Cardiovascular morbidity and mortality in bipolar disorder


  ANNALS OF CLINICAL PSYCHIATRY 2011;23(1):40-47                                                  REVIEW ARTICLE

Cardiovascular morbidity and mortality
in bipolar disorder

Miriam Weiner, BA                                  BACKGROUND: There has been considerable interest in the elevated risk of
Lois Warren, BSW
                                                   cardiovascular disease associated with serious mental illness. Although
Department of Psychiatry                           the contemporary literature has paid much attention to major depres-
Carver College of Medicine
                                                   sion and schizophrenia, focus on the risk of cardiovascular mortality for
The University of Iowa
Iowa City, IA, USA                                 patients with bipolar disorder has been more limited, despite some inter-
                                                   est in the historical literature.
Jess G. Fiedorowicz, MD, MS
Department of Psychiatry
Carver College of Medicine                         METHODS: We reviewed the historical and contemporary literature related
Department of Epidemiology
                                                   to cardiovascular morbidity and mortality in bipolar disorder.
College of Public Health
The University of Iowa
Iowa City, IA, USA                                 RESULTS: In studies that specifically assess cardiovascular mortality, bipolar
                                                   disorder has been associated with a near doubling of risk when compared
                                                   with general population estimates. This may be explained by the elevated
                                                   burden of cardiovascular risk factors found in this population. These find-
                                                   ings predate modern treatments for bipolar disorder, which may further
                                                   influence cardiovascular risk.

                                                   CONCLUSIONS:    Given the substantial risk of cardiovascular disease, rigor-
                                                   ous assessment of cardiovascular risk is warranted for patients with bipo-
                                                   lar disorder. Modifiable risk factors should be treated when identified.
 CORRESPONDENCE                                    Further research is warranted to study mechanisms by which this elevated
 Jess G. Fiedorowicz, MD, MS                       risk for cardiovascular disease are mediated and to identify systems for
 Department of Psychiatry
                                                   effective delivery of integrated medical and psychiatric care for individu-
 Carver College of Medicine
                                                   als with bipolar disorder.
 Department of Epidemiology
 College of Public Health
 The University of Iowa                            KEYWORDS: bipolar disorder, cardiovascular disease, mortality, metabolic
 200 Hawkins Drive, W278GH                         syndrome, obesity, hypertension
 Iowa City, IA 52242 USA


40        February 2011 | Vol. 23 No. 1 | Annals of Clinical Psychiatry

             I N T RO D U C T I O N                              FIGURE 1
                                                                 Excess deaths in bipolar disorder
There has been considerable interest in the mortal-                               1,000
ity that accompanies many psychiatric disorders. The
psychiatric field has long focused on suicide, but in the                          800
past few decades, increasing attention has been given

                                                                  Excess deaths
to cardiovascular mortality with psychiatric disorders.                            600

This article reviews the research related to cardiovascu-
lar morbidity and mortality in bipolar disorder. As illus-                         400

trated in FIGURE 1, vascular disease is a leading cause of
excess death in bipolar disorder.
Early studies                                                                             Respiratory Accidents     Suicide       Vascular
Toward the close of the nineteenth century and begin-
ning of the twentieth, studies of morbidity and mortality        This graph uses aggregate data from one of the largest studies of mortality in mood
                                                                 disorders to illustrate the primary causes of excess death with bipolar disorder.22
in bipolar disorder were based almost entirely on case           In their sample, a total of 2129 excess deaths were identified in those with bipolar
                                                                 disorder, 700 of which were attributable to vascular disease (592 cardiovascular,
studies. The term bipolar disorder was not used, but the         108 cerebrovascular). Thus, nearly one-third of the excess deaths were attributable
cases of mania and the construct of manic-depressive             to vascular disease alone. The top 4 causes of excess deaths are illustrated in this
insanity were analogous to the contemporary construct
of bipolar disorder.1 In an early report, Bell reported on
40 patients with mania seen at the McLean Asylum from            ingly, like Bell, he described “the typical case of exhaus-
1836 to 1849, more than three-quarters of whom died. A           tion” as being characterized by dehydration, fatigue,
patient with Bell’s mania was said to “get so little food, so    increased pulse rate, and “some degree of temperature
little sleep, and be exercised with such constant restless-      elevation.”4 Contemporaneously, from reviewing a series
ness and anxiety, that he will fall off from day to day….        of case studies, Adlund drew the belief that the illness he
At the expiration of two or three weeks, your patient will       referred to as acute exhaustive psychosis “originates as a
sink into death….”2 Bell also compared their illnesses to        psychogenic problem and that the psychopathology…
delirium tremens, inflammation of the brain and menin-           is expressed through dysfunctions of the cardiovascular,
ges, and “passive congestion” of the cerebral circulation,       heat regulatory, and hematopoietic systems.”5
yet none of the patients demonstrated these conditions at             These early reports consisted almost entirely of case
autopsy. Bell concluded that patients experiencing simi-         studies and lacked the methodologic rigor of systematic
lar manias were at great risk of sudden death, or perhaps        observational studies. They occurred during a period
had an illness distinct from any previously defined.2 Sim-       lacking methods that would enable physicians to rule out
ilar cases were later presented, with continued debate           medical illnesses or draw more definitive conclusions on
about whether these cases represented sudden death in            autopsy. Further, the symptoms that characterize “acute
mania or a distinct condition altogether.3                       exhaustive psychoses” (described variously as Bell’s
      Nearly 100 years later, Derby studied mortality in         mania, fatal catatonia, manic-depressive exhaustive
patients with manic depression.4 Between 1912 and                deaths, Scheid’s cyanotic syndrome, and brain death)
1932, 980 of the patients admitted to Brooklyn (NY) State        now suggest agitated delirium, with symptoms includ-
Hospital for manic depression died during their stay. The        ing disorientation, confusion, visual hallucinations, and
cause of death in 40% was determined to be “exhaustion           fever.3,4 Thus, these early studies do not solidly support
from acute mental illness,” a condition perhaps similar to       the connection between bipolar disorder and cardio-
what Bell had described. The next most common cause of           vascular illness, although they are significant in showing
death was cardiac disease, which accounted for an esti-          that a link between the 2 conditions was already a topic of
mated 31% of deaths. In his review, Derby hypothesized           interest more than a century ago.
that “many of these ‘exhaustion’ cases appeared to have               Renewed concern about an association between
actually died of somatic disease” with “cardiovascular           bipolar disorder and cardiovascular disease followed the
disturbance” poised as a potential etiology.4 Interest-          publication of several cohort studies. Two German pub-                                               Annals of Clinical Psychiatry | Vol. 23 No. 1 | February 2011                               41

