Journal of Affective Disorders 115 (2009) 302–308www.elsevier.com/locate/jad

Research report

Cardiovascular risk in patients with bipolar disorder

Maria P. Garcia-Portilla a,b,⁎, Pilar A. Saiz a,b, Maria T. Bascaran a,b, Sara Martíneza b,Antonio Benabarre b,c, Pilar Sierra d, Pedro Torres e, Jose Manuel Montes b,f,

Manuel Bousoño a,b, Julio Bobes a,b

on behalf of the General Health Status in Bipolar Disorder Collaborative Group 1

a Department of Psychiatry, University of Oviedo, Spainb Centro de Investigación Biomédica en Red de Salud Mental, CIBERSAM, Spain

c Hospital Clinic, Barcelona, Spaind Hospital La Fe, Valencia, Spaine Hospital Complex Jaen, Spain

f Department of Psychiatry, University of Alcalá de Henares, Spain

Received 24 July 2008; received in revised form 7 September 2008; accepted 8 September 2008Available online 26 October 2008

Abstract

Background: To date, little is known about cardiovascular risk (CVR) in terms of coronary heart disease (CHD) and cardiovascularmortality risk (CMR) in patients with bipolar disorder. This study provides data on the overall risk of any fatal or non-fatalcoronary heart disease (CHD) and on the cardiovascular mortality risk (CMR) within 10 years in these patients.Methods: Naturalistic, cross-sectional, multicenter study conducted in Spain. Patients were evaluated for cardiovascular risk usingthe Framinghan function (CHD) and the Systematic COronary Risk Evaluation (SCORE) function (CMR).Results: The mean age was 46.6 years and 49% were male. Forty-six percent were in remission. Ten-year CHD risk was 7.6%(males 10.2% versus females 4.7%, pb0.001) and 10-year CMR was 1.8% (males 2.2% versus females 1.3%, p 0.161). Fifty-onepercent smoked and 34% was obese. Metabolic syndrome was present in 22.4% of the sample (35.6% according to AHA andNHLBI criteria). Cardiovascular risk significantly increases with age, body mass index and presence of metabolic syndrome.Limitations: The cross-sectional design of the study.Conclusions: Cardiovascular risk is high in patients with bipolar disorder. It is associated with age, body mass index and metabolicsyndrome. Psychiatrists should be aware of this issue and carefully monitor these patients for cardiovascular risk factors, includingcigarette smoking, as part of the standard of care when treating them.© 2008 Elsevier B.V. All rights reserved.

Keywords: Bipolar disorder; Cardiovascular risk; Framinghan function; SCORE function

⁎ Corresponding author. Department of Psychiatry, School of Medicine, University of Oviedo–Centro de Investigación en Red de Salud Mental,CIBERSAM, Julian Claveria 6-33006 Oviedo, Spain. Tel.: +34 985 104219; fax: +34 985 103552.

E-mail address: albert@uniovi.es (M.P. Garcia-Portilla).1 The General Health Status in Bipolar Disorder Collaborative Group includes the following investigators: Cervilla J (Granada), Fernandez-

Villamor R (Sevilla), Franch J (Valladolid), Gonzalez-Pinto A (Vitoria), Livianos L (Valencia), Medina G (Valladolid), Menendez I (Asturias), MuñizMJ (Madrid), Perez J (Barcelona), Rodriguez A (Barcelona), Sarramea F (Jaen), Valle J (Madrid).

0165-0327/$ - see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.jad.2008.09.008

303M.P. Garcia-Portilla et al. / Journal of Affective Disorders 115 (2009) 302–308

1. Introduction

There is an increasing recognition that bipolar disorderis associated with elevated morbidity and mortality rates.Several studies have foundmortality rates between 1.5 and2.5 times higher than the general population. After suicideand accidents, cardiovascular and all vascular diseases arethe leading causes of death in these patients, with standard-izedmortality ratios (SMR) ranging from1.47 to 2.6 (Ösbyet al., 2007; Angst et al., 2002), although studies in patientsreceiving long-term lithium treatment did not find theseresults (Ahrens et al., 1995; Brodersen et al., 2000).

