SEARCH

SEARCH BY CITATION

Keywords:

  • bipolar disorder;
  • medical burden;
  • staging;
  • age at onset;
  • comorbidity;
  • outcome;
  • treatment;
  • mania;
  • depression;
  • course;
  • diabetes;
  • smoking;
  • cardiovascular

Abstract

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References

Magalhães PV, Kapczinski F, Nierenberg AA, Deckersbach T, Weisinger D, Dodd S, Berk M. Illness burden and medical comorbidity in the Systematic Treatment Enhancement Program for Bipolar Disorder.

Objective:  Coexisting chronic medical conditions are common in bipolar disorder. Here, we report the prevalence and correlates of medical comorbidity in patients enrolled in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). We were particularly interested in associations between variables reflecting illness chronicity and burden with comorbid medical conditions.

Method:  We used intake data from the open-label component of the STEP-BD. History of medical comorbidity was obtained from the affective disorders evaluation, and its presence was the outcome of interest. The sample size in analyses varied from 3399 to 3534. We used multiple Poisson regression to obtain prevalence ratios.

Results:  The prevalence of any medical comorbidity in the sample was 58.8%. In addition to demographic variable, several clinical characteristics were associated with the frequency of medical comorbidity. Having more than 10 previous mood episodes, childhood onset, smoking, lifetime comorbidity with anxiety, and substance use disorders were independently associated with having a medical comorbidity in the final multivariate model.

Conclusion:  The results presented here reveal strong associations between variables related to illness chronicity and medical burden in bipolar disorder. This lends further support to recent multidimensional models incorporating medical morbidity as a core feature of bipolar disorder.


Significant outcomes

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References
  •  Several variables related to cumulative illness burden, such as having more than 10 previous episodes, comorbidity with anxiety and substance use disorders, smoking, and childhood onset, were independently associated with the prevalence of any medical comorbidity.

Limitations

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References
  •  Participants drawn from a clinical study, not population-based.
  •  Relevant mediators such as diet and lifestyle not analyzed.
  •  Comorbidity status based only on clinical interview.

Introduction

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References

Comorbidity with medical conditions is the rule in bipolar disorder (1–5). More often than not, patients will present with a chronic medical comorbidity, with cardiometabolic conditions being prominent among these (5). As such, a considerable part of the costs, disability, and premature mortality accrues from medical illness in bipolar disorder (6–8). This argues for healthcare models that appropriately integrate the treatment for medical comorbidities (6, 9).

Traditionally, the high prevalence of medical comorbidity has been viewed as a consequence of psychotropic medications and an unhealthy lifestyle (10). That, however, may not be the whole story. Recent research on metabolic, inflammation, and oxidative systems provides additional links between systemic and central nervous system pathophysiology. These pathways are possibly shared between comorbid medical disorders and bipolar disorder and may reflect common underlying vulnerabilities (6, 11–14). Preliminary research lends support to this idea, revealing associations between measures of cumulative illness burden, such as a greater chronicity and a higher number of episodes, and a greater medical burden in bipolar disorder (2, 4, 15).

Aims of the study

Here, we report the prevalence of medical comorbidity in patients enrolled in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). We were particularly interested in how variables related to illness chronicity and burden affect the likelihood of the patient presenting with a comorbid medical condition.

Material and methods

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References

Characterizations and complete descriptions of STEP-BD study and its pathways have been extensively published (16–20). The multisite, prospective, open-label component of the study was termed the ‘Standard Care Pathways’. In this, participation was offered to individuals seeking out-patient treatment for bipolar disorder. They needed to meet DSM-IV diagnostic criteria for bipolar I disorder, bipolar II disorder, bipolar disorder not otherwise specified (NOS), cyclothymia, or schizoaffective disorder, bipolar type. Psychiatric diagnoses were confirmed by a clinical rater using the Mini-International Neuropsychiatric Interview (MINI) (21). Patients in open-label treatment could receive any intervention, as clinically indicated. Individual human research committees of all treatment centers approved the study, and informed consent was obtained from participants.

The STEP-BD study enrolled 4107 participants who had baseline data, and information on medical comorbidity was available for 3766 patients. Owing to different patterns of missing data, sample size in bivariate analyses varies from 3399 to 3534. History of medical comorbidities was obtained from the Affective Disorders Evaluation (ADE) (22). Of interest in this category were chronic conditions. These included cardiovascular conditions, diabetes, thyroid disorders, previous head injuries, migraine, epilepsy, multiple sclerosis, peptic ulcer, and cancer. Of these, the STEP-BD had specifically coded information on all but cardiovascular disease and cancer. The latter two were obtained from axis-III diagnoses made by the clinicians in the ADE. Subjects with any of these conditions were coded as having a medical comorbidity.

Information regarding number of previous episodes, age at onset, rapid cycling, history of psychotic episodes, and current medication was also obtained from the ADE. Psychiatric comorbidity status was obtained from the MINI. For this report, the number of previous episodes was categorized as <5, 5–10, and more than 10 previous episodes, consistent with the methodology of previous studies (23). We also coded age of onset in childhood (before or after 13), because it has recently been suggested to be associated with worse outcomes in the STEP-BD (18).