lications suggested that arteriosclerotic disease occurred       Psychiatric Case Register between 1960 and 1966, about
more often, and earlier, in those with manic depression          6% of that county’s population.13 This study found that
than in the general population.6,7 Several comprehensive         “the relative risk for the registered group, when adjusted
studies of state hospital populations were conducted in          for age, sex, marital status, and socioeconomic status,
the 1940s and early 1950s. Alstrom compared the death            is three times the general population.”13 The study also
rates of psychiatric patients in the New York Civil State        found that 4 causes of death were elevated in patients
Hospitals to the death rates of the general population in        with mental illness: circulatory illness, respiratory illness,
New York. He found that the annual death rate of patients        accidents, and suicide.13
with manic depression was more than twice that of                     Although a representative sample was used in the
patients with schizophrenia, at 7.7% vs 3.2%, respectively.      Babigian and Odoroff study, the potential for selection
Alstrom also estimated that the risk of cardiovascular dis-      bias persists in the registry. A patient’s mental disease
ease was twice of the general population in patients with        may have been identified only because medical treat-
manic depression.8 Ødegard reported on mortality from            ment was sought for a separate physical illness, making
Norwegian mental hospitals over a period of approxi-             it possible that the sampled cases included a dispropor-
mately 15 years. There were 3,370 deaths in a sample of          tionate medical burden. This potential selection bias
21,522 first admissions, a mortality rate 5 to 6 times that of   has been called Berkson’s bias.14 Because cases in such
the general population. He also reported that those with         studies are drawn from clinical samples, Berkson’s bias
manic depression had higher mortality rates than those           unavoidably affects research on mortality associated
with schizophrenia. Men and women with schizophrenia             with mental illness.15
had death rates 3.2 and 4.8 times higher than the general
population, respectively, whereas men and women with             Contemporary studies
manic depression had mortality rates 3.8 and 6.4 times           Bipolar disorder has consistently been associated with an
higher, respectively.                                            elevated risk for cardiovascular mortality, relative to the
     For circulatory diseases, these rates in men and            general population. Less consistent elevations in risk are
women with schizophrenia were 1.0 and 2.3 times higher,          seen when bipolar disorder is compared with unipolar
respectively, and 3.9 and 3.5 times higher for those with        depression. Elevations in mortality are often described
other diagnoses, including manic depression. Ødegard             with the standardized mortality ratio (SMR). The SMR
concluded that excess mortality from circulatory diseases        represents the ratio of observed to expected deaths.
was lower in those with schizophrenia compared with                   In a seminal “Iowa 500” study that included 100
the rest of the mentally ill population. He hypothesized         patients with mania, Tsuang et al found an increased
that these individuals “may be protected against circula-        risk of cardiovascular mortality in women (SMR = 1.63)
tory disturbance by their less intensive emotional reac-         but not men.16 A larger study in Denmark reported a
tions and their physical inactivity.”9 Malzberg reviewed         similar estimate for cardiovascular mortality in patients
the rates of mortality and discharge among first admis-          with bipolar disorder when compared with the general
sions to the New York Civil State Hospitals. Although Mal-       population (SMR = 1.60).17 In a subsample that included
zberg did not focus his attention on the cause of death          patients with bipolar and unipolar disorders, those
or on diagnoses, he concluded that mortality rates were          with bipolar disorder were found to have a significantly
lower for individuals with “dementia praecox” (schizo-           increased incidence of cardiovascular mortality com-
phrenia).10-12 These studies show that patients admitted         pared with patients with unipolar disorder.18
to public mental health hospitals had a mortality risk 4 to           A British study also showed a dramatic difference
10 times that of the general population, and they further        between expected and observed deaths from cardiovas-
identified individuals with bipolar disorder as being par-       cular disease in patients with bipolar disorder. During
ticularly at risk for cardiovascular disease.                    the period of study, 57 of 472 patients identified using
     The excess mortality identified by Malzberg, Øde-           the Edinburgh Psychiatric Case Register died; 42.1% of
gard, and Alstrom was largely attributed to the conditions       deaths were from cardiovascular illness, whereas a mor-
of public mental health facilities. More than 20 years later,    tality of 14% was expected.19
Babigian and Odoroff examined mortality in all cases of               A Swiss study followed 406 patients with bipolar
psychiatric illness reported to the Monroe County (NY)           (N = 220) and unipolar (N = 186) disorder who were hos-