Recently, one study was published on cardiovascularrisk factors in patients with bipolar disorder (Birkenaeset al., 2007), demonstrating that they have a prevalenceof cardiovascular risk factors about twice as high as thegeneral population and equal to patients with schizo-phrenia. Similar results were found by Correll et al.(2008). These authors reported a 10-year CHD risk of4.7% for inpatients with bipolar disorder (n=74) and of6.5% for inpatients with schizophrenia (n=111). Nine-teen percent of patients with bipolar disorder and 23.4%of patients with schizophrenia had a CHD risk N10%.

Despite the paper of Correll et al. (2008), little isknown about the prevalence of cardiovascular risk(CVR) in terms of coronary heart disease (CHD) andcardiovascular mortality risk (CMR) in patients withbipolar disorder even though these should be used toguide decision-making on clinical interventions.Furthermore, recent European guidelines on cardiovas-cular disease prevention in clinical practice (De Backeret al., 2003) recommend the identification of asympto-matic individuals who are at high absolute risk as a wayto raise the efficiency of the preventive efforts, andaccording to Newcomer (2006) patients with bipolardisorder should be included in this group.

The Framingham Heart Study (Wilson et al., 1998)and the SCORE project (Conroy et al., 2003) identifiedkey modifiable risk factors for developing CHD andCMR such as dyslipidemia, hyperglycemia, hyperten-sion, obesity and smoking. Several studies (McElroyet al., 2002; Uçok et al., 2004; Fagiolini et al., 2005;Newcomer 2006; Yumru et al., 2007; Correll et al., 2008;Garcia-Portilla et al., 2008; vanWinkel et al., 2008) haveshown that patients with bipolar disorder have anincreased prevalence of these modifiable risk factors,suggesting there is an opportunity to lower their risk.

In addition, individuals with metabolic syndromehave a 2.01 odds ratio for myocardial infarction and a2.16 odds ratio for stroke (Ninomiya et al., 2004). Inpatients with bipolar disorder, metabolic syndrome wasfound to be more prevalent than in the general popu-

lation (Fagiolini et al., 2005; Fagiolini et al., 2008;Yumru et al., 2007; Correll et al., 2008; Garcia-Portillaet al., 2008; Salvi et al., 2008; van Winkel et al., 2008).

In this paper, we report information about the car-diovascular risk in terms of CHD and CMR in patientswith bipolar disorder receiving maintenance treatmentin our setting.

2. Methods

The “General Health Status in Bipolar Disorders”study is a naturalistic, cross-sectional study conducted at13 centers in Spain. The study was approved by theInternal Review Board of one of the centers and is inaccordance with the 1975 Declaration of Helsinki.

2.1. Patients

The study enrolled 194 patients with bipolar disorder,≥18 years of age who consecutively kept their appoint-ments during the period between April and September2006 and gave their inform consent. Diagnoses weredetermined by experienced psychiatrists using ICD-10criteria (WHO, 1992). All subjects were of CaucasianSpanish origin. Exclusion criteria were designed to beminimal according to the naturalistic design and onlyincluded refusal to give informed consent.

2.2. Data collection and measures

Demographic and clinical information were collectedfrom the records of the patients.

Blood pressure consisted of a single, seated determi-nation. Blood samples were collected if patients con-firmed having fasted for at least 8 h.Waist circumferencewas measured at the high point of the iliac crest and at thelevel of the umbilicus. Body mass index (BMI) wascalculated as body weight (kg) divided by height squared(m2). According to WHO overweight was defined asBMI≥25.00 and obesity as BMI≥30.00.

Metabolic syndrome (MetS) was defined accordingto the NHANES 1999–2000 study criteria (Ford et al.,2004) and to the American Heart Association/NationalHeart, Lung and Blood Institute Scientific Statement(Grundy et al., 2005) (Table 1).

Cardiovascular riskwas estimated using the SystematicCoronary Risk Evaluation (SCORE) function (Conroyet al., 2003) for cardiovascular mortality risk (CMR)(coronary death, sudden death, stroke, aortic aneurism,and heart failure) and the Framinghan function (Wilsonet al., 1998) was used to estimate the overall risk for anyfatal or non-fatal coronary heart disease (CHD) (the above

Table 1Criteria for clinical diagnosis of metabolic syndrome.

