tremlett H, Patten SB, et al. Determinants of non-adherence to disease-modifying therapies in multiple sclerosis: A cross-Canada prospective study. Mult Scler 2016; 1–9. 185.  McKay KA, Tremlett H, Zhu F, et al. A population-based study comparing multiple sclerosis clinic users and non-users in British Columbia, Canada. Eur J Neurol 2016; 23: 1093–100. 186.  Goretti B, Portaccio E, Zipoli V, et al. Coping strategies, cognitive impairment, psychological variables and their relationship with quality of life in multiple sclerosis. Neurol Sci 2010; 31: S227-230. 187.  Zorzon M, Zivadinov R, Nasuelli D, et al. Depressive symptoms and MRI changes in multiple sclerosis. Eur J Neurol 2002; 9: 491–496. 188.  Berg D, Supprian T, Thomae J, et al. Lesion pattern in patients with multiple sclerosis and depression. Mult Scler 2000; 6: 156–162. 189.  Roos LL, Gupta S, Soodeen R-A, et al. Data quality in an information-rich environment: Canada as an example. Can J Aging 2005; 24: 153–170. 190.  Lunsford TR, Lunsford BR. The Research Sample, Part I: Sampling. J Prosthetics Orthot 1995; 7: 105–112. 191.  Kristman V, Manno M, Côté P. Loss to Follow-Up in Cohort Studies : How Much Is Too Much? Eur J Epidemiol 2004; 19: 751–760. 192.  Marrie RA, Patten SB, Berrigan LI, et al. Diagnoses of depression and anxiety versus current symptoms and quality of life in MS. Int J MS Care; In Press. 193.  Bradley K a, Bush KR, McDonell MB, et al. Screening for problem drinking : Comparison of CAGE and AUDIT. J Gen Intern Med 1998; 13: 379–88. 194.  Marrie RA, Cutter G, Tyry T, et al. Smoking status over two years in patients with multiple sclerosis. Neuroepidemiology 2008; 32: 72–79. 195.  Vartiainen E, Seppälä T, Lillsunde P, et al. Validation of self reported smoking by serum cotinine measurement in a community-based study. J Epidemiol Community Health 2002; 56: 167–70. 196.  Holmes Rovner M, French M, Sofaer S, et al. A new definition of patient engagement: What  121  is engagement and why is it important? Washington. http://www.cfah.org/pdfs/CFAH%7B_%7DEngagement%7B_%7DBehavior%7B_%7DFramework%7B_%7Dcurrent.pdf (2010). 197.  Thombs BD, Ziegelstein RC. Does depression screening improve depression outcomes in primary care? BMJ Br Med J 2014; 348: Online. 198.  Fiest KM, Walker JR, Bernstein CN, et al. Systematic review and meta-analysis of interventions for depression and anxiety in persons with multiple sclerosis. Mult Scler Relat Disord 2016; 5: 12–26. 199.  Mohr DC, Goodkin DE. Treatment of depression in multiple sclerosis: Review and meta-analysis. Clin Psychol Sci Pract 1999; 6: 1–9.   122  Appendices  Appendix A  Supplementary Figures and Tables Expanded Disability Status Scale (EDSS) Score  Definition  0  Normal Neurological Exam  1.0  No disability, minimal signs on 1 FS  1.5  No disability, minimal signs on 2 of 7 FS  2.0  Minimal disability in 1 of 7 FS  2.5  Minimal disability in 2 FS  3.0  Moderate disability in 1 FS; or mild disability in 3 - 4 FS, though fully ambulatory  3.5  Fully ambulatory but with moderate disability in 1 FS and mild disability in 1 or 2 FS; or moderate disability in 2 FS; or mild disability in 5 FS  4.0  Fully ambulatory without aid, up and about 12hrs a day despite relatively severe disability. Able to walk without aid 500 meters  4.5  Fully ambulatory without aid, up and about much of day, able to work a full day, may otherwise have some limitations of full activity or require minimal assistance. Relatively severe disability. Able to walk without aid 300 meters  5.0  Ambulatory without aid for about 200 meters. Disability impairs full daily activities  5.5  Ambulatory for 100 meters, disability precludes full daily activities  6.0  Intermittent or unilateral constant assistance (cane, crutch or brace) required to walk 100 meters with or without resting  6.5  Constant bilateral support (cane, crutch or braces) required to walk 20 meters without resting  7.0  Unable to walk beyond 5 meters even with aid, essentially restricted to wheelchair, wheels self, transfers alone; active in wheelchair about 12 hours a day  7.5  Unable to take more than a few steps, restricted to wheelchair, may need aid to transfer; wheels self, but may require motorized chair for full day's activities  8.0  Essentially restricted to bed, chair, or wheelchair, but may be out of bed much of day; retains self care functions, generally effective use of arms  8.5  Essentially restricted to bed much of day, some effective use of arms, retains some self care functions  9.0  Helpless bed patient, can communicate and eat  9.5  Unable to communicate effectively or eat/swallow  10.0  Death due to MS  Table A.1 Definitions of Expanded Disability Status Scale scores.18    123   Figure A.1 Schematic of data linkages including health administrative and clinical databases.                      