42      February 2011 | Vol. 23 No. 1 | Annals of Clinical Psychiatry

 Standardized mortality ratios (SMR) for cardiovascular deaths in bipolar disorder
 Study                                                                  Sample                                        (Expected)                Female              Male
 Weeke et al, 1987     17
                                            Inpatients, Denmark, male, index admission 1950 to                             205
                                                                                                                                                  N/A                1.60
                                            1956 (N = 1133) or 1969 to 1976 (N = 2662)                                   (128.5)
 Sharma and Markar, 199419                  Inpatients, Scotland, index admission 1970 to 1975                              24
                                            (N = 472)                                                                       (8)
 Osby et al, 200122                         Inpatients, Sweden, index admission 1973 to 1995                              1073
                                                                                                                                                  1.94               2.65
                                            (N = 15,386)                                                                 (481.5)
 Angst et al, 200220                        Inpatients, Switzerland, index admission 1959 to                                59
                                            1963 (N = 220)                                                                (31.5)
 Laursen et al, 200721                      Inpatients, Denmark, living or born after 1973                                 818
                                                                                                                                                  1.67               1.58
                                            (N = 11,648)                                                                 (502.4)
 Composite                                                                                                                2179
 This table summarizes studies presenting data to estimate cause-specific cardiovascular mortality in bipolar disorder. The available estimates come from inpatient samples and
 suggest an approximate doubling of risk. The observed and expected deaths reflect composite data. Gender-specific estimates were available in only some studies. Observed
 deaths were provided on request for the Laursen et al study. Studies presenting an SMR for cardiovascular death in bipolar disorder and published in the past 25 years were
 selected for inclusion.

pitalized between 1959 and 1963 and followed for up to                                     had bipolar disorder. They reported that individuals with
38 years, at which point 76% of the patients had died.                                     bipolar disorder had higher mortality than those with
The patients with bipolar disorder were more likely to                                     unipolar depression, independent of suicide.23 Sims and
have died from cardiovascular disease than were those                                      Prior estimated mortality in 1482 inpatients treated for
with unipolar depression (an SMR of 1.84 for those with                                    severe neurosis in Birmingham (United Kingdom) hospi-
bipolar depression vs 1.36 for patients with unipolar                                      tals between 1959 and 1968 and found excess mortality in
depression). Additionally, patients with bipolar I disor-                                  diseases of the nervous, respiratory, and circulatory sys-
der had higher rates of death from cardiovascular dis-                                     tems (SMR 1.6), although the sample was not restricted to
ease than did those with bipolar II disorder.20 No subse-                                  individuals with bipolar disorder.24
quent study has compared cardiovascular mortality by                                            The studies detailed above reporting cardiovascular
bipolar subtype.                                                                           SMRs for samples of patients with bipolar disorder are
     A cohort study of over 5.5 million Danes followed                                     summarized in TABLE 1. Included are studies published in
from either their 15th birthday or the beginning of 1973                                   the past quarter century in which estimates of cardiovas-
through the beginning of 2001 found that of the 11,648                                     cular mortality for samples of individuals exclusively with
who were admitted for the first time due to bipolar dis-                                   bipolar disorder could be extracted. Several studies were
order, 3669 had died by the end of the study period. The                                   not included because they presented composite data
SMR for cardiovascular disease was 1.59 for men and                                        from mixed diagnostic samples,25-28 did not specifically
1.47 for women.21 This study strongly supported prior                                      assess cardiovascular mortality,25,28-30 or did not express
evidence of an elevated risk of cardiovascular morbidity                                   mortality relative to the general population.30
for those with bipolar disorder. A similar study in Sweden                                      Of studies from diagnostically mixed samples or
found a cardiovascular SMR for those with bipolar disor-                                   those focusing on natural deaths (inclusive of cardio-
der of 1.9 for men and 2.6 for women, compared with a                                      vascular mortality), some have failed to demonstrate
cardiovascular SMR for individuals with unipolar depres-                                   elevated mortality. In an Iowa study of mortality in
sion of 1.5 for men and 1.7 for women.22                                                   patients hospitalized between 1972 and 1981, Black
     In 1966, Perris and d’Elia investigated mortality in                                  et al found no excess of natural death in patients with
a Swedish sample of 797 patients with “depressive psy-                                     mood disorders (SMR = 0.9).31 Similarly, in 1987, Mel-
choses” admitted between 1950 and 1963, 120 of whom                                        oni et al did not find excess natural deaths in 179 inpa-                                                                    Annals of Clinical Psychiatry | Vol. 23 No. 1 | February 2011                                   43