A theoretical multivariate model was developed to test predictors of medical comorbidity. This is preferred for the confirmation of hypotheses to other data-driven approaches, because it avoids the problem of overfitting (24, 25). The model included baseline age (categorized in four age bins, 15–30, 31–40, 41–64, and 65 or older) and gender, low income (<20 000 US dollars a year), being married or living with a partner, having less than any college education, and being on disability (except when it was the outcome). They also included the following clinical variables: current mood state (depressed, manic, or euthymic), rapid cycling, lifetime substance use or anxiety comorbidity, current smoking, childhood onset, and the most commonly used medications (lithium, other mood stabilizers, atypical antipsychotics, typical antipsychotics, benzodiazepines, and antidepressants). Dummy variables (<5, 5–10, and more than 10 episodes) were created to enter the number of previous episodes in the model. The strength of associations is presented here as prevalence ratios. Because odds ratios tend to overestimate the risk when the outcome is common, we used binary Poisson regression to estimate multivariate effects, as described previously (26). We also used a simple linear-to-linear chi-square test to demonstrate the cumulative associations of five specific factors and the prevalence of medical comorbidity.

Results

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References

The prevalence of any medical comorbidity in the sample was 58.8% (Table 1). Several variables were associated with the presence of any medical comorbidity.

Table 1.   Baseline characteristics of patients in the Standard Care Pathways
CharacteristicTotal sample
Age37.7 ± 12.9
Female sex57%
Married or living with a partner34%
Low income (<20 000 a year)59%
On disability17%
At least some college education82%
Bipolar I disorder66%
Depressed at baseline37%
Manic at baseline5%
Rapid cycling36%
Smoking at baseline33%
Lifetime psychosis39%
Any anxiety disorder at baseline38%
Any substance use disorder at baseline17%
Childhood onset29%
Baseline medication
 Lithium31%
 Other mood stabilizers56%
 Atypical antipsychotics30%
 Typical antipsychotics2%
 Benzodiazepines24%
 Antidepressants45%
Medical comorbidity
 Any58.8%
 Neurological46.3%
 Cardiovascular or diabetes11.7%
 Peptic ulcer disease6.4%
 Cancer1.0%
 Thyroid disease15.1%

In addition to demographic factors, variables related to illness chronicity and psychiatric comorbidity had prominent effects in the final regression model. Of the latter, having had more than 10 previous episodes, a childhood onset, smoking, and lifetime comorbidity with anxiety and substance use disorders were independently associated with having any medical comorbidity (Table 2). We also found a very strong linear-by-linear association between the number of these clinical factors present (more than 10 previous episodes, childhood onset, smoking, and psychiatric comorbidity) and the prevalence of any medical comorbidity (χ= 97.89, P < 0.001; Fig. 1).

Table 2.   Association between demographic and clinical characteristics and the presence of any medical comorbidity in patients with bipolar disorder
 BivariateP valueMultivariateP value
Prevalence ratioPrevalence ratio
  1. Binary Poisson regression used to generate prevalence ratios in bivariate and multivariate analyses.

Age
 15–301Reference1Reference
 31–401.080.0661.000.995
 41–641.26<0.0011.160.001
 65 and over1.34<0.0011.38<0.001
Women1.100.0011.080.018
Low income (<20 000 a year)1.110.0011.050.188
Married or living with a partner0.89<0.0010.920.012
On disability1.23<0.0011.110.010
Less than any college education1.090.0101.010.765
Bipolar I disorder1.030.3901.000.942
Lifetime psychosis1.030.2761.040.237
Childhood onset1.16<0.0011.090.016
Previous episodes
 Between 4 and 101.120.051.050.426
 More than 101.38<0.0011.230.002
Lifetime anxiety disorder1.27<0.0011.16<0.001
Lifetime substance use disorder1.16<0.0011.110.002
Rapid cycling1.070.0200.960.300
Current smoking1.13<0.0011.110.003
Current medication
 Lithium0.980.5441.010.689
 Anticonvulsants0.980.4530.920.010
 Atypical antipsychotics1.050.0801.060.085
 Benzodiazepines1.13<0.0011.000.950
 Antidepressants1.12<0.0011.020.493
 Typical antipsychotics0.910.3940.830.166
Current depression1.100.0010.990.875
Current mania1.050.4720.990.874
image

Figure 1.  Number of associated factors related to cumulative illness burden (more than 10 previous episodes, childhood onset, comorbid anxiety, substance use, or smoking) and the prevalence of any medical comorbidity in bipolar disorder. *χ= 97.89, P < 0.001 for linear-to-linear trend.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References

Consistent with previous research, people with bipolar disorder seeking treatment more often than not presented with a chronic medical disorder. Prominent among the factors associated with medical comorbidity were measures of greater illness chronicity and burden. An early onset and having a highly recurrent illness as well as lifetime comorbidity with anxiety and substance use disorders increased the prevalence of clinical conditions independently of a host of demographical factors.

