UBC Faculty Research and Publications

One size does not fit all in severe infection: obesity alters outcome, susceptibility, treatment, and… Wacharasint, Petch; Boyd, John H; Russell, James A; Walley, Keith R Jun 20, 2013

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

Item Metadata


52383-13054_2013_Article_1901.pdf [ 254.08kB ]
JSON: 52383-1.0223911.json
JSON-LD: 52383-1.0223911-ld.json
RDF/XML (Pretty): 52383-1.0223911-rdf.xml
RDF/JSON: 52383-1.0223911-rdf.json
Turtle: 52383-1.0223911-turtle.txt
N-Triples: 52383-1.0223911-rdf-ntriples.txt
Original Record: 52383-1.0223911-source.json
Full Text

Full Text

RESEARCH Open AccessOne size does not fit all in severe infection:obesity alters outcome, susceptibility, treatment,and inflammatory responsePetch Wacharasint, John H Boyd, James A Russell and Keith R Walley*AbstractIntroduction: Obesity is an increasingly common comorbidity in critically ill patients. Whether obesity alters sepsisoutcome, susceptibility, treatment, and response is not completely understood.Methods: We conducted a retrospective analysis comparing three group of septic shock patients based on theintervals of actual body mass index (BMI) in patients enrolled in the VASST (Vasopressin and Septic Shock Trial)cohort. Primary outcome measurement was 28-day mortality. We tested for differences in patterns of infection bycomparing the primary site of infection and organism. We also compared the treatments (fluids and vasopressors)and inflammatory response, measuring adipose tissue-related cytokine concentrations (interleukin [IL]-6, monocytechemotactic protein [MCP]-1, tumor necrosis factor [TNF]-a, and resistin) in plasma in a subset of 382 patients. Ofthe 778 patients in VASST, 730 patients who had body weight and height measurements were analyzed. Patientswith BMI <25 kg/m2 (n = 276) were grouped as a reference and compared to ‘overweight’ (25< BMI <30 kg/m2,n = 209) and ‘obese’ (BMI ≥30 kg/m2, n = 245) patients.Results: Obese patients had the lowest 28-day mortality followed by overweight patients while patients with BMI<25 kg/m2 had the highest mortality (p = 0.02). Compared to the patients with BMI <25 kg/m2, obese andoverweight patients also had a different pattern of infection with less lung (obese 35%, overweight 45%, BMI<25kg/m2 50%, p = 0.003) and fungal infection (obese 8.2%, overweight 11%, and BMI<25 kg/m2 15.6%, p = 0.03). Perkilogram, obese and overweight patients received less fluid during the first four days (p<0.05) and received lessnorepinephrine (obese 0.14, overweight 0.21, BMI <25 kg/m2 0.26 μg/kg/min, p<0.0001) and vasopressin (obese 0.28,overweight 0.36, BMI <25 kg/m2 0.43 μU/kg/min, p<0.0001) on day 1 compared to patients with BMI <25 kg/m2.Obese and overweight patients also had a lower plasma IL-6 concentration at baseline (obese 106 [IQR 34-686],overweight 190 [IQR 44-2339], BMI <25 kg/m2 235 [IQR 44-1793] pg/mL, p = 0.046).Conclusions: Overall obesity was associated with improved survival in septic shock and differences in pattern ofinfection, fluids, and vasopressors. Importantly, the magnitude of inflammatory IL-6 response is muted in the obese.IntroductionObesity is a major health problem and is the sixth mostimportant risk factor contributing to overall burden of dis-ease worldwide [1]. Acute infection coupled with theimmune response (sepsis) has been linked in multiple waysto obesity [2,3]. Obese patients were over-representedamong those with serious lung injury requiring life supportduring the H1N1 viral pandemic [3,4], and the risk ofwound infections is increased in the obese following majorsurgery [5]. It has also been suggested that obesity itself isa chronic inflammatory condition with increased circulat-ing pro-inflammatory cytokines at rest [6,7]. Despite theapparent risk factors associated with obesity, it appears thatin small case series obese critically ill patients, the majorityof whom have severe infection, may fare better than thoseof normal weight [8,9]. Whether there is a true relationshipbetween obesity and sepsis-induced morbidity and mortal-ity remains unclear [10,11].We conducted a retrospective analysis of the Vasopres-sin and Septic Shock Trial (VASST) [12] to help determine* Correspondence: keith.walley@hli.ubc.caUniversity of British Columbia, Critical Care Research Laboratories, Institutefor Heart + Lung Health, St. Paul’s Hospital, 1081 Burrard Street, Vancouver,BC, Canada V6Z 1Y6Wacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122© 2013 Wacharasint et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.whether being overweight or obese altered (a) mortalityand organ dysfunction outcomes of sepsis, (b) pattern ofsusceptibility to infection, (c) treatment received bypatients, or (d) the inflammatory response to sepsis. Toexplore an inflammatory response of obese and over-weight, compared with lean patients, during the initialphase of septic shock, it is well known that adipose tissuesproduce and release a number of pro-inflammatory media-tors, including cytokines and chemokines, such as inter-leukin-6 (IL-6) [6,13], monocyte chemotactic protein-1(MCP-1) [14], and tumor necrosis factor-alpha (TNF-a)[15], as well as the adipokine resistin [16]. Therefore, weassayed these cytokines at study enrollment in conveni-ence samples of patients with septic shock in our cohort.Materials and methodsPatientsWe conducted a retrospective analysis in patients enrolledin a cohort of septic shock, the VASST. The VASST is amulticenter, randomized, double-blind, controlled trialevaluating the efficacy of vasopressin versus norepinephr-ine on mortality in 778 patients with septic shock [12].Septic shock was defined by the presence of two or morediagnostic criteria for systemic inflammatory response syn-drome (SIRS), proven or suspected infection, at least onenew organ dysfunction by Brussels criteria, and hypoten-sion despite adequate fluid resuscitation [17,18]. Of these,730 patients had body weight and height measured at thetime of enrollment of the VASST (typically about 11hours after they first developed septic shock), and bodymass index (BMI) was calculated as weight (in kilograms)divided by height (in meters) squared. Since our cohorthad a small number of underweight (BMI of less than 18.5kg/m2, n = 26) patients and these patients had similar out-comes in crude analysis compared with patients classifiedas normal BMI (18.5 to 24.9 kg/m2, n = 250), we used amodified form of the BMI categories endorsed by theNational Institutes of Health [19], in which subjects with aBMI of less than 25 kg/m2 were grouped together (n =276) and were compared with 209 patients classified asoverweight (BMI 25 to 29.9 kg/m2) and 245 patients clas-sified as obese (BMI of at least 30 kg/m2). The VASSTprotocol was approved by the research ethics boards of allparticipating centers, and written informed consent wasobtained from all patients or their relatives. The institu-tional review board for the VASST is the University ofBritish Columbia-Providence Health Care (UBC-PHC)Research Ethics Board, and the approval number is H06-50080.Baseline clinical phenotype measurementsBaseline characteristics of the patients were measured atstudy enrollment and included age, gender, and pre-existing conditions identified on the basis of patients’medical history. Hemodynamic variables (mean arterialpressure, central venous pressure (CVP), and heart rate),tidal volume (in patients who received mechanical venti-lation at study enrollment; BMI of less than 25 kg/m2 n =172, overweight n = 143, and obese n = 160), and base-line laboratory variables (lactate, creatinine, white bloodcell count, and platelet count) were also measured.Outcome measurementsOur primary outcome measurement was 28-day mortal-ity. Secondary outcomes were organ dysfunctions as pre-viously described [20]. To assess organ dysfunction andto correct organ dysfunction scoring for deaths in the28-day observation period, we calculated days alive andfree of organ dysfunction. During each 24-hour period(8 a.m. to 8 a.m.) for each variable, days alive and freewas scored as 1 if the patient was alive and free oforgan dysfunction (normal or mild dysfunction usingthe Brussels criteria [18]). Days alive and free was scoredas 0 if the patient had organ dysfunction (moderate orworse) or was not alive. Every day over the 28-dayobservation after intensive care unit (ICU) admissionwas scored in this way. Thus, the lowest score possiblefor each variable was 0 and the highest score possiblewas 28. A low score indicates more organ dysfunction,whereas a high score indicates less organ dysfunction.Pattern of infectionInfection was defined as culture positivity that wasjudged by the attending clinician of each study site tobe an infection rather than just colonization. Pathogenswere identified by the culture from the primary site ofinfection.TreatmentThe amount of infused fluid is associated with hospitaloutcome [21,22]; therefore, we also tested for differencesin the amount of fluid intake, output, and accumulatednet fluid balance during the first four days after ICUadmission (or study enrollment). We also tested for differ-ences in infusion rates of vasopressor medications, which,in this study, were primarily norepinephrine and vasopres-sin. We measured serum vasopressin levels in the over-weight and obese patients receiving vasopressin infusionand compared that vasopressin level with that of patientswith a BMI of less than 25 kg/m2. Of 396 patients in thevasopressin treatment group, 54 patients had measure-ment of serum vasopressin level over time (baseline and 6,24, and 72 hours) and received vasopressin infusion. Ofthese, 46 patients had BMI measurements and were classi-fied as having a BMI of less than 25 kg/m2 (n = 17) orbeing overweight (n = 13) or obese (n = 16) and wereincluded in this analysis. The protocol for serum vasopres-sin level measurement was previously described [12].Wacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122Page 2 of 10Inflammatory cytokine responseWe assayed cytokines that have also been related to adi-pose tissue, including IL-6 [12], MCP-1 [13], TNF-a[14], and resistin [15], in 138 patients with a BMI of lessthan 25 kg/m2, 112 overweight patients, and 132 obesepatients at enrollment. Briefly, whole-blood sampleswere drawn into chilled 7-mL EDTA Vacutainer tubes(BD, Mississauga, ON, Canada), put on ice immediately,and spun at 3,000 revolutions per minute for 15 minutesand then plasma was collected and stored at −70°C untilfurther use [12]. Human 39-plex kits (EMD Millipore,Billerica, MA, USA) were used in accordance with therecommendations of the manufacturer with modifica-tions as described below. Briefly, samples were mixedwith antibody-linked magnetic beads on a 96-well plateand incubated overnight at 4°C with shaking. Plateswere washed twice with wash buffer in a BioTek ELx405washer (BioTek, Winooski, VT, USA). After 1-hourincubation at room temperature with biotinylated detec-tion antibody, streptavidin-phycoerythrin was added for30 minutes with shaking. Plates were washed as above,and phosphate-buffered saline was added to wells forreading by using a Luminex 200 (Illumina Inc., SanDiego, CA, USA) with a lower bound of 100 beads persample per cytokine. Each sample was measured induplicate.Statistical