tients with affective psychosis from an Italian sample,         mediated through traditional risk factors for cardiovas-
although the entire sample of 845 psychiatric inpatients        cular disease.
had a significantly elevated SMR for circulatory disease              There are several possible explanations for the excess
of 2.5.32 Rorsman also found no excess in natural deaths        cardiovascular mortality observed with bipolar disorder.
for those with affective disorders treated at an outpatient     Although the associations between bipolar disorder and
clinic in Sweden in 1962.33 Individuals with bipolar dis-       cardiovascular mortality predate modern pharmacologic
order were not examined separately from other patients          treatments, medication could conceivably contribute to
with affective disorders in these studies. A study by Martin    cardiovascular risk. Lithium can cause weight gain36,37
et al separated diagnostic groups for analysis but did not      and adversely influence glucose metabolism.38 Valproic
find any increase in natural deaths among 19 outpatients        acid has been even more strongly associated with both
with bipolar disorder.34 Although the assessment of natu-       weight gain and insulin resistance.39,40 Second-gener-
ral mortality in bipolar disorder by Martin et al was limited   ation antipsychotics are associated with hyperlipid-
by its small sample size, the presence of these negative        emia,41-45 insulin resistance or increased risk of diabetes
studies, notably the outpatient sample of Rorsman, sup-         mellitus,42,46-51 and weight gain.42,52-54 Beyond iatrogenic
ports some suspicion of Berkson’s bias.                         effects, individuals with bipolar disorder may have poor
     Several issues influence the results of mortality stud-    diets and obtain inadequate exercise.55 Smoking is more
ies in psychiatry. Choice of study population may be par-       common among those with bipolar disorder, even when
ticularly relevant, given the concern for Berkson’s bias.       compared with other serious mental illnesses.56
To facilitate case selection, studies to date have mainly             Several cardiovascular risk factors are more com-
involved clinical samples. The potential impact of selec-       mon in individuals with bipolar disorder than in the gen-
tion bias on these samples must consider illness acuity (eg,    eral population, which may help to explain the elevated
inpatient or outpatient), the nature of the cohort design       risk of cardiovascular mortality. These risk factors include
(retrospective or prospective), and secular trends. Addi-       obesity, hypertension, diabetes, and hyperlipidemia.
tional considerations include determination of death,           Each could contribute to excess cardiovascular mortality.
psychiatric assessment, and statistical inference.35 Limi-            Obesity has been associated with bipolar disorder. A
tations in statistical power may be particularly evident        study of 644 patients with bipolar disorder from private,
when estimating a cause-specific mortality for a specific       academic, and community mental health clinics found
psychiatric diagnosis, as highlighted in this review of car-    that 79% were overweight or obese, in contrast to 60%
diovascular mortality in bipolar disorder. When consid-         of the general population.57 Another study found that
ering the aggregate data, mindful of these limitations, an      45% of patients with bipolar disorder were obese, while
association between bipolar disorder and cardiovascular         another 29% were overweight.58 In a case-control study,
mortality is likely, particularly after weighting the larger,   participants with bipolar disorder weighed more, had a
contemporary studies of Laursen et al21 and Osby et al.22       greater body mass index, and had a higher percentage
     Despite data supporting an association between             of fat than controls; interestingly, the premorbid weights
bipolar disorder and cardiovascular morbidity, many             of the participants with bipolar disorder were not signifi-
questions remain. Whether bipolar disorder leads to an          cantly different than those of controls, leading the authors
increased risk of cardiovascular disease or whether car-        to conclude that weight gain is caused by the illness or its
diovascular disease elevates the likelihood that a person       treatment.59 A Norwegian study of 110 patients with bipo-
will suffer from bipolar disorder remain unanswered.            lar disorder also found more obesity among the patients
Whether there is a temporal association between affec-          than in the general population (24.9% vs 14.1%, respec-
tive disorder and cardiovascular comorbidity is also            tively). The differences were more pronounced when
unknown, as is the nature of that association. Most rel-        central obesity was measured (defined as a waist circum-
evant studies included patients who had been admit-             ference of >102 cm in men or >88 cm in women). In the
ted to hospitals because of their mental illness, leading       general population, only 16% met criteria for central obe-
to ascertainment bias and potential overestimation of           sity, compared with 39.9% of individuals with schizophre-
the risk of cardiovascular mortality. Lastly, the ques-         nia and 54.2% with bipolar disorder.60 In a recent study at
tion remains as to whether excess mortality is related to       the University of Iowa, evaluation of available weight and
an unidentified, inherent feature of mental illness or is       height information of 161 patients with bipolar disorder