Any 3 of 5 constitute diagnosisof metabolic syndrome

NHANES 1999–2000 study criteria (Ford et al., 2004) AHA and NHLBI criteria (Grundy et al., 2005)

Elevated waist circumference N102 cm in men ≥102 cm in menN88 cm in women ≥88 cm in women

Elevated triglycerides ≥150 mg/dL or on lipid-lowering medication ≥150 mg/dL or on triglycerides-lowering medicationReduced HDL-C b40 mg/dL in men b40 mg/dL in men

b50 mg/dL in women b50 mg/dL in women or on drug treatment forreduced HDL-C

Elevated blood pressure Systolic pressure ≥130 and diastolic pressure≥85 mm Hg or on antihypertensive medication

≥130 mm Hg systolic pressure or ≥85 mm Hgdiastolic pressure or on antihypertensive drug treatment

Elevated fasting glucose ≥110 mg/dL or on antidiabetic medication ≥100 mg/dL or on drug treatment for elevated glucose

HDL-C: HDL cholesterol.

304 M.P. Garcia-Portilla et al. / Journal of Affective Disorders 115 (2009) 302–308

CMR factors plus any type of angina, myocardial infarc-tion, other type of coronary ischemia, congestive heartfailure, intermittent claudication, or peripheral arterialischemia) within 10 years. The SCORE function wasemployed since the Framingham algorithms overestimatethe risk of cardiovascular events in the lower-risk popula-tions of Southern Europe, i.e. Spain (Empana et al., 2003).

Both functions are mathematical probability modelsthat relate individual risk factors of each subject (age,gender, total cholesterol concentration, systolic bloodpressure and current smoking status, — presence ofdiabetes and HDL cholesterol are also included in theFramingham function) to the incidence of a fatal or non-fatal cardiovascular event. In addition, patients wereclassified as “very high/high” CHD risk if they had a

Table 2Individual prevalence of the main cardiovascular risk factors, 10-year CHD

Risk factors Total %

Age ≥40 (men) or ≥45 (women) 56.5Smoker 51.5Diabetes (type 1/2 or glucose ≥126 mg/dL) 6.5Hyperglycemia 9.5Total cholesterol ≥200 mg/dL 41.4HDL cholesterol b45 (men) or b50 (women) mg/dL 49.3SBP ≥140 mm Hg 19.8DBP ≥90 mm Hg 16.0

Mean (SD10-year CHD risk (Framingham) 7.57 (7.4)10-year CMR (SCORE) 1.78 (4.4)

%Very high/high 10-year CHD risk (Framingham ≥10%) 23.4High 10-year CMR (SCORE ≥5%) 8.2

CHD: Coronary Heart Disease.CMR: Cardiovascular Mortality Risk.DBP: Diastolic Blood Pressure.SBP: Systolic Blood Pressure.a Chi-square.b Student's t test.

Framingham score of ≥10% and “high” CMR risk ifthey had a SCORE of ≥5% (Hartz et al., 2005).

2.3. Statistical analysis

A statistical analysis was performed using SPSS/PC(version 15.0). The association between cardiovascularrisks and demographic and clinical variableswas examinedusing chi-square and Student's t tests. The adopted two-tailed level of significance was .05.

3. Results

Themean agewas 46.6 years (SD 13.9, 18 to 81 years);49.2% were male. Forty-six percent were in remission,

risk and CMR, in the total sample and by gender.

Males % Females % Statistics a — p

60.0 53.1 0.945–0.33164.2 39.8 11.517–b0.0055.4 7.7 0.405–0.5255.3 13.8 3.932–b0.0537.6 45.2 1.086–0.29753.8 44.6 1.300–0.25424.7 15.1 2.733–0.09819.4 12.9 1.431–0.232

) Mean (SD) Mean (SD) Statistics b — p10.20 (8.6) 4.67 (4.4) 4.805–b0.0012.23 (3.8) 1.32 (4.8) 1.406–0.161

% % Statistics a — p34.2 11.6 10.306–b0.0112.1 4.4 3.560–0.059

Table 3Prevalence of very high/high 10-year CHD risk and high CMR,according to age group and overall, in the total sample and by gender.

