lable at ScienceDirect

Journal of Psychiatric Research 45 (2011) 1036e1041

Contents lists avai

Journal of Psychiatric Research

journal homepage: www.elsevier .com/locate/psychires

Gender-specific lipid profiles in patients with bipolar disorder

Mytilee Vemuri*, Heather A. Kenna, Po W. Wang, Terence A. Ketter, Natalie L. RasgonDepartment of Psychiatry and Behavioral Sciences, Stanford University, 401 Quarry Road, Stanford, CA 94305-5723, United States

a r t i c l e i n f o

Article history:Received 15 April 2010Received in revised form31 January 2011Accepted 4 February 2011

Keywords:DyslipidemiaBipolar disorderGenderInsulin resistanceMetabolic

* Corresponding author. Tel.: þ1 650 725 1774; faxE-mail address: mvemuri@stanford.edu (M. Vemu

0022-3956/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.jpsychires.2011.02.002

a b s t r a c t

Objective: High rates of dyslipidemia and insulin resistance (IR) are reported in patients with bipolardisorder (BD). We assessed gender effects upon rates of dyslipidemia/IR in outpatients with BD.Methods: Data from 491 outpatients (ages 18e88) seen in the Stanford Bipolar Disorders clinic between2000 and 2007 were evaluated. Patients were followed longitudinally and received naturalistic treat-ment. BD patients (n ¼ 234; 61% female; 42% Type I, 47% Type II, 11% NOS) with a mean age of 40.3 � 14.0years, mean BMI 26.8 � 6.4, and 81% Caucasian, who had one of four lipid measures (total cholesterol,LDL, HDL, TG) at clinicians’ discretion, a psychiatry clinic visit within 2 months of laboratory, and werenot medicated for dyslipidemia were included. IR was imputed from TG/HDL ratio.Results: Women, compared with men, had significantly lower mean triglycerides (105.58 � 64.12 vs.137.99 � 105.14, p ¼ 0.009), higher mean HDL cholesterol (60.17 � 17.56 vs. 46.07 � 11.91 mg/dl,p < 0.001), lower mean LDL cholesterol (109.84 � 33.47 vs. 123.79 � 35.96 mg/dl, p ¼ 0.004), and lowerTG/HDL ratio (1.98 � 1.73 vs. 3.59 � 3.14 p < 0.001). Compared to men, women had a significantly lowerprevalence of abnormal total cholesterol, LDL, TG, HDL, and TG/HDL ratio. No significant differences werefound between men and women with regard to age, BMI, ethnicity, educational attainment, smokinghabits, bipolar illness type, illness severity or duration, or weight-liable medication exposure.Discussion: In outpatients with BD, women had more favorable lipid profiles than men despite similardemographic variables. This sample of primarily Caucasian and educated patients, receiving vigilantclinical monitoring, may represent a relatively healthy psychiatric population demonstrating genderdifferences similar to those in the general population.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

High rates of metabolic abnormalities such as dyslipidemia,diabetes, obesity, and the metabolic syndrome (MetS) occur inpatients with bipolar disorder (BD) (Fagiolini et al., 2005;Fiedorowicz et al., 2008; Garcia-Portilla et al., 2008; Kilbourneet al., 2004; Sicras et al., 2008; Van Winkel et al, 2008), all ofwhich are modifiable risk factors for cardiovascular disease (CVD)(Newcomer, 2009).

Evaluating dyslipidemia is a component of routine medicationmonitoring in BD, and thus places the treating psychiatrist in theposition of identifying modifiable risk factors for CVD. Low-densitylipoprotein (LDL) cholesterol, triglycerides (TG), and high-densitylipoprotein cholesterol (HDL) are independent risk factors forcoronary heart disease (CHD) (NCEP, 2002). LDL remains theprimary treatment target of the National Cholesterol EducationProgram (NCEP) Adult Treatment Protocol (ATP) III (NCEP, 2002).

: þ1 650 724 9900.ri).

All rights reserved.

The ratio of TG to HDL cholesterol (TG/HDL) has been proposed asa proxy marker of insulin resistance (IR) (McLaughlin et al., 2005),a condition in which tissue responsiveness to the normal action ofinsulin is impaired, and is related to the development of diabetes,dyslipidemia, and possibly mood disorders (NCEP, 2002; Rasgonand Jarvik, 2004; Rasgon et al., 2002).