44      February 2011 | Vol. 23 No. 1 | Annals of Clinical Psychiatry

 Estimates of NCEP-defined metabolic syndrome prevalence with bipolar disorder
 Study                                                              Sample                               N                   Prevalence
                                                                                                   (male/female)            (male/female)
 Cardenas et al, 200869                      Outpatients from a West Los Angeles                         98                      49%
                                             Veterans Affairs clinic                                   (90/8)                 (49%/50%)
 Fagiolini et al, 200558                     Consecutive recruits from 2003 to 2004                     171                      30%
                                             for bipolar disorder center in Pennsylvania              (67/104)                (31%/29%)
 Fiedorowicz et al, 200861                   Outpatients from a tertiary care center                  60 to 125               36% to 55%
                                             with primary diagnosis of bipolar disorder                (46/79)               (52% to 64%/
                                                                                                                             27% to 46%)
 Garcia-Portilla et al, 200870               Naturalistic, multicenter, cross-sectional                  194                    22.4%a
                                             study in Spain                                            (95/99)                (19%/26%)
 Teixeira and Rocha, 200773                  Consecutive sample of psychiatric inpatients                47                     38.3%b
                                                                                                       (35/12)                (43%/25%)
 van Winkel et al, 200871                    Prescreening for patients with bipolar                      60                     16.7%c
                                             disorder started on antipsychotics                        (34/26)                (19%/15%)
 Yumru et al, 200774                         Young sample of outpatients with bipolar                    125                     32%d
                                             disorder in Turkey                                        (78/47)                (30%/36%)
 NCEP: National Cholesterol Education Program.
  Nearly 60% higher than expected in the general Spanish population.
  More than 60% higher than expected in the general Brazilian population (75% of patients were overweight, and nearly                             als with schizophrenia. The incident rate ratio in this
half of these patients met criteria for obesity.61                                  study was 1.3, indicating that those with bipolar disor-
     Hypertension has been less consistently linked                                 der were significantly more likely to be newly diagnosed
with bipolar disorder. Although 2 studies indicate that                             with hypertension than were individuals in the general
hypertension was not more common among those with                                   population.64 The assessment of incidence rather than
bipolar disorder,58,62 some studies suggest otherwise. An                           prevalence may lessen some of the bias inherent in the
Iowa study showed an increased prevalence of hyper-                                 study of medical comorbidity in psychiatric popula-
tension among those with bipolar disorder but not                                   tions. With its size and rigorous assessment of incident
among those with unipolar mood disorders.63 Although                                cases, the study by Johannessen et al strongly supports a
the prevalence of hypertension in the control popula-                               link between bipolar disorder and hypertension.64
tion was 5.6%, the prevalence of hypertension was 14%                                    The association between bipolar disorder and dia-
in patients with bipolar disorder and 5% in patients with                           betes was first suggested nearly a century ago.65,66 Clini-
unipolar depression.63 In the Yates and Wallace study,                              cal studies have supported a greater prevalence of diabe-
hypertension was identified by a diagnosis of hyperten-                             tes among patients with bipolar disorder. Another study
sion, treatment with an antihypertensive, or systolic or                            found that 9.9% of inpatients with bipolar disorder had
diastolic blood pressure >160/95 mm Hg. A Norwegian                                 diabetes—3 times that expected in the general population
study estimated a prevalence of hypertension of 61% in                              (3.3%).67 A study of 4210 veterans with an average age of
those with bipolar disorder as compared with a preva-                               53 also showed a statistically significant greater prevalence
lence of 41% in the general population.60 This study used                           of diabetes among patients with bipolar disorder (17.2%
a much lower threshold, with a systolic or diastolic pres-                          vs 15.6%).62 Finally, in Norway, 5.5% of 113 patients with
sure ≥130/85 mm Hg. The largest study (involving 25,339                             bipolar disorder were found to have diabetes, as compared
people with bipolar disorder and a control population                               with 2.2% of the general population.60 Overall, the data
of 113,698) showed an increased rate of new-onset                                   support a link between bipolar disorder and diabetes.
hypertension among those with bipolar disorder, com-                                     A possible association between bipolar disorder
pared with both the control population and individu-                                and hyperlipidemia has also been suggested.68 In one                                                                  Annals of Clinical Psychiatry | Vol. 23 No. 1 | February 2011       45

study, almost half of the patients with bipolar disorder                                  tality than expected, based on general population
met metabolic syndrome criteria for hypertriglyceride-                                    estimates. Further, there is evidence that this risk may
mia, in contrast to only 32% of the general population.58                                 exceed that seen with other mental disorders. The
At the University of Iowa Hospitals and Clinics, of 77                                    strength and robustness of this association as indicated
patients with bipolar disorder and a recorded lipid pro-                                  by the evidence suggests this association cannot be
file, almost one-third were diagnosed with hypertriglyc-                                  dismissed. Although there may be features inherent in
eridemia, though some potential for surveillance bias                                     bipolar disorder that contribute to cardiovascular risk,
existed.61 Available evidence indicates that individuals                                  the preponderance of cardiovascular risk factors in this
with bipolar disorder may be at increased risk for hyper-                                 population warrants public health focus on traditional
lipidemia, specifically hypertriglyceridemia.                                             risk factors. Cardiovascular risk factors are readily iden-
      The metabolic syndrome can be conceptualized as a                                   tifiable with established screening approaches, and risk
composite measure of many of these cardiovascular risk                                    factors can be modified. Unfortunately, patients with
factors: visceral obesity, hypertriglyceridemia, low high-                                serious mental illness may be less likely to be monitored
density lipoprotein, hypertension, and insulin resistance.                                for75 and appropriately treated for cardiovascular risk
US studies indicate that patients with bipolar disorder may                               factors.76 Treatment for bipolar disorder may further
have an elevated risk of metabolic syndrome. Estimates                                    increase cardiovascular risk and require more rigorous
suggest that metabolic syndrome has a prevalence of 30% to                                monitoring. Additional research is needed to enable us
53% among those with bipolar disorder, as compared with                                   to better understand the many potential mediators of
a national prevalence of 27%.58,61,69 Several studies outside                             cardiovascular risk in this at-risk population. ■
the United States have also found evidence of an elevated
risk for metabolic syndrome in bipolar disorder. In Spain,                                DISCLOSURES: Dr. Fiedorowicz is supported by the National
22.4% of those with bipolar disorder had metabolic syn-                                   Institute of Mental Health (1K23MH083695-01A210), the
drome, as opposed to the national prevalence of 14.2%,70                                  Nellie Ball Trust Research Fund, and a NARSAD Young
and more than double the prevalence was seen in a Belgian                                 Investigator Award, and the Institute for Clinical and
study.71,72 Similar trends were found in studies from Bra-                                Translational Science at the University of Iowa (3 UL1
zil and Turkey (38.3% vs 23.7% and 32% vs 17.9%, respec-                                  RR024979-03S4). Dr. Fiedorowicz currently serves on
tively).73,74 TABLE 2 summarizes estimates of the prevalence                              colleagues’ studies with Neurosearch, Vitalin/Enzymatic
of metabolic syndrome, as defined by the National Choles-                                 Therapy, and the National Center for Complementary
terol Education Program, in bipolar disorder.                                             and Alternative Medicine/Food and Drug Administration
                                                                                          Orphan Products division. He also has received research
                                                                                          support for participating in a colleague’s investigator-
                     CO N C LU S I O N S                                                  initiated study with Eli Lilly and Company. Ms. Weiner
                                                                                          and Ms. Warren report no financial relationship with any
Individuals with bipolar disorder have a significantly                                    company whose products are mentioned in this article,
greater burden of cardiovascular morbidity and mor-                                       or with manufacturers of competing products.