analysisWe tested for the differences in baseline characteristicsbetween the three groups of patients (obese, overweight,and patients with a BMI of less than 25 kg/m2) by usinga Kruskal-Wallis test for continuous data or a chi-squaretest for categorical data, and we report the median andinterquartile range (IQR). We evaluated the primaryoutcome variable (28-day mortality) by using log-rank(Mantel-Cox) test to compare Kaplan-Meier curves forpatients with a BMI of less than 25 kg/m2, overweightpatients, and obese patients. The risk of 28-day hospitalmortality was expressed as hazard ratio (HR) with 95%confidence interval (CI). We tested for the influence ofcovariates, including APACHE II (Acute Physiology andChronic Health Evaluation II) score, gender, diabeteshistory, lung infection, fungal infection, and BMI byusing Cox regression analysis. Secondary clinical out-comes were days alive and free of organ dysfunction(including cardiovascular, respiratory, renal, hepatic, neu-rologic, and coagulation), days alive and free of SIRS (twoof four SIRS criteria), and days alive and free of artificialorgan support (including mechanical ventilation andhemodialysis). We tested for differences in the site ofinfection and the frequency of Gram-positive, Gram-negative, or fungal infection by using a chi-square test.We tested for differences in fluid amount and vasopres-sor administration dosages by using a Kruskal-Wallistest. We tested for differences in the vasopressin concen-tration over time by using one-way repeated-measuresanalysis of variance, and we report the mean and stan-dard error of the mean. Serum cytokine concentrationsof three patient groups were analyzed by using a Kruskal-Wallis test, and we report the median and IQR. Differ-ences were considered significant by using a two-tailedP value of less than 0.05. Analyses were performed byusing SPSS (version 17.0; IBM Corporation, Armonk,NY, USA) statistical software packages.ResultsBaseline characteristicsThe median BMIs of the patients with a BMI of less than25 kg/m2, overweight patients, and obese groups were 23,28, and 34 kg/m2, respectively. At initial presentation, theobese group - compared with the other two patientgroups - least frequently had male patients (obese 55.5%,overweight 67.9%, BMI of less than 25 kg/m2 62%; P =0.03), most frequently had diabetes (obese 29.8%, over-weight 20.6%, BMI of less than 25 kg/m2 14.9%; P <0.0001), had the highest CVP value (obese 15 mm Hg,overweight 14 mm Hg, BMI of less than 25 kg/m214 mm Hg; P = 0.002), had the highest serum creatinineconcentration (obese 191 μmol/L, overweight 150 μmol/L, BMI of less than 25 kg/m2 130 μmol/L; P < 0.0001),and was mechanically ventilated with a highest tidalvolume per kilogram of predicted body weight [23](obese 9.3 mL/kg, overweight 8.8 mL/kg, BMI of lessthan 25 kg/m2 8.3 mL/kg; P < 0.0001) on enrollment.There were no differences in age, APACHE II severityscore, or other laboratory variables and calculations atbaseline, including serum lactate and glomerular filtra-tion rate, across the three patient groups (Table 1).Obese and overweight patients with septic shock havelower 28-day mortality and less organ dysfunction thanpatients with a BMI of less than 25 kg/m2 despitesimilar severity at presentationWe found that the obese and overweight patients had sig-nificantly lower 28-day mortality compared with thosewith a BMI of less than 25 kg/m2 (P = 0.02) (BMI of lessthan 25 kg/m2 versus overweight, P = 0.10; BMI of lessthan 25 kg/m2 versus obese, P = 0.01; overweight versusobese, P = 0.2) (Figure 1). We introduced BMI as a contin-uous variable in our regression model to validate ourhypothesis, following adjustment for factors that areknown to influence mortality (APACHE II score) and thatdiffer between groups at enrollment (gender, pre-existingdiabetes, lung infection and fungal infection). We foundthat for every 1-unit increase in BMI, the HR-adjustedmortality was 2% lower (95% CI 0.97 to 0.99, P = 0.04)(Table 2). We also found that obese and overweightpatients had significantly lower coagulation dysfunctionWacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122Page 3 of 10(obese 28, overweight 23, and BMI of less than 25 kg/m221 days alive and free of dysfunction; P <0.0001) and SIRS(obese 8, overweight 7, and BMI of less than 25 kg/m2 4days alive and free of two of four SIRS criteria; P = 0.02)compared with those with a BMI of less than 25 kg/m2.Differences in other organ dysfunctions were not signifi-cantly different across the three patient groups. Therewere no differences in days alive and free of mechanicalventilation (obese 11, overweight 10, BMI of less than 25kg/m2 6 days; P = 0.36) and days alive and free of renalreplacement therapy (obese 25, overweight 24, BMI of lessthan 25 kg/m2 24 days; P = 0.93) among the three patientgroups.Pattern of infectionCompared with patients with a BMI of less than 25 kg/m2,obese and overweight patients had a significantly lowerrate of lung infection (obese 35%, overweight 45%, andBMI of less than 25 kg/m2 49.8%; P = 0.003) as the sourceof their severe sepsis and had significantly fewer fungalTable 1 Baseline characteristics compared on the basis of actual body mass index category in three groups of patientswith septic shockBaseline characteristics Actual body mass indexa category, kg/m2 P valueBMI <25(n = 276)BMI 25 to <30(n = 209)BMI ≥30(n = 245)DemographicsAge, years 63 (48-74) 64 (50-73) 63 (53-72) 0.62Gender: Male, number (percentage) 171 (62) 142 (67.9) 136 (55.5) 0.03bRace: Caucasian, number (percentage) 231 (83.7) 177 (84.7) 209 (85.3) 0.88Body mass indexa, kg/m2 23 (20-24) 28 (26-29) 34 (32-39) <0.0001bSeverity of illnessAPACHE II score 27 (21-32) 26 (22-31) 27 (23-32) 0.51Surgical condition, number (percentage) 52 (18.8) 45 (21.5) 58 (23.7) 0.4In-hospital steroid treatment, number (percentage) 211 (76.7) 167 (79.9) 172 (70.8) 0.07Pre-existing conditions, number (percentage)Chronic obstructive pulmonary disease 44 (15.9) 31 (14.8) 41 (16.7) 0.86Congestive heart failure 19 (6.9) 14 (6.7) 22 (9.0) 0.57Chronic liver disease 27 (9.8) 30 (14.4) 26 (10.6) 0.26Chronic renal failure 29 (10.5) 24 (11.5) 27 (11) 0.94Diabetes 41 (14.9) 43 (20.6) 73 (29.8) <0.0001bIschemic heart disease 40 (15.4) 38 (20) 50 (21.8) 0.18Chronic steroid use 67 (24.3) 42 (20.1) 51 (20.8) 0.48Hemodynamic variables at day 1Central venous pressure, mm Hg 14 (10-17) 14 (11-18) 15 (12-18) 0.002bMean arterial pressure, mm Hg 55 (50-60) 56 (51-61) 55 (49-61) 0.09Heart rate, beats per minute 130 (115-144) 125 (112-140) 125 (106-140) 0.05Mechanical ventilation variables at day 1Tidal volume, mL 530 (450-630) 570 (480-650) 571 (490-650) 0.03bTidal volume, mL/kg of PBWc 8.3 (7.0-9.7) 8.8 (7.4-10.3) 9.3 (7.9-10.8) <0.0001bLaboratory variables at day 1White blood cells, x103/mm3 14.5 (6.7-21.8) 13.1 (8.0-19.2) 12.8 (7.4-20.2) 0.72Platelets, x103/mm3 145 (55-256) 144 (72-232) 165 (90.5-254.5) 0.13Creatinine, μmol/L 130 (80-218) 150 (104-266) 191 (120-297) <0.0001bGFR, mL/minute 44 (26-76) 45 (25-72) 47 (29-77) 0.27Lactate, mmol/L 2.0 (1.2-4.2) 2.1 (0.9-3.8) 1.7 (1.1-3.5) 0.27PaO2/FiO2 180 (120-255) 182 (123-255) 190 (135-259) 0.41Data are presented as median (interquartile range) for continuous variables. P values were calculated for statistical difference between three patient groups withthe use of chi-square test and Kruskal-Wallis test. aBody mass index (BMI) was measured at enrollment of the Vasopressin and Septic Shock Trial (average about11 hours after onset of septic shock). bP <0.05. cPredicted body weight (PBW) (male: 50 + 0.91 (centimeters of height: 152.4), female: 45.5 + 0.91 (centimeters ofheight: 152.4)). APACHE II, Acute Physiology and Chronic Health Evaluation II; GFR, glomerular filtration rate (Cockcroft-Gault formula); PaO2/FiO2, arterial oxygenpartial pressure per fraction of inspired oxygen.Wacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122Page 4 of 10infections (obese 8.2%, overweight 11%, and BMI of lessthan 25 kg/m2 15.6%; P = 0.03) (Table 3).Obese and overweight patients receive less intravenousfluids and vasopressors than patients with a BMI ofless than 25 kg/m2The updated Surviving Sepsis Campaign guidelines sug-gest titrating fluid therapy to body mass with an initialprescription of at least 30 mL/kg [24]. Further mass-adjusted fluid administration is prescribed as required. Itis now appreciated that, whereas adequate early fluidresuscitation is crucial, overzealous fluid administrationis harmful [21,22]. During the initial resuscitation forseptic shock, obese and overweight patients received sig-nificantly less fluids per kilogram than those with a BMIof less than 25 kg/m2 (Table 4). This appears driven bynon-body weight-adjusted fluid prescription, becausethere was no significant difference in absolute fluidadministered between groups (Table 4). Although vaso-pressin is traditionally prescribed without body weightadjustment, some have suggested that vasopressin be pre-scribed with it [25]. During the initial phase of septicshock (day 1), both norepinephrine (obese 0.14, IQR 0.09to 0.25; overweight 0.21, IQR 0.12 to 0.34; BMI of lessthan 25 kg/m2 0.26, IQR 0.15 to 0.44 μg/kg per minute;P < 0.0001) and body weight-adjusted vasopressin (obese0.28, IQR 0.23 to 0.33; overweight 0.36, IQR 0.31 to 0.40;BMI of less than 25 kg/m2 0.43, IQR 0.38 to 0.50 μU/kgper minute; P <0.0001) dosage - similar to fluid adminis-tration - were prescribed at significantly lower doses inobese and overweight patients compared with those witha BMI of less than 25 kg/m2. At 72 hours of vasopressininfusion, obese and overweight patients had a trendtoward lower serum vasopressin level (obese 28.9 ± 7.9,overweight 51.5 ± 16.5, BMI of less than 25 kg/m2 69.9 ±17.5 pmol/L; P = 0.08) compared with patients with aFigure 1 Kaplan-Meier curves of 28-day survival compared by actual body mass index (BMI) across three groups of patients with septicshock. Obese patients with septic shock had a significantly lowest 28-day mortality following overweight patients, whereas the septic shock patientswith a BMI of less than 25 kg/m2 had a highest 28-day mortality (P = 0.02, log-rank analysis). *P <0.05, compared among three groups.Table 2 Hazard ratio of 28-day mortality in patients with septic shockHazard ratio 95% confidence interval P valueAPACHE II score (per score) 1.06 1.05-1.08 <0.0001aMale 1.12 0.87-1.43 0.39Diabetes 1.05 0.78-1.41 0.77Lung infection 1.02 0.79-1.33 0.86Fungal infection 1.26 0.87-1.82 0.22BMI (per score) 0.98 0.97-0.99 0.041aP values were calculated by using Cox regression analysis. aP <0.05. APACHE II, Acute Physiology and Chronic Health Evaluation II; BMI, body mass index.Wacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122Page 5 of 10BMI of less than 25 kg/m2 receiving vasopressin infusion(Figure 2).IL-6 levels of obese and overweight patients are mutedcompared with those of patients with a BMI of less than25 kg/m2There was no significant difference in body temperature(obese 38.6, IQR 37.8 to 39.4; overweight 38.5, IQR 37.8to 39.1; BMI of less than 25 kg/m2 38.4, IQR 37.5 to39.1°C; P = 0.053), platelet count (obese 165,000, IQR90,500 to 254,500; overweight 144,000, IQR 72,000 to232,000; BMI of less than 25 kg/m2 145,000, IQR 55,000to 256,000/mm3; P = 