Age group Total%

Males%

Females%

Statistics a — p

Very high/high CHD risk(Framingham ≥10%)

b c d

b35 years 0 0 0 NP35–44 years 5.6 13.3 0 2.965–0.08545–54 years 32.3 64.3 5.9 11.984–b0.0155–64 years 57.6 59.1 54.5 0.062–0.803≥65 years 25.0 66.7 11.1 3.704–0.054Overall 23.4 34.2 11.6 10.306–b0.01

High CMR(SCORE ≥5%)

e f g

b35 years 0 0 0 NP35–44 years 0 0 0 NP45–54 years 0 0 0 NP55–64 years 14.3 23.1 0 4.308–b0.05≥65 years 52.9 83.3 36.4 3.438–0.064Overall 8.2 12.1 4.4 3.560–0.059

CHD: Coronary Heart Disease.CMR: Cardiovascular Mortality Risk.NP: statistical significance could not be calculated because all patientsbelong to the same patient risk group.a Chi-square test between males and females.b Chi-square test among age groups=39.297, pb0.001.c Chi-square test among age groups=27.426, pb0.001.d Chi-square test among age groups=24.538, pb0.001.e Chi-square test among age groups=57.991, pb0.001.f Chi-square test among age groups=39.726, pb0.001.g Chi-square test among age groups=30.428, pb0.001.

Table 4Prevalence of very high/high 10-year CHD risk and high CMR,according to the presence of metabolic syndrome, in the total sampleand by gender.

Age group Total%

Males%

Females%

Statistics a — p

Very high/high CHD risk(Framingham ≥10%)

b c d

MetS− 15.2 26.2 2.0 12.707–b0.001MetS+ 53.1 66.7 41.2 2.079–0.149

High CMR (SCORE ≥5%) e f g

MetS− 6.1 8.7 4.8 0.960–0.327MetS+ 15.8 29.4 0.0 2.601–0.107

MetS−: no metabolic syndrome; MetS+: presence of metabolicsyndrome.CHD: Coronary Heart Disease.CMR: Cardiovascular Mortality Risk.a Chi-square test between males and females.b Chi-square test between MetS− and MetS+=19.870, pb0.001.c Chi-square test between MetS− and MetS+=8.747, pb0.005.d Chi-square test between MetS− and MetS+=18.889, pb0.001.e Chi-square test between MetS− and MetS+=3.635, p 0.057.f Chi-square test between MetS− and MetS+=5.248, pb0.05.g Chi-square test between MetS− and MetS+=0.349, p 0.555.

305M.P. Garcia-Portilla et al. / Journal of Affective Disorders 115 (2009) 302–308

16.8% were depressed (mild/moderate), 11.5% werehypomanic and 11% manic with psychotic symptoms.The mean length of illness was 173.3 (SD 116.6) monthsand in 64.1% of the patients the length of illness was≥10 years. The mean Clinical Global Impression-Severity (CGI-S) score was 2.5 (SD 1.4). On average,patients were receiving 2.9 (SD 1.3) drugs for treatment oftheir bipolar disorder on enrolment in the study. Fifty-ninepercent were receiving one, 30.1% two, and 1.6% threemood stabilizers. Lithium was used by 64.1% of patientsreceiving mood stabilizers. Fifty-seven percent of thesample was receiving one antipsychotic, 8.3% two, and0.5% three antipsychotic medications. Ninety percent ofthe antipsychotics prescribed were atypical. Twenty-fivepercent of the sample was receiving one antidepressant,3.1% two, and 0.5% three different antidepressants.

3.1. Cardiovascular risk factors

The main cardiovascular risk factors are shown inTable 2. The mean body mass index (BMI) was 27.9

(SD 4.6) kg/m2. By categories, 34.7% of the patientswere classified as overweight and 34.1% as obese.

The overall prevalence of metabolic syndrome was22.4% according to NHANES 1999–2000 study criteriaand 35.6% according to AHA and NHLBI criteria.

3.2. CHD risk (Framingham function) and CMR(SCORE function)

Table 2 shows the CHD and CMR 10-year risks. Theoverall risk of CHD was 7.6% (SD 7.4) and wassignificantly higher in males than in females (10.2%versus 4.7, pb0.001). In addition, there was asignificantly greater proportion of males in the veryhigh/high CHD (34.2% versus 11.6%, pb0.01) groups.The overall CMR was 1.8% (SD 4.4). We did not findstatistically significant differences in CMR by gender.

The risk of CHD and the CMR in ten years signif-icantly increased with age in both males and females(pb0.001), with the exception of females aged 65 ormore. The proportion of patients considered high CMRand very high/high CHD risk significantly increasedwith age as well (Table 3).