1. Kraepelin E. Manic-depressive insanity and para-         siven Irreseins. Die Eltern und Kinder von Manisch-       tality among first admissions to the New York Civil State
noia. Edinburgh, Scotland: E and S Livingstone; 1921.       Depressiven. [Hereditary pathology of manic-depressive    Hospitals. III. Ment Hyg. 1953;37:619-654.
2. Bell LV. On a form of disease resembling some            illness. Parents and offspring of manic-depressives]. Z   13. Babigian HM, Odoroff CL. The mortality experience
advanced stages of mania and fever, but so contradis-       Gesamte Neurol Psychiatr. 1938;163:1-47.                  of a population with psychiatric illness. Am J Psychiatry.
tinguished from any ordinarily described combination        7. Bumke O. Handbuch der Geisteskrankheiten. Ber-         1969;126:470-480.
of symptoms, as to render it probable that it may be        lin, Germany: Julius Springer Verlag; 1928.               14. Merikangas KR, Kalaydjian A. Magnitude and impact
overlooked and hitherto unrecorded malady. American         8. Alstrom CH. Mortality in mental hospitals. Acta Psy-   of comorbidity of mental disorders from epidemiologic
Journal of Insanity. 1849;6:97-127.                         chiatr Neurol. 1942;17:1-42.                              surveys. Curr Opin Psychiatry. 2007;20:353-358.
3. Ray I. On undescribed forms of acute maniacal dis-       9. Ødegard Ø. Mortality in Norwegian mental hospi-        15. Berkson J. Limitations of the application of fourfold
ease. American Journal of Insanity. 1853;10:95-111.         tals 1926-1941. Acta Genet Stat Med. 1951;2:141-173.      tables to hospital data. Biometrics Bulletin. 1946;2:47-53.
4. Derby IM. Manic-depressive “exhaustion” deaths:          10. Malzberg B. Rates of discharge and rates of mortal-   16. Tsuang MT, Woolson RF, Fleming JA. Causes of
an analysis of “exhaustion” case histories. Psychiatr Q.    ity among first admissions to the New York civil state    death in schizophrenia and manic-depression. Br J Psy-
1933;7:436-449.                                             hospitals. Ment Hyg. 1952;36:104-120.                     chiatry. 1980;136:239-242.
5. Adland ML. Review, case studies, therapy, and            11. Malzberg B. Rates of discharge and rates of mortal-   17. Weeke A, Juel K, Vaeth M. Cardiovascular death
interpretation of the acute exhaustive psychoses. Psychi-   ity among first admissions to the New York civil state    and manic-depressive psychosis. J Affect Disord.
atr Q. 1947;21:38-69.                                       hospitals. II. Ment Hyg. 1952;36:618-638.                 1987;13:287-292.
6. Slater E. Zur Erbpathoogie des manisch-depres-           12. Malzberg BM. Rates of discharge and rates of mor-     18. Weeke A, Vaeth M. Excess mortality of bipolar and