0.13), and white blood cell count(obese 12,800, IQR 7,400 to 20,200; overweight 13,100,IQR 8,000 to 19,200; BMI of less than 25 kg/m2 14,500,IQR 6,700 to 21,800/mm3; P = 0.72) between groups.Table 3 Pattern of infection of patients with septic shockActual body mass index category, kg/m2 P valueBMI <25(n = 276)BMI 25 to <30(n = 209)BMI ≥30(n = 245)Primary source of infection, number (percentage)Lung 137 (49.8) 94 (45) 85 (35) 0.003aAbdomen 69 (25.1) 54 (25.8) 73 (30) 0.41Genitourinary 10 (3.6) 11 (5.3) 13 (5.3) 0.58Blood 13 (4.7) 14 (6.7) 16 (6.6) 0.57Skin 17 (6.2) 14 (6.7) 26 (10.7) 0.12Otherb 23 (8.4) 21 (10) 25 (10.3) 0.72Pathogen type in cultures, number (percentage)Gram-positive 81 (29.5) 53 (25.4) 81 (33.3) 0.18Gram-negative 68 (24.7) 52 (24.9) 46 (18.9) 0.21Fungal 43 (15.6) 23 (11) 20 (8.2) 0.03aaP <0.05. bBone/joint/central nervous system infection. P values were calculated for statistical difference between three patient groups with the use of chi-squaretest. BMI, body mass index.Table 4 Fluid administration during the first four days after septic shockAmount of fluid, mL Amount of fluid per body weight, mL/kgBMI <25(n = 276)BMI 25 to <30(n = 209)BMI ≥30(n = 245)P value BMI <25(n = 276)BMI 25 to <30(n = 209)BMI ≥30(n = 245)P valuePredicted body weighta, kg 67 ± 2.1 64 ± 0.7 61 ± 1.0 0.003b - - -Day 1Intake 11,000 ± 310 10,500 ± 360 10,500 ± 310 0.27 180 ± 5 130 ± 5 100 ± 3 <0.0001bOutput 4,100 ± 200 3,900 ± 240 3,900 ± 190 0.7 60 ± 3 50 ± 3 40 ± 2 <0.0001bNet accumulated balance 7,000 ± 330 6,600 ± 360 6,700 ± 330 0.43 110 ± 5 84 ± 5 70 ± 3 <0.0001bDay 2Intake 16,000 ± 420 15,200 ± 500 15,400 ± 420 0.28 250 ± 7 190 ± 6 150 ± 5 <0.0001bOutput 6,500 ± 270 6,400 ± 320 6,400 ± 290 0.95 100 ± 5 80 ± 4 60 ± 3 <0.0001bNet accumulated balance 9,500 ± 430 8,800 ± 490 9,000 ± 460 0.28 150 ± 7 110 ± 6 90 ± 5 <0.0001bDay 3Intake 19,900 ± 520 19,000 ± 600 19,400 ± 510 0.37 320 ± 9 240 ± 8 190 ± 5 <0.0001bOutput 9,100 ± 360 8,900 ± 410 9,300 ± 410 0.95 140 ± 6 110 ± 5 90 ± 4 <0.0001bNet accumulated balance 10,800 ± 490 10,200 ± 580 10,200 ± 540 0.31 170 ± 8 130 ± 7 100 ± 5 <0.0001bDay 4Intake 23,600 ± 590 23,000 ± 710 23,300 ± 590 0.41 380 ± 10 290 ± 9 230 ± 7 <0.0001bOutput 12,000 ± 430 12,300 ± 530 12,600 ± 530 0.9 190 ± 7 160 ± 7 120 ± 6 <0.0001bNet accumulated balance 11,700 ± 540 10,700 ± 660 10,700 ± 600 0.17 190 ± 9 140 ± 8 110 ± 6 <0.0001bData are presented as mean ± standard error of the mean. P values were calculated for statistical difference between three patient groups with the use of Kruskal-Wallistest. aPredicted body weight (male: 50 + 0.91 (centimeters of height: 152.4), female: 45.5 + 0.91 (centimeters of height: 152.4)). bP <0.05. BMI, body mass index.Wacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122Page 6 of 10We found that, in the early phase of septic shock, plasmaIL-6 (obese 106, IQR 34 to 686; overweight 190, IQR 44 to2,339; BMI of less than 25 kg/m2 235, IQR 44 to 1,793 pg/mL; P = 0.046) levels were significantly lower in obese andoverweight patients than in those with a BMI of less than25 kg/m2 (Figure 3). There was no difference in plasmaMCP-1 (obese 567, IQR 220 to 1,817; overweight 842, IQR284 to 2,386; BMI of less than 25 kg/m2 782, IQR 315 to2,686 pg/mL; P = 0.06), resistin (obese 77, IQR 39 to 137;overweight 66, IQR 37 to 134; BMI of less than 25 kg/m273, IQR 38 to 134 ng/mL; P = 0.85), or plasma TNF-a inobese or overweight patients versus those with a BMI ofless than 25 kg/m2 (obese 15, IQR 6 to 29; overweight14, IQR 7 to 31; BMI of less than 25 kg/m2 13, IQR 6 to33 pg/mL; P = 0.86).DiscussionOur major finding was that despite equal severity of ill-ness upon presentation, mortality in obese and over-weight patients was significantly lower than in patientswith a BMI of less than 25 kg/m2 (Figure 1). Comparedwith those with a BMI of less than 25 kg/m2, obese andoverweight patients had less frequent lung and fungalinfections as the site and organism causing septic shock.We also found that obese and overweight patients weretreated differently, in what appeared to be a ‘one size fitsall’ non-weight-adjusted dosing, so that obese and over-weight patients received less fluids and vasopressors perkilogram (norepinephrine and vasopressin) than patientswith a BMI of less than 25 kg/m2. Unlike previousreports of obesity augmenting the inflammatory response[6,7,13], we found that the IL-6 inflammatory responsewas muted in overweight and obese patients comparedwith those with a BMI of less than 25 kg/m2 in early sep-tic shock. This surprising result is novel and consistentwith current and previous reports of improved survivaloutcomes in obese patients [8,9].Figure 2 Serum vasopressin level during septic shock compared across the three patient groups. At 24 and 72 hours after vasopressininfusion, the overweight and obese patients had a trend toward lower mean vasopressin concentrations compared with those with body massindex (BMI) of less than 25 kg/m2 (P = 0.08, one-way repeated-measures analysis of variance). Error bars indicate standard error of the mean.Figure 3 Plasma interleukin-6 (IL-6) expression differences atthe initial phase of septic shock compared across the threepatient groups. Compared with patients with a body mass index(BMI) of less than 25 kg/m2, overweight and obese patients hadsignificantly lower IL-6 (P = 0.046) plasma levels at the initial phaseof septic shock. P values were calculated by using the Kruskal-Wallistest. Error bars indicate interquartile range. *P <0.05.Wacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122Page 7 of 10Conceivably, the inclusion of very underweight or veryoverweight patients could