Neither length of illness nor bodymass index category(normal weight, overweight and obesity) were associatedwithCMRorCHD risk. However, when consideringBMIdimensionally significant correlations were foundbetween BMI and CMR (Pearson's r=0.165, p 0.03)

306 M.P. Garcia-Portilla et al. / Journal of Affective Disorders 115 (2009) 302–308

and CHD risk (Pearson's r=0.227, p 0.008). In addition,in males, body mass index was significantly greater inthose with very high/high CHD risk than with lower/moderate CHD risk (29.93 kg/m2 versus 27.68, pb0.05).

Patients with metabolic syndrome (NHANES 1999–2000 study criteria) had significantly greater CMR (2.9versus 1.2, pb0.05) and CHD risk (12.9 versus 6.0,pb0.001). Males with metabolic syndrome had sig-nificantly greater CHD risk (17.8 versus 8.3, pb0.005),but not CMR. In the case of females, both CMR andCHD risk were significantly greater in those withmetabolic syndrome (1.6 versus 0.5, pb0.05 and 8.7versus 3.3, pb0.005, respectively). Identical resultshave been obtained using AHA and NHLBI criteria(data not shown). In addition, the presence of metabolicsyndrome was significantly associated with belongingto the high CMR and very high/high CHD risk groups(Table 4).

4. Discussion

Our results indicate that patients with bipolar disorderhave a high risk of CHD and CMR. The risk profile ischaracterized by high rates of cigarette smoking, obesity,and metabolic syndrome, increased levels of total cho-lesterol, and reduced HDL-cholesterol levels.

Compared to the study on CHD risk in bipolar patients(Correll et al., 2008) the prevalence of coronary heartdisease found in our study was higher (7.6 versus 4.7%),particularly in males (10.2 versus 6.0%). This is an aston-ishing result taking into account that the prevalence ofmetabolic syndromewas almost the double in the study ofCorrell et al. (2008), and the mean age, body mass index,proportion of males and smokers were similar in bothstudies (46.6 versus 44.4 years, 27.9 versus 27.1 kg/m2,50.8% versus 48.6%, and 51.5% versus 53.4%). Oneexplanation may be that a smaller proportion of ourpatients seem to have metabolic abnormalities and hyper-tension. However, the more plausible explanation isrelated to the well-known fact that the Framingham func-tion overestimates the CHD risk in Southern Europeancountries (Empana et al., 2003). Taking this into account,we strongly recommend the employment of the SCOREfunction when determining cardiovascular risks inpatients with severe mental disorders from countrieswith low cardiovascular disease risk, in order to avoidoverestimations.

The Spanish CLAMORS study (Bobes et al., 2007)conducted in patients with schizophrenia reported similarresults for CHD risk, 10.2% for males and 4.7% forfemales, but lower results for CMR (males: 2.2% versus1.1%; females: 1.3% versus 0.5%). This difference may

be due to the fact that patients with bipolar disorder wereolder than the patients with schizophrenia. Correll et al.(2008) in two samples closely matched for age and race,and balanced for gender, smoking status and history ofdiabetes and past CHD events, found similar CHD risk inpatients with schizophrenia compared to patients withbipolar disorder (6.5 versus 4.7%, p 0.15).

Several studies (Birkenaes et al., 2007; McElroy et al.,2002; Uçok et al., 2004; Fagiolini et al., 2005; Fagioliniet al., 2008; Newcomer 2006; Yumru et al., 2007; Correllet al., 2008; Garcia-Portilla et al., 2008; van Winkel et al.,2008) have reported high rates of the main cardiovascularrisk factors in patientswith bipolar disorder,which is in linewith our findings. Similar rates of smoking – 50–55% –(Uçok et al., 2004; Birkenaes et al., 2007; Correll et al.,2008), diabetes – 5.5–6.7% – (Birkenaes et al., 2007;van Winkel et al., 2008), metabolic syndrome – 18.3–25.3% – (Birkenaes et al., 2007; Salvi et al., 2008; vanWinkel et al., 2008) and mean BMI – 26.3–29.4 kg/m2 –(Birkenaes et al., 2007) were reported in studiesconducted in European countries (Belgium, Italy andNorway). Higher prevalence of metabolic syndrome wasreported in the United States by Fagiolini et al. (2005,2008), Cardenas et al. (2008) and Correll et al. (2008) –30%, 49% and 43.2% respectively – and in Turkey byYumru et al. (2007) – 32% – . All these findings supportthe claim made by Fleischhacker et al. (2008) about thereintegration of psychiatry and medicine in order toprovide optimal services to patients with severe mentalillness such as schizophrenia and bipolar disorder.