124    Number of doses required in order to fulfill the definition of: Disease-modifying therapy (route and frequency)  Expected number of doses in 30 days ‘Adherent’ (i.e, ≥  80% of expected doses taken) ‘Non-adherent’ (i.e, < 80% of expected doses taken) Interferon β-1a (intramuscular, weekly) 4 4 < 4 Interferon β-1b   (subcutaneous, every other day) 15 ≥ 12 < 12 Interferon β-1a (subcutaneous, three times per week) 12 ≥ 10 < 10 Glatiramer acetate (subcutaneous, daily)  30 ≥ 24 < 24 Table A.2 Definitions of adherent and non-adherent based on the medication possession ratio of 80% for the four disease-modifying therapy (DMT) types. Key: Each number shown in the table represents the number of doses of injectable DMTs during a 30-day period. The ‘expected number’ for each DMT were derived from the relevant product monograph and represent the full, licensed dosing schedule for each drug.                              125  Psychiatric Comorbidity ICD -9 codes ICD-10 codes ATC system classification codes Algorithm156 Mood or anxiety disorder (omnibus definition) 300.0, 300.2, 296.0, 296.1, 296.04, 296.14, 296.4, 296.44, 296.5, 296.54, 296.6, 296.7, 296.8, 296.2, 296.3, 298.0, 300.4, 50B F40, F41, F31, F32, F33, F34 N06AA01, N06AA02, N06AA04, N06AA11, N06AA12, N06AA17, N06AA21, N06AB03, N06AB04, N06AB05, N06AB06, N06AB08, N06AB10, N06AF03, N06AF04, N06AG02, N06AX06, N06AX11, N06AX16, N06AX21, N06AX23, N05BA12, N05BA06, N05AN01, N03AF01, N03AG01, N03AX09 ≥1 Hospitalization or ≥5 Physician visits or (≥ 1 physician visit and ≥4 prescriptions dispensed) in 2 years  Depression 296.2, 296.3, 298.0, 300.4, 311  F32, F33, F34  N06AA01, N06AA02, N06AA04, N06AA11, N06AA12, N06AA17, N06AA21, N06AB03, N06AB04, N06AB05, N06AB06, N06AB08, N06AB10, N06AF03, N06AF04, N06AG02, N06AX06, N06AX11, N06AX16, N06AX21, N06AX23 ≥1 Hospitalization or ≥5 Physician visits or (≥ 1 physician visit and ≥7 prescriptions dispensed) in 2 years Anxiety disorders  300.0, 300.2  F40, F41 N05BA12 N05BA06 ≥1 Hospitalization or ≥2 Physician visits or (≥1 physician visit and ≥2 prescriptions dispensed) in 2 years Bipolar disorder 296.0, 296.1, 296.04, 296.14, 296.4, 296.44, 296.5, 296.54, 296.6, 296.7, 296.8 F31 N05AN01, N03AF01, N03AG01, N03AX09  ≥1 Hospitalization or ≥3 Physician visits or (≥1 physician visit and ≥3 prescriptions dispensed) in 2 years Table A.3 Diagnosis (ICD-9 and ICD-10) codes and algorithms used to identify psychiatric comorbidities.          126  Comorbidity ICD-9 Code ICD-10 Code Algorithm 141,156,159 Chronic Lung Disease 493, 491, 492, 496 J40, J42, J43, J44, J45, J46 ≥ 1 H or ≥ 2 P in 5 years Diabetes 250 E10, E11, E12, E13, E14 ≥ 1 H or ≥ 1 P in 3 years Epilepsy 345 G40, G41 ≥ 1 Hospitalization or ≥ 1 Physician visits in 3 years Heart Disease 410, 411, 412, 413, 414 I20, I21, I22, I23, I24, I25 ≥ 1 Hospitalization or ≥ 2 Physician visits in 5 years Hyperlipidemia 272 E780, E782, E784, E785 ≥ 1 H or ≥ 2 P in 4 years Hypertension 401, 402, 403, 404, 405 I10, I11, I12, I13, I15 ≥ 1 H or ≥ 2 P in 2 years  Migraine 345, 625.4 G43 ≥ 1 H or ≥ 2 P in 4 years Mood or Anxiety Disorder* 296, 298, 300, 311, 50B F40, F41, F31, F25, F32, F33, F34 ≥ 1 H or ≥ 5 P in 5 years Table A.4 Diagnosis (ICD-9 and ICD-10) codes and algorithms used to identify comorbidities from administrative data.  * Definition used for Chapter 6 only. Algorithm used in Chapter 5 is outlined in Table A.3.    Demyelinating disease of the CNS ICD-9 Code ICD-10 Code Multiple Sclerosis 340 G35 Optic Neuritis 377.3 H46 Acute transverse myelitis 341.2 G37 Acute disseminated encephalomyelitis 323 G36.9 Demyelinating disease of the CNS, unspecified 341.9 G37.8 Other acute disseminated demyelination NA G36.8 Neuromyelitis optica 341.0 G36.0 Table A.5 Diagnosis (ICD-9 and ICD-10) codes used to identify MS and demyelinating diseases of the CNS from administrative data.        127  Appendix B  Informed Consent Form  SUBJECT INFORMATION AND CONSENT FORM The Epidemiology and Impact of Comorbidity on MS in Canada (ECoMS): Characterization of Comorbidities in MS  Principal Investigator: Dr. Helen Tremlett, PhD Department of Medicine (Neurology) Rm S178 – 2211 Wesbrook Mall University of British Columbia Vancouver, BC V6T 2B5 Canada Email: Tremlett@interchange.ubc.ca Telephone: 604-822-0759  Fax: 604-822-7131  INTRODUCTION  You are being invited to take part in this research study because you have been identified as a patient with multiple sclerosis attending the UBC MS Clinic.  