alter this analysis and inter-pretation. We removed the underweight group (n = 26)and re-analyzed for mortality difference and found thesame significant result in which obese patients had thelowest 28-day mortality followed by overweight and nor-mal-BMI patients. A similar post hoc analysis of themorbidly obese (BMI ≥40 kg/m2) showed that thisgroup had a lower 28-day mortality than all otherweight groups, consistent with a previous report byAbhyankar and colleagues [26]. We also found thatdespite a higher prevalence of diabetes, those with obe-sity had a decreased risk of hospital mortality. Covariateadjustment of diabetes appeared to augment this protec-tion for the obese. We found that BMI was an indepen-dent predictor of lower mortality with an HR of 0.98per kg/m2 (95% CI 0.97 to 0.99, P = 0.041), suggestingthat although diabetics had a poorer prognosis, diabetesitself had a weaker association with mortality than BMI(Table 2). Our results are in agreement with data fromthe Acute Decompensated Heart Failure National Regis-try, which studied patients with acute heart failure andfound that hospital mortality rates decreased in a near-linear fashion across higher BMI. For every 5-unitincrease in BMI, the odds of risk-adjusted mortalitywere 10% lower (95% CI 0.88 to 0.93, P < 0.0001) [10].Our findings also align with a meta-analysis in criticallyill patients which demonstrated that patients who wereobese (had a BMI of 30 to 39.9 kg/m2) had significantlylower mortality compared with the non-obese patients(BMI of less than 30 kg/m2) (relative risk 0.86, 95% CI0.81 to 0.91; P <0.001) [11].Previous studies demonstrate that low tidal volumeventilation is associated with a lower mortality [23],which might have contributed to our findings. However,we found the opposite. Obese and overweight patientsreceived higher tidal volume ventilation compared withpatients with a BMI of less than 25 kg/m2, particularlywhen corrected to predicted body weight (Table 1).Although our data do not allow us to determine whetherthere was an absolute difference in susceptibility to infec-tion (since we would need to know the number of obeseand non-obese patients at risk), we did observe a substan-tial difference in pattern of infection. We observed a largeand significant reduction in the fraction of patients havinglung infections and the fraction of patients having fungalinfections in obese patients. One possible limitation to thisfinding is that, in this patient cohort, the diagnosis of lunginfection was based primarily on physician judgment,which depended on results from chest radiographs andsputum culture. This could have been confounded by theincreasing inability to obtain clear or definitive chestradiographs in obese patients. Although there was no sta-tistically significant difference in white blood cell countsor a baseline difference in immunosuppression or chronicsteroid use, we did find that obese and overweight patientsexpressed significantly lower plasma IL-6 levels during theinitial phase of septic shock compared with those with aBMI of less than 25 kg/m2. This is in agreement with astudy of patients with blunt injury which found that obesepatients had a significantly lower inflammatory cytokineprofile than those with a normal BMI [27]. Why obesepatients have an altered IL-6 inflammatory response is notknown. However, the finding of decreased circulating IL-6concentrations in obese patients with sepsis is consistentwith the observation of improved survival outcome. Wealso measured resistin, an adipokine that antagonizes theeffects of insulin [16] and that functions as a pro-inflam-matory cytokine [28]. Unlike in stable obese patients inwhom resistin levels are much higher than those with anormal BMI, we found no significant difference betweenobese and normal-BMI patients in early septic shock.Unlike in most medical therapies which rely uponweight-based dosing, we hypothesized that the empiricalmanner (1 L of fluid at a time mentality) in which fluidand vasopressor therapy are recommended for sepsis[22,24] would cause obese and overweight patients toreceive relatively less weight-adjusted fluids and vaso-pressors than normal-BMI patients. Alternatively, thetreating physicians may have recognized that bloodvolume is not linearly related to weight [29,30] and,therefore, may have followed the current Surviving SepsisCampaign guidelines that suggest that fluid administra-tion be based on ideal body weight, which did not differbetween the three groups in this study (Table 4). Indeed,obese and overweight patients received less body weight-adjusted fluids, norepinephrine, and vasopressin in earlyseptic shock compared with the patients with a BMI ofless than 25 kg/m2. As excessive fluid and vasopressorresuscitation is associated with increased mortality[21,22], it may be that using doses more typically usedfor smaller individuals was, ironically, protective. Todirectly assess whether this was the case, we measuredvasopressin levels in patients randomly assigned toreceive this drug in the protocolized non-body weight-adjusted dose used in the VASST. We found that, com-pared with those with a BMI of less than 25 kg/m2, obeseand overweight patients had a trend toward lower serumvasopressin concentrations after 24 hours of vasopressininfusion. The net accumulated fluid balance differs byonly 1 L between patients of normal weight and obesepatients. It could be argued that this small difference isunlikely to account for the observed outcome differences.However, indexed to body mass, this amounts to a differ-ence of 80 mL/kg, a remarkable difference that conceiva-bly could be related to