In the general population, women compared to men havehistorically lived longer, developed cardiovascular disease at anolder age, and had more favorable lipid profiles (Regitz-Zagroseket al., 2007). Internationally, women compared to men consis-tently have more favorable lipid profiles, lower TC, LDL, TG andhigher HDL (Gardner et al., 2000; Gostynski et al., 2004; Seidellet al., 1991; Williams, 2004). These gender differences in lipidsappear to be highly correlated with age and abdominal obesity(Gostynski et al., 2004; Seidell et al., 1991). In the U.S., genderdifferences in lipid distribution appear to be changing. NationalHealth and Nutrition Examination Survey (NHANES) 2003e2006data showed a higher prevalence of lipid abnormalities in womenversus men (American Heart Association, 2010; Regitz-Zagroseket al., 2007), possibly driven by increasing obesity in women(Regitz-Zagrosek et al., 2007). Gender differences in lipids are

M. Vemuri et al. / Journal of Psychiatric Research 45 (2011) 1036e1041 1037

attributed to sex steroid hormone influence on lipids and on sexdifferences in fat distribution; specifically abdominal fat tissue ismore abundant in men compared to premenopausal womenfavoring development of insulin resistance (IR), dyslipidemia andhypertension (Regitz-Zagrosek et al., 2007). After menopause,lipoprotein concentrations and fat distribution shift to a more malepattern (Ley et al., 1992).

In contrast to gender differences in the general population,women in psychiatric populations appear to have higher rates ofmodifiable CVD risk factors, such as obesity (Wang et al., 2006) andMetS (McEvoy et al., 2005; Teixeira and Rocha, 2007), relative tomen, suggesting substantive vulnerability to CVD and its riskfactors in women with mental illness. While women with BDhave high rates of dyslipidemia (Stemmle et al., 2009), genderdifferences in dyslipidemia have not been well characterized inpsychiatric populations. In this study, we compared lipid valuesbetween female and male outpatients with bipolar disorder (BD).Given existing published data suggesting women with mentalillness have a higher vulnerability to CVD and its risk factors, wehypothesized that women with BD would have less favorable lipidprofiles than men.

2. Methods

The study was approved by the Stanford University Adminis-trative Panel on Human Subjects, and patients provided verbal andwritten informed consent prior to participation. Data wereanalyzed from all outpatients (age 18e88) seen in the Stanfordoutpatient Bipolar Disorders clinic between January 2000 and June2007 who had consented to longitudinal data review of theirmedical records and who were seen for ongoing evaluation. Indi-viduals seen for one-time consultations were not included. A totalof 491 patient records were reviewed. Patients were included inanalysis (n ¼ 234) if the following criteria were met: a) patient metDSM-IV diagnostic criteria for bipolar disorder b) patient receivedone of four concurrent lipid measures within twomonths of a clinicvisit including: Total cholesterol (TC) (n ¼ 256), Triglycerides (TG)(n ¼ 253), Low-Density Lipoprotein (LDL) (n ¼ 238), High-DensityLipoprotein (HDL) (n ¼ 250), and c) were not concurrently treatedwith lipid-lowering medications. In this sample, neither genderwas significantly more likely to be treated with a lipid-loweringmedication (c2 ¼ 2.05, p ¼ 0.152), nor were there significantdifferences in lipid values by gender within the 22 patients treatedwith lipid-lowering medication.

2.1. Intake evaluations and clinical follow-up

Baseline clinical evaluations were performed using theSystematic Treatment Enhancement Program for Bipolar Disorder(STEP-BD) Affective Disorders Evaluation (ADE) (Sachs et al., 2003)to establish psychiatric diagnosis of BD, and to derive clinicalinformation including age, current and previous medication treat-ment, years since first episode, bipolar illness type, and height.Demographic information, including self-reported gender, race,and education attainment was collected by patient self-assess-ments and interview. Patients received naturalistic treatmentguided by model practice procedures, which included publishedpharmacotherapy guidelines (Dennehy et al., 2007). Clinical infor-mation, including psychotropic treatment, weight, and illnessseverity, measured by Clinical Global Inventory (CGI), was longi-tudinally collected using STEP-BD Clinical Monitoring Form (CMF)(Sachs et al., 2002).

For analytic purposes, patients were classified into three medi-cation categories based on the medication taken at time of veni-puncture: 1) weight-liable mood stabilizer/antipsychotic (i.e.

associated with substantive weight-gain risk): lithium (Bowdenet al., 2000), valproate (Bowden et al., 2000), oxcarbazepine(Vieta et al., 2008), olanzapine (Tohenet al., 2006), clozapine (Allisonet al., 1999), quetiapine (Weisler et al., 2009), risperidone (Quirozet al., 2010), or aripiprazole (Keck et al., 2006); 2) non-weight-liable mood stabilizer/antipsychotic (associated with non-substan-tive weight-gain risk): lamotrigine (Sachs et al., 2006), ziprasidone(Bowden et al., 2009) or carbamazepine (Weisler et al., 2006), or 3)no mood stabilizer/antipsychotic. Only second-generation antipsy-chotics, if any, were received by patients in the present analysis.Patients were only included in a mood stabilizer/antipsychoticcategory if they had been treated for aminimum of sixmonths withthe above mentioned medication classes. If patients were concur-rently treated with both weight-liable and non-weight-liablemedications, they were included in the weight-liable group.