YOUR PARTICIPATION IS VOLUNTARY Your participation is entirely voluntary. Before you decide whether or not to take part in this study, it is important for you to understand what the research involves. This consent form will tell you about the study, why the research is being done, what will happen to you during the study and the possible benefits, risks and discomforts.  If you wish to participate, you will be asked to sign this form. If you do decide to take part in this study, you are still free to withdraw at any time and without giving any reasons for your decision.  If you do not wish to participate, you do not have to provide any reason for your decision not to participate nor will you lose the benefit of any medical care to which you are entitled or are presently receiving.  WHO IS CONDUCTING THE STUDY? This study is being conducted by members of the University of British Columbia, Division of Neurology who are affiliated with the UBC MS Clinic, as part of a larger national multi-centre study initiative funded by the Canadian Institutes of Health Research (CIHR).  128  BACKGROUND Multiple Sclerosis (MS) is a chronic neurological disease affecting the nervous system. MS exemplifies many chronic neurological diseases; it is incurable, affects the brain and spinal cord and causes substantial morbidity and mortality.  Canada has among the highest prevalence of MS in the world. MS is a physically and mentally disabling disease with vast unexplained variability in outcomes and adversely affects employment, social relationships and quality of life.  Health related quality of life (HRQoL) refers to the functional impact of a health condition on physical and mental well being from the perspective of the affected individual. Persons with MS report lower HRQoL as compared to general and other chronic disease populations with factors such as age, sex, socioeconomic status, and disability status know to exert effects on HRQoL in MS. While the presence of a co-existing health condition (i.e. comorbidity) is known to be associated with reduced HRQoL in other chronic disease, little is known about its impact in MS. Identifying potentially modifiable comorbidities which affect HRQoL can create an avenue for improving HRQoL as many of the comorbidities we will evaluate can be managed successfully.  WHAT IS THE PURPOSE OF THE STUDY? The purpose of this study is to determine how comorbidity affects health-related quality of life (HRQoL) in MS.  WHO CAN PARTICIPATE IN THE STUDY? BC-resident patients identified to have definite MS and are 18 years or older will be invited to participate. Patients must be able to provide informed consent, as well as have adequate knowledge of the English language so as to complete the questionnaires administered for this study.  WHO SHOULD NOT PARTICIPATE IN THE STUDY? Patients unable or unwilling to commit to completing questionnaires at yearly intervals over three years will be ineligible to participate in this study.  WHAT DOES THE STUDY INVOLVE? This study is a national multi-centre study taking place in various MS Clinics in Canada. The University of British Columbia site will conduct the study at the UBC MS Clinic located at the UBC Hospital in Vancouver. If you agree to participate, you will be asked to fill out a questionnaire. You will fill out the questionnaire at the UBC MS Clinic during your usual clinic visit.  It will take about 15-20 minutes to complete and will include questions about your quality  129  of life, fatigue, medical history, use of alcohol and mood. You will be asked to fill out the questionnaire again in 12 months, and again in 24 months when you are here for your usual clinic visits. If you cannot attend those clinic visits we will mail you the questionnaire to complete at home.  Participation in the study will end when you finish the third questionnaire.  We will review your medical records to learn more about your MS, MS-related treatment and your other health conditions.  