outcomes. In addition, althoughthere was a statistically significant difference in CVPvalues between the three patient groups, which wasWacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122Page 8 of 10possibly due to differences in chest wall compliance inthe obese compared with the normal-BMI patients, theclinical significance of this small difference is uncertain.The strengths of this study are, first, that our cohort is alarge prospective multicenter cohort of patients with well-defined septic shock [12]. Second, our results after covari-ate-adjusted and unadjusted analyses were consistent.There are several important limitations to our study.APACHE II score has never been validated in obesepatients. For example, oxygenation may be worseningdespite normal lung function or creatinine level may beincreased despite normal kidney function. Thus, obesepatients may have artificially elevated APACHE IIscores. However, there was no difference of serum lac-tate levels between the three patient groups at enroll-ment. Because our analysis was retrospective, someadditional useful data such as coexisting hypertension,serum cholesterol, or other serum adipokines such asleptin and adiponectin which regulate insulin sensitivitywere not recorded in the dataset. In addition, adiposetissue-related cytokines in our analysis are released notonly from adipose tissue but also from other tissues andthis could affect our interpretation. Also, our study mayrepresent not the ‘real’ population of obese patients withsepsis but only an obese population included in a rando-mized controlled study with a strict inclusion criteria.ConclusionsWe demonstrated that, in a cohort of patients with sep-tic shock, the obese and overweight patients haddecreased mortality compared with patients with a BMIof less than 25 kg/m2. Obese and overweight patientsmay have had less aggressive disease, as shown by ourevaluations of blunted host inflammatory response topathogens (lower IL-6), and an altered pattern of infec-tion (less frequent lung and fungal infection). Obese andoverweight patients may also have been protected byreceiving less fluids and vasopressors compared with thepatients with a BMI of less than 25 kg/m2.Key messages• In septic shock, the obese and overweight patientshad decreased mortality compared with patientswith a BMI of less than 25 kg/m2.• Compared with patients with a BMI of less than25 kg/m2, obese and overweight patients with sepsishad a lower rate of lung infection as the source oftheir severe sepsis as well as fewer fungal infections.• In septic shock, the obese and overweight patientsreceive less intravenous fluids and vasopressors perkilogram than patients with a BMI of less than25 kg/m2.• At the initial phase of septic shock, plasma IL-6concentrations in obese and overweight patients aremuted compared with those of patients with a BMIof less than 25 kg/m2.AbbreviationsAPACHE II: Acute Physiology and Chronic Health Evaluation II; BMI: bodymass index; CI: confidence interval; CVP: central venous pressure; HR: hazardratio; ICU: intensive care unit; IL-6: interleukin-6; IQR: interquartile range;MCP-1: monocyte chemotactic protein-1; SIRS: systemic inflammatoryresponse syndrome; TNF-α: tumor necrosis factor-alpha; VASST: Vasopressinand Septic Shock Trial.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsPW and KRW contributed to study conception and design, acquisition ofdata, statistical analysis, interpretation of data, and drafting of themanuscript. JAR and JHB contributed to study conception and design,acquisition of data, interpretation of data, and drafting of the manuscript. Allauthors read and approved the final manuscript.AcknowledgementsThe following institutions participated in the VASST, which was funded byCanadian Institutes of Health Research: Canada - British Columbia: St. Paul’sHospital, Vancouver General Hospital, Royal Jubilee Hospital, KelownaGeneral Hospital, Richmond General Hospital, Royal Columbian Hospital,Manitoba: St. Boniface Hospital, Winnipeg Health Science Centre, Ontario:Ottawa Hospital, General Campus, University Health NetworkToronto General& Toronto Western Hospitals, St. Joseph’s Hospital, Mount Sinai Hospital,Ottawa Hospital, Civic Campus, St. Michael’s Hospital, Sunnybrook andWomen’s College Health Science Centre, Hamilton Health Sciences Centre,London Health Sciences Centre, Sudbury Regional Hospital, Charles LeMoyneHospital, Hôtel-Dieu Grace Hospital. Australia - Victoria: Alfred Hospital, RoyalMelbourne Hospital, Monash Medical Centre, Western Australia: Royal PerthHospital, South Australia: Flinders Medical Centre. US - Phoenix, AZ: MayoClinic Hospital.Received: 21 January 2013 Revised: 12 April 2013Accepted: 20 June 2013 Published: 20 June 2013References1. Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S, Murray CJ, ComparativeRisk Assessment Collaborating Group: Selected major risk factors andglobal and regional burden of disease. Lancet 2002, 360:1347-1360.2. Huttunen R, Syrjanen J: Obesity and the outcome of infection. LancetInfect Dis 2010, 10:442-443.3. Milner JJ, Beck MA: The impact of obesity on the immune response toinfection. Proc Nutr Soc 2012, 71:298-306.4. Huttunen R, Syrjanen J: Obesity and the risk and outcome of infection. IntJ Obes (Lond) 2013, 37:333-340.5. Harrington G, Russo P, Spelman D, Borrell S, Watson K, Barr W, Martin R,Edmonds D, Cocks J, Greenbough J, Lowe J, Randle L, Castell J, Browne E,Bellis K, Aberline M: Surgical-site infection rates and risk factor analysis incoronary artery bypass graft surgery. Infect Control Hosp Epidemiol 2004,25:472-476.6. Wisse BE: The inflammatory syndrome: the role