Compared with the prevalence of main cardiovascu-lar risk factors reported for the Spanish generalpopulation aged 25 to 64 years (Aranceta et al., 2004)our patients had a higher prevalence of cigarette smoking(males: 64.2% versus 48.1%; females: 39.8% versus30.2%), obesity (males: 27.1% versus 13.2%; females:24.1% versus 17.5%) and diabetes (only in females:7.7% versus 2.4%) and a lower prevalence of hyperch-olesterolemia (males: 37.6% versus 57.3%; females:45.2% versus 53%), even when adding those patientswho were on lipid-lowering medication and had normallevels of cholesterol (3.3% of the total sample, males:5.4%, females: 1.1%). However, we must be cautious ininterpreting these data, since age could be a confoundingfactor; 60% of the males were older than 39 years and53.1% of the females were older than 44 years.

Another important finding is that, in the studypopulation, the presence of metabolic syndrome, usingboth NHANES 1999–2000 study and AHA-NHLBIcriteria, was associated with higher cardiovascular risk,which confirms this well-known association (Ninomiyaet al., 2004; Bobes et al., 2007) in this population as

307M.P. Garcia-Portilla et al. / Journal of Affective Disorders 115 (2009) 302–308

well. Furthermore, body mass index significantly cor-related with cardiovascular risk. However, it is neces-sary to highlight that only a very small percentage of thecardiovascular risk is explained by this factor (5.1% ofthe variance in the case of CHD risk and 2.7% in thecase of CMR).

Our results demonstrate that Spanish patients withbipolar disorder are exposed to the same cardiovascularrisk level as Spanish patients with schizophrenia, and to ahigher risk level than the general population. Spanishpsychiatrists must be aware of this fact and carefullymonitor and control these patients for the different riskfactors, such as smoking, obesity, diabetes, dyslipidemia,and hypertension following the recommendations of theSpanish Consensus on Physical Health of Patients withBipolar Disorder Consensus (Bobes et al., 2008).Furthermore, there is a need for better coordinationbetween psychiatrists and general practitioners to controlthese modifiable risk factors and lower the cardiovascularrisk in patients with severe mental disorders. In addition,specific programs should be implemented for patientswith bipolar disorder, aimed at reducing cigarettesmoking, increasing exercise, and improving dietaryhabits, as in the case of patients with schizophrenia.

This study has some limitations. Firstly, the cross-sectional design reflects only a specific moment in time.Patients were not followed up 1 year later to verifywhether the cardiovascular risk remained similar orweather it increased or decreased. However, this studyprovides a good idea of the scope of the cardiovascularrisk and associated risk factors that clinicians need tomonitor when treating patients with bipolar disorder. Onthe other hand, this sample may not be representative ofSpanish patients with bipolar disorder, but the number ofpatients enrolled was large enough to merit taking theseresults into consideration. Nevertheless, this is the firststudy that determines the prevalence of cardiovascularrisk (CVR) in terms of coronary heart disease (CHD) andcardiovascular mortality risk (CMR) in patients withbipolar disorder in a population from a lower-risk country.

In conclusion, this study confirms the elevatedcardiovascular risk in patients with bipolar disorderand the high prevalence of associated risk factors.Cardiovascular risk is associated with age, body massindex and metabolic syndrome. Clinicians should beaware of this issue and monitor patients for cardiovas-cular risk factors, including cigarette smoking, as part ofthe standard of care when treating them.

Role of funding sourceSupported in part by the Institute de Salud Carlos III, Centro de

Investigación Biomédica en Red de Salud Mental, CIBERSAM.

Conflict of interestAntonio Benabarre has been a consultant for Pfizer Inc.Julio Bobes has been a consultant for Astra-Zeneca, Bristol-Myers

Squibb, Johnson & Johnson, Ely Lilly, and Pfizer Inc.All other authors declare that they have no conflicts of interest.

Acknowledgements

We would like to acknowledge the other investiga-tors who participated in the “General Health Status inBipolar Disorders” study: Cervilla J (Granada), Fernan-dez-Villamor R (Sevilla), Franch J (Valladolid), Gonza-lez-Pinto A (Vitoria), Livianos L (Valencia), Medina G(Valladolid), Menendez I (Asturias), Muñiz MJ(Madrid), Perez J (Barcelona), Rodriguez A (Barce-lona), Sarramea F (Jaen), Valle J (Madrid).