2.2. Assessment of lipids and metabolic function

Fasting blood lipid panels and thyroid function laboratorieswere obtained at clinician discretion during each patient’s treat-ment as part of routine screening and follow-up. The most recentlaboratory assessment for each patient during period from January2000 to June 2007 was used in this analysis. Body Mass Index (BMI;in kg/m2) was assessed based on height at intake and weight notedat the time of clinical assessment, within two months of laboratorydate.

2.3. Statistics

All statistical analyses were performed using the SPSS softwarefor windows, version 17.0 (SPSS, Chicago, Ill). Gender differences indemographic and clinical variables and lipid values were evaluatedusing t-tests, chi-square tests and Fisher’s exact tests. Genderdifferences in prevalence of abnormal lipid values were evaluatedusing chi-square tests. Pearson correlations were used to assessassociations of the blood lipid values with age, BMI, illness dura-tion, education level, CGI score, and years since first episode in thesample as awhole, and variables that were found to be significantlyassociated with mean lipid values were included in the finalmultivariate analysis of variance (MANOVA) of gender differencesin blood lipid parameters. One-way ANOVAs were used to assessdifferences in blood lipid levels by medication categories (weight-liable, non-weight-liable, or none), bipolar illness type, hypothy-roidism, or nicotine use. All statistical tests used a 0.05 significancelevel, with no correction for multiple comparisons.

3. Results

Of the 234 patients included in the analysis 61% were female,81% were Caucasian, 42% had Bipolar I Disorder, 47% had Bipolar IIDisorder, and 11% had Bipolar Disorder Not Otherwise Specified.The mean � standard deviation age was 40.3 � 14.0 years, andmean BMI was 26.8 � 6.4. Demographic and clinical characteristicsof all 234 patients included in the analysis are presented by genderin Table 1. Patients in our sample tended to be relatively affluent,well-educated, and Caucasian. No significant differences werefound between men and womenwith regard to age, BMI, ethnicity,educational attainment, smoking habits, TSH, bipolar illnesssubtype, illness severity or duration, or weight-liable medicationexposure. Mean lipid values for men and women are presented inTable 2 and Fig.1. Prevalence of abnormal lipid values are presentedin Table 3. Compared to men, women had significantly lower meantriglycerides (105.58 � 64.12 vs. 137.99 � 105.14 mg/dl, p ¼ 0.009),LDL (109.84 � 33.47 vs.123.79 � 35.96 mg/dl, p ¼ 0.003) and TG/HDL ratio (1.98 � 1.73 vs.3.59 � 3.14, p < 0.001), and significantly

Table 1Demographic and clinical characteristics of study sample by gender (N ¼ 234).

Men(n ¼ 91)

Women(n ¼ 143)

Statistic p value

Mean � SD

Age (yrs) 40.59 � 14.66 40.05 � 13.66 t ¼ 0.287 0.774BMI 27.61 � 6.06 26.33 � 6.53 t ¼ 1.444 0.150CGI 2.71 � 1.16 2.8 � 12.1 t ¼ 0.544 0.587Yrs since first

episode21.36 � 15.11 22.03 � 12.93 t ¼ 0.363 0.717

n (%)Caucasian 72 (79%) 118 (83%) c2 ¼ 0.420 0.517College graduates 30 (33%) 46 (32%) c2 ¼ 0.068 0.794Daily smokers 13 (14%) 19 (13%) c2 ¼ 0.054 0.816

Bipolar typeBipolar I 41 (45%) 57 (40%) c2 ¼ 5.760 0.056Bipolar II 35 (38%) 75 (52%)Bipolar NOS 14 (15%) 11 (7%)

TSH category>Reference range 2 (2%) 4 (3%) c2 ¼ 0.303 0.859Normal range 60 (66%) 93 (65%)<Reference range 6 (7%) 12 (8%)

Medicationcategorya

WL MS or AP 54 (61%) 80(59%) c2 ¼ 0.125 0.939NWL MS or AP 13 (15%) 22 (16%)No mood MSor AP

21 (24%) 3 (24%)

Weight-liable (WM) mood stabilizer (MS) or antipsychotic (AP): lithium, valproate,oxcarbazepine, olanzapine, clozapine, quetiapine, risperidone, or aripiprazole. Non-weight-liable (NWL) MS or AP: lamotrigine, carbamazepine, or ziprasidone. Cate-gories were not mutually exclusive.

a Medication status categories refers to presence or absence of medication for �6months prior to venipuncture.