of adipose tissuecytokines in metabolic disorders linked to obesity. J Am Soc Nephrol2004, 15:2792-2800.7. Benson S, Arck PC, Tan S, Mann K, Hahn S, Janssen OE, Schedlowski M,Elsenbruch S: Effects of obesity on neuroendocrine, cardiovascular, andimmune cell responses to acute psychosocial stress in premenopausalwomen. Psychoneuroendocrinology 2009, 34:181-189.8. Habbu A, Lakkis NM, Dokainish H: The obesity paradox: Fact or fiction?Am J Cardiol 2006, 98:944-948.9. Kalantar-Zadeh K, Horwich TB, Oreopoulos A, Kovesdy CP, Younessi H,Anker SD, Morley JE: Risk factor paradox in wasting diseases. Curr OpinClin Nutr Metab Care 2007, 10:433-442.10. Fonarow GC, Srikanthan P, Costanzo MR, Cintron GB, Lopatin M, ADHEREScientific Advisory Committee and Investigators: An obesity paradox inWacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122Page 9 of 10acute heart failure: analysis of body mass index and inhospital mortalityfor 108,927 patients in the Acute Decompensated Herat Failure NationalRegistry. Am Heart J 2007, 153:74-81.11. Akinnusi ME, Pineda LA, El Solh AA: Effect of obesity on intensive caremorbidity and mortality: a meta-analysis. Crit Care Med 2008, 36:151-158.12. Russell JA, Walley KR, Singer J, Gordon AC, Hebert PC, Cooper DJ,Holmes CL, Mehta S, Granton JT, Storms MM, Cook DJ, Presneil JJ, Ayers D,VASST Investigators: Vasopressin versus norepinephrine infusion inpatients with septic shock. N Engl J Med 2008, 358:877-887.13. Roytblat L, Rachinsky M, Fisher A, Greemberg L, Shapira Y, Douvdevani A,Gelman S: Raised interleukin-6 levels in obese patients. Obese Res 2000,8:673-675.14. Gerhardt CC, Romero IA, Cancello R, Camoin L, Strosberg AD: Chemokinescontrol fat accumulation and leptin secretion by cultured humanadipocytes. Mol Cell Endocrinol 2001, 175:81-92.15. Hotamisligil GS, Arner P, Caro JF, Atkinson RL, Spiegelman BM: Increasedadipose tissue expression of tumor necrosis factor-alpha in humanobesity and insulin resistance. J Clin Invest 1995, 95:2409-2415.16. Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright CM, Patel HR,Ahima RS, Lazar MA: The hormone resistin links obesity to diabetes.Nature 2001, 409:307-312.17. American College of Chest Physicians/Society of Critical care medicineconsensus conference: definitions for sepsis and organ failure andguidelines for the use of innovative therapies in sepsis. Crit Care Med1992, 20:864-874.18. Sibbald WJ, Vincent JL: Roundtable conference on clinical trials for thetreatment of sepsis. Brussels, march 12-14, 1994. Chest 1995, 107:522-527.19. National Institutes of Health BMI categories. [http://nhlbisupport.com/bmi/].20. Sutherland AM, Walley KR, Manocha S, Russell JA: The association ofinterleukin 6 haplotype clades with mortality in critically ill adults. ArchIntern Med 2005, 165:75-82.21. National Heart, Lung, and Blood Institute Acute Respiratory DistressSyndrome (ARDS) Clinical Trials Network, Wiedemann HP, Wheeler AP,Bernard GR, Thompson BT, Hayden D, deBoisblanc B, Connors AF Jr,Hite RD, Harabin AL: Comparison of two fluid-management strategies inacute lung injury. N Engl J Med 2006, 354:2564-2575.22. Boyd JH, Forbes J, Nakada TA, Walley KR, Russell JA: Fluid resuscitation inseptic shock: a positive fluid balance and elevated central venouspressure are associated with increased mortality. Crit Care Med 2011,39:259-265.23. Ventilation with lower tidal volumes as compared with traditional tidalvolumes for acute lung injury and the acute respiratory distresssyndrome. The Acute Respiratory Distress Syndrome Network. N Engl JMed 2000, 342:1301-1308.24. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM,Sevransky JE, Sprung CL, Douglas IS, Jaeschke R, Osborn TM, Nunnally ME,Townsend SR, Reinhart K, Kleinpell RM, Angus DC, Deutschman CS,Machado FR, Rubenfeld GD, Webb SA, Beale RJ, Vincent J-L, Moreno R, andthe Surviving Sepsis Campaign Guidelines Committee including thePediatric subgroup: Surviving sepsis campaign: international guidelinesfor the management of severe sepsis and septic shock: 2012. Crit CareMed 2013, 41:580-637.25. Miller JT, Welage LS, Kraft MD, Alaniz C: Does body weight impact theefficacy of vasopressin therapy in the management of septic shock? JCrit Care 2012, 27:289-293.26. Abhyankar S, Leishear K, Callaghan FM, Demner-Fushman D, McDonald CJ:Lower short- and long-term mortality associated with overweight andobesity in a large cohort study of adult intensive care unit patients. CritCare 2012, 16:R235.27. Winfield RD, Delano MJ, Cuenca AG, Cendan JC, Lottenberg L, Efron PA,Maier RV, Remick DG, Moldawer LL, Cuschieri J: Obese patients show adepressed cytokine profile following severe blunt injury. Shock 2012,37:253-256.28. Sunden-Cullberg J, Nystrom T, Lee ML, Mullins GE, Tokics L, Andersson J,Norrby-Teglund A, Treutiger CJ: Pronounced elevation of resistincorrelates with severity of disease in severe sepsis and septic shock. CritCare Med 2007, 35:1536-1542.29. Gibson JG, Evans WA: Clinical studies of blood volume. I. Clinicalapplication of a method employing the azo dye ‘evans blue’ and thespectrophotometer. J Clin Invest 1937, 16:301-316.30. Gregersen MI, Nickerson JL: Relation of blood volume and cardiac outputto body type. J Appl Physiol 1950, 3:329-341.doi:10.1186/cc12794Cite this article as: Wacharasint et al.: One size does not fit all in severeinfection: obesity alters outcome, susceptibility, treatment, andinflammatory response. Critical Care 2013 17:R122.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitWacharasint et al. Critical Care 2013, 17:R122http://ccforum.com/content/17/3/R122Page 10 of 10


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items