46         February 2011 | Vol. 23 No. 1 | Annals of Clinical Psychiatry

unipolar manic-depressive patients. J Affect Disord.          gon in the hyperglycaemic response. Br J Pharmacol.           order. J Clin Psychiatry. 2002;63:207-213.
1986;11:227-234.                                              1994;111:861-865.                                             58. Fagiolini A, Frank E, Scott JA, et al. Metabolic syn-
19. Sharma R, Markar HR. Mortality in affective disor-        39. Dinesen H, Gram L, Andersen T, et al. Weight gain         drome in bipolar disorder: findings from the Bipolar
der. J Affect Disord. 1994;31:91-96.                          during treatment with valproate. Acta Neurol Scand.           Disorder Center for Pennsylvanians. Bipolar Disord.
20. Angst F, Stassen HH, Clayton PJ, et al. Mortality of      1984;70:65-69.                                                2005;7:424-430.
patients with mood disorders: follow-up over 34-38            40. Pylvänen V, Knip M, Pakarinen A, et al. Serum insu-       59. Shah A, Shen N, El-Mallakh RS. Weight gain occurs
years. J Affect Disord. 2002;68:167-181.                      lin and leptin levels in valproate-associated obesity. Epi-   after onset of bipolar illness in overweight bipolar
21. Laursen TM, Munk-Olsen T, Nordentoft M, et al.            lepsia. 2002;43:514-517.                                      patients. Ann Clin Psychiatry. 2006;18:239-241.
Increased mortality among patients admitted with              41. Huang TL, Chen JF. Serum lipid profiles and schizo-       60. Birkenaes AB, Opjordsmoen S, Brunborg C, et al.
major psychiatric disorders: a register-based study com-      phrenia: effects of conventional or atypical antipsy-         The level of cardiovascular risk factors in bipolar disor-
paring mortality in unipolar depressive disorder, bipolar     chotic drugs in Taiwan. Schizophr Res. 2005;80:55-59.         der equals that of schizophrenia: a comparative study. J
affective disorder, schizoaffective disorder, and schizo-     42. Henderson DC, Cagliero E, Gray C, et al. Clozap-          Clin Psychiatry. 2007;68:917-923.
phrenia. J Clin Psychiatry. 2007;68:899-907.                  ine, diabetes mellitus, weight gain, and lipid abnor-         61. Fiedorowicz JG, Palagummi NM, Forman-Hoffman
22. Osby U, Brandt L, Correia N, et al. Excess mortality      malities: a five-year naturalistic study. Am J Psychiatry.    VL, et al. Elevated prevalence of obesity, metabolic syn-
in bipolar and unipolar disorder in Sweden. Arch Gen          2000;157:975-981.                                             drome, and cardiovascular risk factors in bipolar disor-
Psychiatry. 2001;58:844-850.                                  43. Spivak B, Lamschtein C, Talmon Y, et al. The impact       der. Ann Clin Psychiatry. 2008;20:131-137.
23. Perris C, d’Elia G. A study of bipolar (manic-depres-     of clozapine treatment on serum lipids in chronic schizo-     62. Kilbourne AM, Cornelius JR, Han X, et al. Burden
sive) and unipolar recurrent depressive psychoses. X.         phrenic patients. Clin Neuropharmacol. 1999;22:98-101.        of general medical conditions among individuals with
Mortality, suicide and life-cycles. Acta Psychiatr Scand      44. Gaulin BD, Markowitz JS, Caley CF, et al. Clozapine-      bipolar disorder. Bipolar Disord. 2004;6:368-373.
Suppl. 1966;194:172-189.                                      associated elevation in serum triglycerides. Am J Psy-        63. Yates WR, Wallace R. Cardiovascular risk factors in
24. Sims A, Prior P. The pattern of mortality in severe       chiatry. 1999;156:1270-1272.                                  affective disorder. J Affect Disord. 1987;12:129-134.
neuroses. Br J Psychiatry. 1978;133:299-305.                  45. Osser DN, Najarian DM, Dufresne RL. Olanzapine            64. Johannessen L, Strudsholm U, Foldager L, et al.
25. Kallner G, Lindelius R, Petterson U, et al. Mortal-       increases weight and serum triglyceride levels. J Clin        Increased risk of hypertension in patients with bipolar
ity in 497 patients with affective disorders attending a      Psychiatry. 1999;60:767-770.                                  disorder and patients with anxiety compared to back-
lithium clinic or after having left it. Pharmacopsychiatry.   46. Guo JJ, Keck PE Jr, Corey-Lisle PK, et al. Risk of dia-   ground population and patients with schizophrenia.
2000;33:8-13.                                                 betes mellitus associated with atypical antipsychotic use     J Affect Disord. 2006;95:13-17.
26. Müller-Oerlinghausen B, Wolf T, Ahrens B, et al.          among patients with bipolar disorder: a retrospective,        65. Raphael T, Parsons JP. Blood sugar studies in
Mortality during initial and during later lithium treat-      population-based, case-control study. J Clin Psychiatry.      dementia praecox and manic-depressive insanity. Arch
ment. A collaborative study by the International Group        2006;67:1055-1061.                                            Neurol Psychiatry. 1921;5:687-709.
for the Study of Lithium-treated Patients. Acta Psychiatr     47. Lambert BL, Chou CH, Chang KY, et al. Antipsy-            66. Kasanin J. The blood sugar curve in mental disease:
Scand. 1994;90:295-297.                                       chotic exposure and type 2 diabetes among patients            II. The schizophrenia (dementia praecox) groups. Arch
27. Zilber N, Schufman N, Lerner Y. Mortality among           with schizophrenia: a matched case-control study of           Neurol Psychiatry. 1926;16:414-419.
psychiatric patients—the groups at risk. Acta Psychiatr       California Medicaid claims. Pharmacoepidemiol Drug            67. Cassidy F, Ahearn E, Carroll BJ. Elevated frequency
Scand. 1989;79:248-256.                                       Saf. 2005;14:417-425.                                         of diabetes mellitus in hospitalized manic-depressive