M. Vemuri et al. / Journal of Psychiatric Research 45 (2011) 1036e10411038

higher mean HDL (60.17� 17.56 vs. 46.07� 11.91, p< 0.001), but nosignificant difference in total cholesterol. Compared to men,women had a significantly lower prevalence of abnormal totalcholesterol, LDL, TG, HDL, and TG/HDL ratio.

Statistical analyses found that age, BMI, bipolar illness subtype,and years since first episode correlated significantly with at leastone blood lipid level. These variables were subsequently includedas covariates in the final MANOVA model of effects of genderdifferences on the dependent variables TC, TG, HDL, LDL, andTG/HDL Results showed that after controlling for age, BMI, bipolartype, and illness duration, there was still a significant main effectof gender on TG [F(1, 28591) ¼ 5.295; p ¼ 0.022]; HDL,[F(1,6567)¼ 32.598; p< 0.001]; LDL, [F(1,5391)¼ 4.937; p¼ 0.027];and TG/HDL, [F(1,68) ¼ 14.380; p¼<0.001]. Specifically, womenhad lower TG, LDL, and TG/HDL, and higher HDL than men (seeTable 1). Years since first episode also showed a significant effect onLDL levels [F(1,4820) ¼ 4.414; p ¼ 0.037] and there were significantBMI effects on TG, [F(1,112985) ¼ 20.924; p < 0.001], HDL,[F(1,5308)¼ 26.350; p< 0.001]. LDL, [F(1,6659)¼ 6.098; P¼ 0.014],

Table 2Mean lipid values by gender.

Men n Women n t value pvalue

Mean � SD Mean � SD

Total cholesterol(mg/dl)

195.68 � 42.07 90 189.58 � 40.80 143 t ¼ 1.097 0.274

Triglycerides(mg/dL)

137.99 � 105.14 91 105.58 � 64.12 139 t ¼ 2.637 0.009

HDL (mg/dl) 46.07 � 11.91 91 60.17 � 17.56 137 t ¼ 7.227 <0.001LDL (mg/dl) 123.79 � 35.96 82 109.84 � 33.47 134 t ¼ 2.891 0.004TG/HDL 3.59 � 3.14 91 1.98 � 1.73 137 t ¼ 3.942 <0.001

and TG/HDL [F(1,126) ¼ 26.479; p < 0.001]. Age had a significanteffect on TC, [F(1,13640) ¼ 8.503; p ¼ 0.004], HDL, [F(1,939)¼ 4.660; p¼ 0.32], and LDL levels [F(1,6093)¼ 5.580; p¼ 0.19].

4. Discussion

To our knowledge this is the first study to compare compre-hensive lipid profiles (including TC, TG, LDL, HDL, and TG/HDL)between men and women with BD. The main finding in this studywas that women compared to men with BD have more favorablelipid profiles (lower TG, LDL, and TG/HDL and higher HDL). Simi-larly, the prevalence of abnormal lipid values was lower in womencompared to men. The association between gender and lipid levelsheld evenwhen controlling for age, BMI, Illness duration, or bipolarillness subtype, weight-liable medication exposure, illness severity,TSH, nicotine use and demographic variables.

In the past the influence of gender on dyslipidemia in BD hasbeen understudied. Comparisons of dyslipidemia by gender inpsychiatric patients occur most commonly as subcomponentanalyses within studies of MetS. Such comparisons includemeasures of TG and HDL, but not TC or LDL, as NCEP ATP-IIIguidelines for diagnosis of MetS call for the presence of three ormore of the following: abdominal obesity, elevated TG, low HDL,elevated blood pressure, and elevated fasting glucose, but do notinclude TC and LDL (NCEP, 2002). Similar to our results, Sicras et alfoundmore favorable lipid measurese higher mean HDL and lowermean TG in women with BD compared to men (Sicras et al., 2008),but found no difference in prevalence of elevated TG or low HDLbetween gender. This discrepancy may be related to differences inthresholds for abnormally lowHDL inmen andwomen (40mg/dl inmen, 50 mg/dl in women (NCEP, 2002)). The remaining studies ofMetS in BD that included gender differences in lipid valuesprimarily evaluated prevalence of abnormal components categor-ically, and did not include gender-related differences in mean lipidconcentrations as continuous parameters. Two studies found nogender-related differences in prevalence of abnormal TG (Garcia-Portilla et al., 2008; Van Winkel et al., 2008), or HDL (Van Winkelet al., 2008), perhaps related to lower statistical power achievedby categorical rather than continuous analyses. In contrast to theresults of our study, Garcia-Portilla et al found that womencompared to men with BD had a higher prevalence of abnormallylow HDL (Garcia-Portilla et al., 2008).