We are also seeking your permission to combine (link) the information collected from your questionnaires to information collected by BC Ministry of Health (administrative data) at a later date. Administrative data includes information on such things as hospitalizations, hospital visits, physician visits and medication use related to MS and co-existing health conditions.  WHAT ARE MY RESPONSIBILITIES? If you choose to participate in this study, you will be asked to complete an initial questionnaire at the time of your annual clinic visit as a baseline. After 12 months, you will be asked to fill out the questionnaire again, as well as after 24 months. For your convenience, the questionnaire can be completed at each of your annual clinic visits; however, if you are unable to attend your clinic visit, we will mail you the questionnaires to complete at home.  WHAT ARE THE POSSIBLE HARMS AND SIDE EFFECTS OF PARTICIPATING? There are no physical risks associated with this study. There is, however, the potential risk of loss of confidentiality. Every possible effort will be made to keep your information confidential, however, absolute confidentiality cannot be guaranteed. Some of the questions we ask you as part of this study may make you feel uncomfortable. You may refuse to answer any of the questions and you may take a break at any time during the study. You may stop your participation in this study at any time.  WHAT ARE THE BENEFITS OF PARTICIPATING IN THIS STUDY? There are no direct benefits to you from participating in this study. We hope the information learned from this study will benefit other people with multiple sclerosis in the future.  WHAT HAPPENS IF I DECIDE TO WITHDRAW MY CONSENT TO PARTICIPATE? Your participation in this research is entirely voluntary. You may withdraw at any time. If you decide to enter the study and to withdraw at any time in the future, there will be no penalty or loss of benefits to which you are otherwise entitled, and your future medical care  130  will not be affected. If you choose to enter the study and then decide to withdraw at a later time, all data collected about you during your enrolment in the study will be retained for analysis. By law, this data cannot be destroyed.  WHAT HAPPENS IF SOMETHING GOES WRONG? Signing this consent form in no way limits your legal rights against investigators or anyone else.  WHAT WILL THE STUDY COST ME? There is no cost to participating in this study. If you agree to participate in this study, you will not be paid for participating.  WILL MY TAKING PART IN THIS STUDY BE KEPT CONFIDENTIAL? Your confidentiality will be respect. No information that discloses your identity will be released or published without your specific consent to the disclosure. However, research records identifying you may be inspected in the presence of the investigator or his or her designate by representatives of Health Canada and the UBC Research Ethics Board for the purpose of monitoring the research. However, no records which identify you by name or initials will be allowed to leave the Investigator’s offices. Your records will be stored for at least five years prior to being destroyed.  WHO DO I CONTACT IF I HAVE QUESTIONS ABOUT THE STUDY DURING MY PARTICIPATION?  If you have any questions or desire further information about this study before or during participation, you can contact the Study Investigator, Dr. Helen Tremlett, at 604-822-0759 (or tremlett@interchange.ubc.ca) or the Study Coordinator, Anna-Marie Bueno at 604-822-7880 (or abueno@interchange.ubc.ca).  WHO DO I CONTACT IF I HAVE ANY QUESTIONS OR CONCERNS ABOUT MY RIGHTS AS A SUBJECT DURING THE STUDY?  If you have any concerns about your rights as a research subject and/or your experiences while participating in this study, contact the Research Subject Information Line in the University of British Columbia, Office of Research Services at 604-822-8598 (or RSIL@ors.ubc.ca).  SUBJECT CONSENT TO PARTICIPATE By signing below, I acknowledge that the following is true:  131    I have read and understood the subject information and consent form.  I understand that all of the information collected will be kept confidential and that the result will only be used for scientific objectives.  I understand that my participation in this study is voluntary and that I am completely free to refuse to participate or to withdraw from this study at any time without changing in any way the quality of care that I receive.  