References

Ahrens, B., Müller-Oerlinghausen, B., Schou, M., Wolf, T., Alda, M.,Grof, E., Grof, P., Lenz, G., Simhandl, C., Thau, K., Vestergaard,P., Wolf, R., Möller, H.J., 1995. Excess cardiovascular and suicidemortality of affective disorders may be reduced by lithiumprofilaxis. J. Affect. Disord. 33, 67–75.

Angst, F., Stassen, H.H., Clayton, P.J., Angst, J., 2002. Mortality ofpatients with mood disorders: follow-up over 34–38 years. J. Affect.Disord. 68, 167–181.

Aranceta, J., et al., Grupo Colaborativo para el estudio DORICA faseII, 2004. Tablas de evolución del riesgo coronario adaptadas a lapoblación española. Estudio DORICA. Med. Clin. (Barc.) 123,686–691.

Birkenaes, A.B., Opjordsmoen, S., Brunborg, C., Engh, J.A.,Jonsdottir, H., Ringen, P.A., Simonsen, C., Vaskinn, A., Birkeland,K.I., Friis, S., Sundet, K., Andreassen, O.A., 2007. The level ofcardiovascular risk factors in bipolar disorder equals that ofschizophrenia: a comparative study. J. Clin. Psyhiatry 68, 917–923.

Bobes, J., et al., on behalf of the CLAMORS Study CollaborativeGroup, 2007. Cardiovascular and metabolic risk in outpatients withschizophrenia treated with antipsychotics: results of the CLA-MORS Study. Schizophr. Res. 90, 162–173.

Bobes, J., et al., on behalf of the Consensus Group of Experts onPhysical Health in Patients with Bipolar Disorders, 2008. SpanishConsensus on Physical Health of Patients with Bipolar Disorder.Revista de Psiquiatría y Salud Mental 1, 26–37.

Brodersen, A., Licht, R.W., Vestergaard, P., Olesen, A.V., Mortensen,P.B., 2000. Sixteen-year mortality in patients with affective dis-order commenced on lithium. Br. J. Psychiatry 176, 429–433.

Cardenas, J., Frye, M.A., Marusak, S.L., Levander, E.M., Chirichigno,J.W., Lewis, S., Nakelsky, S., Hwang, S., Mintz, J., Altshuler, L.L.,2008. Modal subcomponents of metabolic syndrome in patientswith bipolar disorder. J. Affect. Disord. 106, 91–97.

Conroy, R.M., et al., SCORE project group, 2003. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCOREproject. Eur. Heart J. 24, 987–1003.

Correll, C.U., Frederickson, A.M., Kane, J.M., Manu, P., 2008.Equally increased risk for metabolic syndrome in patients withbipolar disorder and schizophrenia treated with second-generationantipsychotics. Bipolar Disord. 10, 1–10.

De Baker, G., Ambrosini, E., Borch-Johnsen, K., Brotons, C., Cifkova,R., Dallongeville, J., Ebrahim, S., Faergeman, O., Graham, I.,

308 M.P. Garcia-Portilla et al. / Journal of Affective Disorders 115 (2009) 302–308

Mancia, G., Manger Cats, V., Orth-Gomer, K., Perk, J., Pyorala, K.,Rodicio, J.L., Sans, S., Sansoy, V., Sechtem, U., Silber, S., Thomsen,T., Wood, D., 2003. European guidelines on cardiovascular diseaseprevention in clinical practice. Third Joint Task Force of EuropeanandOther Societies on Cardiovascular Disease Prevention in ClinicalPractice. Eur. Heart J. 24, 1601–1610.

Empana, J.P., et al., PRIME Study Group, 2003. Are the Framinghamand PROCAM coronary heart disease risk functions applicable todifferent European populations? The PRIME Study. Eur. Heart J.24, 1903–1911.

Fagiolini, A., Frank, E., Scott, J.A., Turkin, S., Kupfer, D.J., 2005.Metabolic syndrome in bipolar disorder: findings from the BipolarDisorder Center for Pennsylvanians. Bipolar Disord. 7, 424–430.