Studies of dyslipidemia in other psychiatric populations havesuggested a higher or similar prevalence of abnormally low HDL inwomen versus men, but a lower or similar prevalence of abnor-mally high TG in women versus men. In patients with a lifetimehistory of a major depressive episode, women compared to menwere more likely to have abnormally low HDL, and less likely tohave abnormally high TG (Kinder et al., 2004). Likewise, inschizophrenia patients evaluated for MetS in the Clinical Antipsy-chotic Trials of Intervention Effectiveness (CATIE), womencompared to men were more likely than to have abnormally lowHDL, and less likely to have abnormally high TG (McEvoy et al.,2005). Also, a higher prevalence of abnormally low HDL inwomen compared to men was seen in a general psychiatric pop-ulation with heterogeneous diagnoses receiving antipsychotictherapy (Sicras-Mainar et al., 2008). Gender-related differences inTC or LDL are largely unstudied in psychiatric populations.

Our findings did not support the hypothesis that women withBD would have greater dyslipidemia than man, as has beenobserved in other general psychiatric populations with regards toCVD risk factors. However, our results are consistent with genderdifferences in lipids observed in the general population in the U.S.and Europe (Gardner et al., 2000; Razay et al., 1992; Seidell et al.,1991). One explanation for these differences is that in some

Fig. 1. Unadjusted group means � standard error for each assessed lipid value (Total cholesterol, Triglyceride (TG), LDL cholesterol, HDL cholesterol, TG/HDL by gender).

M. Vemuri et al. / Journal of Psychiatric Research 45 (2011) 1036e1041 1039

populations, such as individuals with certain psychiatric illnesses,women compared to men may be equally or more vulnerable toobesity, abdominal fat deposition and to consequent dyslipidemia.However, in our sample of relatively affluent and educated bipolarpatients women compared to men had a statistically similar rate ofobesity, perhaps accounting for gender-related differences in lipidshaving a pattern similar to that seen in the general population.

In our study, no significant differences were found in BMI or anylipid value by weight liability of medication in either women ormen. This finding was in contrast to evidence suggesting a rela-tionship between weight-liable medications, such as olanzapineand clozapine, to weight-gain and dyslipidemia (Birkenaes et al.,2008; Weiden, 2007), and evidence demonstrating lower HDL

Table 3Prevalence of abnormal lipid values by gender.

Women Men c2 p value

% (n/N)

Total cholesterol �200 mg/dl 36 (51/143) 49 (44/90) 4.0 0.045LDL �160 mg/dl 9 (12/134) 18 (15/82) 4.055 0.044TG �150 mg/dl 19 (27/139) 31 (28/91) 3.890 0.049HDL <40 (men) or <50 (women) 12 (16/137) 64 (58/91) 67.594 <0.011TG/HDL �3.5 12 (16/137) 34 (31/91) 16.746 <0.001

and higher TG values in women taking olanzapine or clozapineversus those not taking antipsychotics (Birkenaes et al., 2008). Theabsence of such findings in our study may be related to not onlypatient (primarily Caucasian race and life-style decisions of ourwell-educated and relatively affluent sample), but also clinician(vigilant monitoring of BMI, careful consideration when selectingweight-liable versus non-weight-liable medications, and pharma-cological efforts to manage weight-gain) factors. Such patient andclinician factors may also be reflected in the low mean BMI of 26.8found in our study relative to the mean BMI of 27.7 found across allSTEP-BD sites (Wang et al., 2006).

In our study, BMI was also associated with TG, HDL, LDL, and TG/HDL, controlling for gender, age, illness duration and bipolar type.These findings are consistent with other studies demonstratinga relationship between obesity and MetS in patients with bipolardisorder (Fiedorowicz et al., 2008; Garcia-Portilla et al., 2008; VanWinkel et al., 2008) and in the general population (Gostynski et al.,2004). Controlling for age, gender, diagnosis and BMI, an associa-tion between years since first episode and LDL was found in thisstudy. This may suggest that chronic mental illness or treatment forchronic mental illness may play a role in dyslipidemia.