I understand that I am not waiving any of my legal rights as a result of signing this consent form.  I have read this form and I freely consent to participate in this study.  I have been told that I will receive a dated and signed copy of this form.  Signatures      Printed name of subject  Signature  Date Printed name of witness  Signature  Date Printed name of Principal Investigator/ Designated Representative  Signature  Date          132  Appendix C  Self-report questionnaire  The Influence of Comorbidity on Health Related Quality of Life in Multiple Sclerosis  The following questions relate to your treatments for multiple sclerosis.  1. If you are currently taking a disease-modifying therapy please mark which therapy you are taking with a √.   i. Avonex  ii. Betaseron  iii. Rebif  iv. Copaxone  v. Tysabri  If you are not taking a disease-modifying therapy, please place a √ in this box and then skip to question # 3:   □  2. If you are currently taking a disease-modifying therapy please indicate how many injections you missed in last 30 days: _____________  The following questions are about your other health conditions   Here is an example of how to complete the table on the next page. Imagine that you are a 50 year old with MS. You have high blood pressure (hypertension) that was diagnosed in 2002. You are taking medication for your high blood pressure and it is under good control. You do not have diabetes or high cholesterol. You would fill this portion of the table out like this:  Condition NO  YES   Year diagnosed   Currently treated?  NO    YES High cholesterol (hyperlipidemia) X      High blood pressure (hypertension)        X 2002  X Diabetes X        133  3. Has a doctor ever told you that you have any of the following conditions?      For each condition please mark NO or YES.   If you do NOT have the problem, skip to the next problem.   If you do have the problem, please write the year you were diagnosed in the second column. In the third column please indicate if you take medication or some other treatment for the problem.   Condition NO  YES   Year diagnosed   Currently treated?    NO     YES    High cholesterol (hyperlipidemia)       High blood pressure (hypertension)            Heart trouble (such as angina, congestive heart failure, or coronary artery disease)       Disease of arteries in the legs (peripheral vascular disease)       Lung trouble (asthma, emphysema, chronic bronchitis, or COPD)       Diabetes mellitus       Glaucoma       Cataracts       Migraine       Thyroid disease (such as Graves’ disease, Hashimoto’s thyroiditis; not thyroid cancer)       Lupus (systemic lupus erythematosus, SLE)        Inflammatory bowel disease (Crohn’s disease, ulcerative colitis)       Rheumatoid arthritis        134  Osteoporosis (bone disease causing thinning of the  bones)       Fibromyalgia       Irritable bowel syndrome       Depression        Anxiety disorder       Bipolar disorder (manic depression)       Schizophrenia         The following questions are about behaviours that can affect your health. 4. Have you smoked at least 100 cigarettes in your ENTIRE LIFE?       1. Yes                     2. No (SKIP TO # 11)  5. How old were you when you FIRST started to smoke fairly regularly?   ____ years  6. Do you NOW smoke:      1. not at all                2. every day (SKIP to #8)       3. some days (SKIP to #8)  7. How old were you when you quit smoking cigarettes?  _____ years  (SKIP to #10)  8. On the average, how many cigarettes do you now smoke a day?  _____ cigarettes   9. On how many of the PAST 30 DAYS did you smoke a cigarette?  _____ days   10. On the average, during the years that you have smoked, about how many cigarettes did you smoke a day? _____ cigarettes  11. For each question please circle NO or YES as appropriate.  A. Have you ever felt you should Cut down on your drinking? NO YES  135   B. Have people Annoyed you by criticizing your drinking?       NO YES  C. Have you ever felt bad or Guilty about your drinking?       NO YES  D. Have you ever had a drink first thing in the morning to steady  your  nerves or to get rid of a hangover (Eye opener)?       NO YES                          136  Appendix D  Briefing Note   137      138   

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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0348724/manifest

Comment

Related Items