Fagiolini, A., Frank, E., Turkin, S., Houck, P.R., Soreca, I., Kupfer, D.J.,2008. Metabolic syndrome in patients with bipolar disorder. J. Clin.Psychiatry 69, 678–679.

Fleischhacker, W.W., Cetkovich-Bakmas, M., De Hert, M., Henne-kens, C.H., Lambert, M., Leucht, S., Maj, M., McIntyre, R.S.,Naber, D., Newcomer, J.W., Olfson, M., Osby, U., Sartorius, N.,Lieberman, J.A., 2008. Comorbid somatic illnesses in patients withsevere mental disorders: clinical, policy, and research challenges.J. Clin. Psychiatry 69, 514–519.

Ford, E.S., Giles, W.H., Mokdad, A.H., 2004. Increasing prevalence ofthe metabolic syndrome among U.S. adults. Diabetes Care 27,2444–2449.

Garcia-Portilla, M.P., Saiz, P.A., Benabarre, A., Sierra, P., Perez, J.,Rodriguez, A., Livianos, L., Torres, P., Bobes, J., 2008. The prev-alence of metabolic syndrome in patients with bipolar disorder.J. Affect. Disord. 106, 197–201.

Grundy, S.M., Cleeman, J.I., Daniels, S.R., Donato, K.A., Eckel, R.H.,Franklin, B.A., Gordon, D.J., Krauss, R.M., Savage, P.J., Smith Jr.,S.C., Spertus, J.A., Costa, F., 2005. Diagnosis and management ofthe metabolic syndrome. An American Heart Association/NationalHeart, Lung, and Blood Institute Scientific Statement. Circulation112, 2735–2752.

Hartz, I., Njolstad, I., Eggen, A.E., 2005. Does implementation of theEuropean guidelines based on the SCORE model double the

number of Norwegian adults who need cardiovascular drugs forprimary prevention? The Tromso study 2001. Eur. Heart J. 26,2673–2680.

McElroy, S.L., Frye, M.A., Suppes, T., Dhavale, D., Keck Jr., P.E.,Leverich, G.S., Altshuler, L., Denicoff, K.D., Nolen, W.A., Kupka,R., Grunze, H., Walden, J., Post, R.M., 2002. Correlates of over-weight and obesity in 644 patients with bipolar disorder. J. Clin.Psychiatry 63, 207–213.

Newcomer, J.W., 2006. Medical risk in patients with bipolar disorderand schizophrenia. J. Clin. Psychiatry 67 (suppl 9), 25–30.

Ninomiya, J.K., L'Italien, G., Criqui, M.H., Whyte, J.L., Gamst, A.,Chen, R.S., 2004. Association of the metabolic syndrome withhistory of myocardial infarction and stroke in the Third NationalHealth and Nutrition Examination Survey. Circulation 109, 42–46.

Ösby, U., Brand, L., Correia, N., Ekbom, A., Sparén, P., 2007. Excessmortality in bipolar and unipolar disorder in Sweden. Arch. Gen.Psychiatry 58, 844–850.

Salvi, V., Albert, U., Chiarle, A., Soreca, I., Bogetto, F., Maina, G.,2008. Metabolic syndrome in Italian patients with bipolar disorder.Gen. Hosp. Psych. 30, 318–323.

Uçok, A., Polat, A., Bozkurt, O., Meteris, H., 2004. Cigarette smokingamong patients with schizophrenia and bipolar disorders. Psy-chiatry Clin. Neurosci. 58, 434–437.

van Winkel, R., De Hert, M., Van Eyck, D., Hanssens, L., Wampers,M., Scheen, A., Peuskens, J., 2008. Prevalence of diabetes and themetabolic syndrome in a sample of patients with bipolar disorder.Bipolar Disord. 10, 342–348.

Wilson, P.W.F., D'Agostino, R.B., Levy, D., Belanger, A.M.,Silbershatz, H., Kannel, W.B., 1998. Prediction of coronary heartdisease using risk factor categories. Circulation 97, 1837–1847.

World Health Organization, 1992. The ICD-10Classification ofMentaland Behavioural Disorders: Clinical Descriptions and DiagnosticGuidelines. WHO, Geneve.

Yumru,M., Savas, H.A., Kurt, E., Kaya,M.C., Selek, S., Savas, E., Oral,E.T., Atagun, I., 2007. Atypical antipsychotics related metabolicsyndrome in bipolar patients. J. Affect. Disord. 98, 247–252.