This study had some noteworthy limitations. First, our relativelyaffluent and well-educated primarily Caucasian patients likelymade life-style decisions that would not permit generalization of

M. Vemuri et al. / Journal of Psychiatric Research 45 (2011) 1036e10411040

our findings to individuals with less education and financialresources. Information about dietary practices, alcohol use andexercise levels were not ascertained. Treatment was naturalisticrather than randomized and lipid values were obtained at cliniciandiscretion. Therefore it is possible that a gender selection biasoccurred with respect to medication choice and available lipidparameters. The lack of a healthy control group did not allow forassessment of gender differences in comparison to healthy indi-viduals without bipolar disorder. Waist circumference may havebeen an important covariate in this analysis, but these data werenot available. Many of our patients were treated with multiplemood stabilizers/antipsychotics of unspecified duration, makingthe contribution of medication type and duration to dyslipidemiachallenging to ascertain. Likewise, patients may have been treatedconcomitantly with antidepressants, possibly affecting dyslipide-mia, which were not accounted for in this study. Finally, the effectsof menopausal status and use of exogenous hormones were nottaken into account.

Future controlled studies are needed to evaluate gender differ-ences in dyslipidemia in bipolar disorder, with attention to concom-itant hormone use and menopausal status, and with attention todifferences within specific age and BMI subgroups. Of particularimportance, research is needed to assess whether or not Caucasianrace and life-style decisions made by relatively affluent and well-educated women with bipolar disorder protect them from obesity,dyslipidemia, and metabolic problems reported in other studies.

Contributors

Natalie L. Rasgon, Terence A. Ketter designed the study. Diag-nosis of patients and clinical data was performed and collected byTerence A. Ketter and Po W. Wang. Heather A. Kenna performedstatistical analysis. Mytilee Vemuri wrote the first draft of themanuscript. All other authors contributed to and have approvedthe final manuscript.

Role of funding source

Dr. Vemuri’s scholarship was supported by NIMH T-32 TrainingGrant. The NIMH had no further role in study design, in thecollection, analysis and interpretation of data, in the writing of thereport, and in the decision to submit the paper for publication.

Conflict of interestDr. Rasgon has received grant/research support from Bayer

Pharmaceuticals, Abbott Laboratories, GlaxoSmithKline, ForrestLaboratories, Pfizer Inc, Wyeth-Arest; has served as a consultant forForest Laboratories and Wyeth-Arest; and has received lecturehonoraria from Bristol-Myers Squibb Corp., and Pfizer Inc.

Dr. Ketter has received grant/research support from AbbottLaboratories, Inc. AstraZeneca Pharmaceuticals LP, Bristol-MyersSquibb Co., Cephalon Inc., Eli Lilly and Co., GlaxoSmithKline, PfizerInc., Repligen Corporation, Wyeth Pharmaceuticals; has served asa consultant for Abbott Laboratories Inc., AstraZeneca Pharmaceuti-cals LP, Astellas Pharmaceuticals, Bristol-Myers Squibb Co., CephalonInc., Dainippon Sumitomo Pharmaceuticals, Eli Lilly and Co., Glax-oSmithKline, Janssen Pharmaceutical ProductsLP, Jazz Pharmaceuti-cals Inc., Novartis Pharmaceuticals Corp., Organon International Inc.,a part of Schering-Plough Corp., Sepracor Inc., Solvay Pharmaceuti-cals Inc., Valeant Pharmaceuticals, Vanda Pharmaceuticals, WyethPharmaceuticals, XenoPort, Inc. and has received lecture honorariafrom Abbott Laboratories Inc., AstraZeneca Pharmaceuticals LP,Bristol-Myers Squibb Co., Eli Lilly and Co., GlaxoSmithKline, NovenPharmaceuticals, Otsuka Pharmaceuticals, Pfizer Inc.

Dr. Wang has received grant/research support from AbbottLaboratories AstraZeneca, Bristol-Myers Squibb, Eisai Inc., ElanPharmaceuticals, Inc., Eli Lilly and Co., GlaxoSmithKline, JanssenPharmaceutica Products LP, Novartis Pharmaceuticals Corp., Repli-gen Corp., Shire Pharmaceuticals Group, Solvay PharmaceuticalsInc.. Wyeth Pharmaceuticals; served as a consultant for AbbottLaboratories, Corcept Therapeutics, Pfizer, Sanofi-Aventis andreceived lecture honoraria from Abbott Laboratories, AstraZeneca,Briston-Myers Squibb, Eli Lilly and Co., GlaxoSmithKline, Pfizer, andSanofi-Aventis.

Acknowledgement

We thank Anna Morenkova, M.D. and Dr. Uma Saha, M.D. forassistance with data processing

References

Allison DB, Mentore JL, Heo M, Chandler LP, Cappelleri JC, nfante MC, et al. Anti-psychotic-induced weight gain: a comprehensive research synthesis. AmericanJournal of Psychiatry 1999;156:1686e96.

American Heart Association. Cholesterol statistics. American Heart Association;2010.