28. Vestergaard P, Aagaard J. Five-year mortality in lith-    48. Ollendorf DA, Joyce AT, Rucker M. Rate of new-            patients. Am J Psychiatry. 1999;156:1417-1420.
ium-treated manic-depressive patients. J Affect Disord.       onset diabetes among patients treated with atypical or        68. Brandrup E, Randrup A. A controlled investigation
1991;21:33-38.                                                conventional antipsychotic medications for schizophre-        of plasma lipids in manic-depressives. Br J Psychiatry.
29. Tsai SY, Lee CH, Kuo CJ, et al. A retrospective analy-    nia. MedGenMed. 2004;6:5.                                     1967;113:987-992.
sis of risk and protective factors for natural death in       49. Sernyak MJ, Gulanski B, Rosenheck R. Undiagnosed          69. Cardenas J, Frye MA, Marusak SL, et al. Modal sub-
bipolar disorder. J Clin Psychiatry. 2005;66:1586-1591.       hyperglycemia in patients treated with atypical antipsy-      components of metabolic syndrome in patients with
30. Craig TJ, Ye Q, Bromet EJ. Mortality among first-         chotics. J Clin Psychiatry. 2005;66:1463-1467.                bipolar disorder. J Affect Disord. 2008;106:91-97.
admission patients with psychosis. Compr Psychiatry.          50. Carlson C, Hornbuckle K, Delisle F, et al. Diabetes       70. Garcia-Portilla MP, Saiz PA, Benabarre A, et al. The
2006;47:246-251.                                              mellitus and antipsychotic treatment in the United King-      prevalence of metabolic syndrome in patients with bipo-
31. Black DW, Warrack G, Winokur G. The Iowa record-          dom. Eur Neuropsychopharmacol. 2006;16:366-375.               lar disorder. J Affect Disord. 2008;106:197-201.
linkage study. III. Excess mortality among patients with      51. Gianfrancesco F, White R, Wang RH, et al. Anti-           71. van Winkel R, De Hert M, Van Eyck D, et al. Preva-
‘functional’ disorders. Arch Gen Psychiatry. 1985;42:82-88.   psychotic-induced type 2 diabetes: evidence from a            lence of diabetes and the metabolic syndrome in a
32. Meloni D, Miccinesi G, Bencini A, et al. Mortality        large health plan database. J Clin Psychopharmacol.           sample of patients with bipolar disorder. Bipolar Disord.
among discharged psychiatric patients in Florence, Italy.     2003;23:328-335.                                              2008;10:342-348.
Psychiatr Serv. 2006;57:1474-1481.                            52. Simpson GM. Atypical antipsychotics and the               72. De Hert M, van Winkel R, Van Eyck D, et al. Preva-
33. Rorsman B. Mortality among psychiatric patients.          burden of disease. Am J Manag Care. 2005;11(suppl):           lence of diabetes, metabolic syndrome and metabolic
Acta Psychiatr Scand. 1974;50:354-375.                        S235-S241.                                                    abnormalities in schizophrenia over the course of the ill-
34. Martin RL, Cloninger CR, Guze SB, et al. Mortality        53. Volavka J, Czobor P, Sheitman B, et al. Clozapine,        ness: a cross-sectional study. Clin Pract Epidemol Ment
in a follow-up of 500 psychiatric outpatients. II. Cause-     olanzapine, risperidone, and haloperidol in the treatment     Health. 2006;2:14.
specific mortality. Arch Gen Psychiatry. 1985;42:58-66.       of patients with chronic schizophrenia and schizoaffec-       73. Teixeira PJ, Rocha FL. The prevalence of metabolic
35. Martin RL. Methodological and conceptual prob-            tive disorder. Am J Psychiatry. 2002;159:255-262.             syndrome among psychiatric inpatients in Brazil. Rev
lems in the study of mortality in psychiatry. Psychiatr       54. Zipursky RB, Gu H, Green AI, et al. Course and            Bras Psiquiatr. 2007;29:330-336.
Dev. 1985;3:317-333.                                          predictors of weight gain in people with first-episode        74. Yumru M, Savas HA, Kurt E, et al. Atypical antipsy-
36. Vendsborg PB, Bech P, Rafaelsen OJ. Lithium               psychosis treated with olanzapine or haloperidol. Br J        chotics related metabolic syndrome in bipolar patients. J
treatment and weight gain. Acta Psychiatr Scand.              Psychiatry. 2005;187:537-543.                                 Affect Disord. 2007;98:247-252.
1976;53:139-147.                                              55. Kilbourne AM, Rofey DL, McCarthy JF, et al. Nutri-        75. Kilbourne AM, Post EP, Bauer MS, et al. Thera-
37. Sachs G, Bowden C, Calabrese JR, et al. Effects of        tion and exercise behavior among patients with bipolar        peutic drug and cardiovascular disease risk monitor-
lamotrigine and lithium on body weight during main-           disorder. Bipolar Disord. 2007;9:443-452.                     ing in patients with bipolar disorder. J Affect Disord.
tenance treatment of bipolar I disorder. Bipolar Disord.      56. Lasser K, Boyd JW, Woolhandler S, et al. Smok-            2007;102:145-151.
2006;8:175-181.                                               ing and mental illness: a population-based prevalence         76. Kreyenbuhl J, Dickerson FB, Medoff DR, et al.
38. Hermida OG, Fontela T, Ghiglione M, et al. Effect         study. JAMA. 2000;284:2606-2610.                              Extent and management of cardiovascular risk factors in
of lithium on plasma glucose, insulin and glucagon in         57. McElroy SL, Frye MA, Suppes T, et al. Correlates of       patients with type 2 diabetes and serious mental illness.
normal and streptozotocin-diabetic rats: role of gluca-       overweight and obesity in 644 patients with bipolar dis-      J Nerv Ment Dis. 2006;194:404-410.                                                                       Annals of Clinical Psychiatry | Vol. 23 No. 1 | February 2011                                     47
You can also read
Next part ... Cancel