Birkenaes A, Birkeland K, Engh J, Farden A. Dyslipidemia independent of body massin antipsychotic-treated patients under Real-life conditions. Journal of ClinicalPsychopharmacology 2008;28:132e7.

Bowden CL, Calabrese JR, McElroy SL, Gyulai L, Wassef A, Petty F, et al.A randomized, placebo-controlled 12-month trial of divalproex and lithium intreatment of outpatients with bipolar I disorder. Divalproex Maintenance StudyGroup. Archives of General Psychiatry 2000;57:481e9.

Bowden CL, Vieta E, et al. A 6-month, randomized, placebo-controlled, double-blindtrial of ziprasidone plus a mood stabilizer in subjects with bipolar I disorder. In:2009, Eighth International Conference on Bipolar Disorder, Pittsburgh, June25e27. 2009.

Dennehy EB, Bauer MS, Perlis RH, Kogan JN, Sachs GS. Concordance with treatmentguidelines for bipolar disorder: data from the systematic treatment enhance-ment program for bipolar disorder. Psychopharmacology Bulletin2007;40:72e84.

Fagiolini A, Frank E, Scott JA, Turkin S, Kupfer DJ. Metabolic syndrome in bipolardisorder: findings from the Bipolar Disorder Center for Pennsylvanians. BipolarDisorder 2005;7:424e30.

Fiedorowicz JG, Palagummi NM, Forman-Hoffman VL, Miller del D, Haynes WG.Elevated prevalence of obesity, metabolic syndrome, and cardiovascular riskfactors in bipolar disorder. Annals of Clinical Psychiatry 2008;20:131e7.

Garcia-Portilla MP, Saiz PA, Benabarre A, Sierra P, Perez J, Rodriguez A, et al. Theprevalence of metabolic syndrome in patients with bipolar disorder. Journal ofAffective Disorders 2008;106:197e201.

Gardner CD, Winkleby MA, Fortmann SP. Population frequency distribution of non-high-density lipoprotein cholesterol (Third National Health and NutritionExamination Survey [NHANES III], 1988e1994). American Journal of Cardiology2000;86:299e304.

Gostynski M, Gutzwiller F, Kuulasmaa K, Doring A, Ferrario M, Grafnetter D, et al.Analysis of the relationship between total cholesterol, age, body mass indexamong males and females in the WHO MONICA Project. Int J Obes Relat MetabDisord 2004;28:1082e90.

Keck Jr PE, Calabrese JR, McQuade RD, Carson WH, Carlson BX, Rollin LM, et al.A randomized, double-blind, placebo-controlled 26-week trial of aripiprazole inrecently manic patients with bipolar I disorder. Journal of Clinical Psychiatry2006;67:626e37.

Kilbourne AM, Cornelius JR, Han X, Pincus HA, Shad M, Salloum I, et al. Burden ofgeneral medical conditions among individuals with bipolar disorder. BipolarDisorder 2004;6:368e73.

Kinder LS, Carnethon MR, Palaniappan LP, King AC, Fortmann SP. Depression andthe metabolic syndrome in young adults: findings from the third NationalHealth and Nutrition Examination Survey. Psychosomatic Medicine2004;66:316e22.

Ley CJ, Lees B, Stevenson JC. Sex- and menopause-associated changes in body-fatdistribution. American Journal of Clinical Nutrition 1992;55:950e4.

McEvoy JP, Meyer JM, Goff DC, Nasrallah HA, Davis SM, Sullivan L, et al. Prevalenceof the metabolic syndrome in patients with schizophrenia: baseline resultsfrom the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE)schizophrenia trial and comparison with national estimates from NHANES III.Schizophrenia Research 2005;80:19e32.

McLaughlin T, Reaven G, Abbasi F, Lamendola C, Saad M, Waters D, et al. Is therea simple way to identify insulin-resistant individuals at increased risk ofcardiovascular disease? American Journal of Cardiology 2005;96:399e404.

NCEP. Third report of the National Cholesterol Education Program (NCEP) expertpanel on detection, evaluationand treatment of high blood cholesterol in adults(Adult Treatment Panel III) final report. Circulation 2002;106:3143e421.

M. Vemuri et al. / Journal of Psychiatric Research 45 (2011) 1036e1041 1041

Newcomer JW. Comparing the safety and efficacy of atypical antipsychotics inpsychiatric patients with comorbid medical illnesses. Journal of ClinicalPsychiatry 2009;70(Suppl. 3):30e6.

Quiroz JA, Yatham LN, Palumbo JM, Karcher K, Kushner S, Kusumakar V. Risperidonelong-acting injectable monotherapy in the maintenance treatment of bipolar Idisorder. Biological Psychiatry 2010;68:156e62.

Rasgon N, Jarvik L. Insulin resistance, affective disorders, and Alzheimer’s disease:review and hypothesis. The Journals of Gerontology, Series A: Medical Sciences2004;59:178e83. discussion 184e192.

Rasgon NL, Carter MS, Elman S, Bauer M, Love M, Korenman SG. Common treatmentof polycystic ovarian syndrome and major depressive disorder: case report andreview. Curr Drug Targets Immune Endocr Metabol Disord 2002;2:97e102.

Razay G, Heaton KW, Bolton CH. Coronary heart disease risk factors in relation tothe menopause. QJM 1992;85:889e96.

Regitz-Zagrosek V, Lehmkuhl E, Mahmoodzadeh S. Gender aspects of the role of themetabolic syndrome as a risk factor for cardiovascular disease. Gender Medi-cine 2007;4(Suppl. B):S162e77.

Sachs G, Bowden C, Calabrese JR, Ketter T, Thompson T, White R, Bentley B. Effects oflamotrigine and lithium on body weight during maintenance treatment ofbipolar I disorder. Bipolar Disorder 2006;8:175e81.

Sachs GS, Guille C, McMurrich SL. A clinical monitoring form for mood disorders.Bipolar Disorder 2002;4:323e7.

Sachs GS, Thase ME, Otto MW, Bauer M, Miklowitz D, Wisniewski SR, et al. Ratio-nale, design, and methods of the systematic treatment enhancement programfor bipolar disorder (STEP-BD). Biological Psychiatry 2003;53:1028e42.

Seidell JC, Cigolini M, Charzewska J, Ellsinger BM, Bjorntorp P, Hautvast JG. Fatdistribution and gender differences in serum lipids in men and women fromfour European communities. Atherosclerosis 1991;87:203e10.

Sicras A, Rejas J, Navarro R, Serrat J, Blanca M. Metabolic syndrome in bipolardisorder: a cross-sectional assessment of a Health Management Organizationdatabase. Bipolar Disord 2008;10:607e16.

Sicras-Mainar A, Blanca-Tamayo M, Rejas-Gutierrez J, Navarro-Artieda R. Metabolicsyndrome in outpatients receiving antipsychotic therapy in routine clinical

practice: a cross-sectional assessment of a primary health care database.European Psychiatry 2008;23:100e8.

Stemmle PG, Kenna HA, Wang PW, Hill SJ, Ketter TA, Rasgon NL. Insulin resistanceand hyperlipidemia in women with bipolar disorder. Journal of PsychiatricResearch 2009;43:341e3.

Teixeira PJ, Rocha FL. The prevalence of metabolic syndrome among psychiatricinpatients in Brazil. Revista Brasileira de Psiquiatria 2007;29:330e6.

Tohen M, Calabrese JR, Sachs GS, Banov MD, Detke HC, Risser R, et al. Randomized,placebo-controlled trial of olanzapine as maintenance therapy in patients withbipolar I disorder responding to acute treatment with olanzapine. AmericanJournal of Psychiatry 2006;163:247e56.

Van Winkel R, De Hert M, Hanssens L, Wampers M, Scheen A, Peuskens J. Preva-lence of diabetes and the metabolic syndrome in a sample of patients withbipolar disorder. Bipolar Disorder 2008;10:342e8.

Vieta E, CruzN, Garcia-Campayo J, deArceR,Manuel Crespo J, Valles V, et al. Adouble-blind, randomized, placebo-controlled prophylaxis trial of oxcarbazepine asadjunctive treatment to lithium in the long-term treatment of bipolar I and IIdisorder. International Journal of Neuropsychopharmacology 2008;11:445e52.

Wang PW, Sachs GS, Zarate CA, Marangell LB, Calabrese JR, Goldberg JF, et al.Overweight and obesity in bipolar disorders. Journal of Psychiatric Research2006;40:762e4.

Weiden PJ. Switching antipsychotics as a treatment strategy for antipsychotic-induced weight gain and dyslipidemia. Journal of Clinical Psychiatry2007;68(Suppl. 4):34e9.

Weisler RH, Hirschfeld R, Cutler AJ, Gazda T, Ketter TA, Keck PE, et al. Extended-release carbamazepine capsules as monotherapy in bipolar disorder: pooledresults from two randomised, double-blind, placebo-controlled trials. CNSDrugs 2006;20:219e31.

Weisler RH, Nolen WA. Quetiapine or lithium versus placebo for maintenancetreatment of bipolar I disorder after stabilization on Quetiapine. In: Interna-tional Conference on Bipolar Disorder, Pittsburgh, June 25e27, 2009.

Williams CM. Lipid metabolism in women. Proceedings of the Nutrition Society2004;63:153e60.