These data support previous research associating illness chronicity with medical burden in bipolar disorder. It has been previously argued that allostasis is a relevant framework for understanding these findings (11). In this model, mood episodes, together with stress and medical and other comorbidities, are seen as contributing to allostatic stress states. With recurrence, the additive burden results in an overload of relevant regulatory systems. These include hypothalamic–pituitary–adrenal (HPA) axis hyperactivity and pro-inflammatory and pro-oxidant states, providing a mechanistic link between neurobiological and systemic networks (5, 6, 11). In this fashion, a high psychiatric illness burden may engender vulnerability to (systemic) medical illness. Alternatively, a shared diathesis to medical and psychiatric morbidity is a viable explanation (27–29).

Empirical prospective data show more advanced illness stage, defined per the number of previous episodes (as a measure of cumulative illness burden), is associated with poorer treatment response in bipolar disorder (23). The basis for the utility of such models is that patients in different stages may have different needs and prognoses (30). The results here further support this notion, and patients in a late stage (i.e. highly relapsing, comorbid, or very early onset) may especially require a healthcare model that incorporates attention to chronic systemic illness (31). Conversely, early stage bipolar disorder may be the ideal period for interventions that aim for the reduction in medical risk factors (32, 33). Of note, there have been proposals of specific ‘packages’ of treatment interventions specifically for those with bipolar disorder comorbid with chronic systemic conditions (34, 35). Even if preliminary, these data suggest better outcomes with more complex interventions.

The STEP-BD data provide a large sample of people with established bipolar disorder. Whether these results apply to population data, however, is not known. Variables related to diet and lifestyle have been associated with mood disorders (36–40) and could be relevant confounders. Another issue is that medical comorbidity was probably underestimated in this sample. Ideally, medical charts should be analyzed for more reliable results (4). Their absence from our model is a limitation. Research specifically designed to test causal hypotheses and pathways leading to medical comorbidity and disability is in order. Early intervention studies would be ideal to that end. In these, lifestyle (including diet and exercise, for example) and oxidative and inflammatory markers could be tested as mediators of medical comorbidity in patients initially free from them.

Current views of bipolar disorder are moving away from the idea of it being a relatively benign cyclical illness. Instead, accumulating evidence points to an active process of illness progression. Associated pathways (related, for instance, to neurotrophic factors, oxidative stress, and inflammation) are known to be risk indicators for the development of medical disorders such as cardiovascular disease, diabetes, and osteoporosis (14, 27, 28). Variables associated with a greater illness burden, such as number of recurrences and early onset, are associated with a greater likelihood of medical comorbidity. The results presented here support recent, more complex, multidimensional models incorporating medical morbidity as a core feature of illness progression in bipolar disorder.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References

The Systematic Treatment Enhancement Program for Bipolar Disorder was funded by the National Institute of Mental Health, which had no role in drafting or analyzing and preparing this manuscript. The data set was obtained from the National Institutes of Mental Health by a request from Prof. Berk that was specifically approved by Barwon Health’s Human Research Ethics Committee. STEP-BD was registered at clinicaltrials.gov with the identifier NCT00012558. Dr. Magalhães is supported by the National Institute for Translational Medicine, Brazil.

Declaration of interests

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References

Prof. Kapczinski has received grant/research support from Astra-Zeneca, Eli Lilly, the Janssen-Cilag, Servier, CNPq, CAPES, NARSAD, and the Stanley Medical Research Institute; has been a member of the speakers’ boards for Astra-Zeneca, Eli Lilly, Janssen, and Servier; and has served as a consultant for Servier.

Prof. Berk has received research support from Stanley Medical Research Foundation, MBF, NHMRC, Beyond Blue, Geelong Medical Research Foundation, Bristol Myers Squibb, Eli Lilly, Glaxo SmithKline, Organon, Novartis, Mayne Pharma, Servier. He has been a speaker for Speakers: Astra Zeneca, Bristol Myers Squibb, Eli Lilly, Glaxo SmithKline, Janssen Cilag, Lundbeck, Merck, Pfizer, Sanofi Synthelabo, Servier, Solvay, and Wyeth. He has been a consultant for Astra Zeneca, Bristol Myers Squibb, Eli Lilly, Glaxo SmithKline, Janssen Cilag, Lundbeck, and Servier.

As of September 2, 2009, Dr Nierenberg is a consultant for Appliance Computing (Mindsite), BrainCells, Brandeis University, PGx Health, Shire, Schering-Plough, Targacept, and Takeda; has received grant/research support from National Institute of Mental Health, Pamlab, Pfizer, and Shire; has received honoraria from Belvior Publishing, University of Texas Southwestern Dallas, Hillside Hospital, American Drug Utilization Review, American Society for Clinical Psychopharmacology, Baystate Medical Center, Columbia University, IMEDEX, MJ Consulting, New York State, MBL Publishing, Physicians Postgraduate Press, SUNY Buffalo, University of Wisconsin, and the University of Pisa; is on the advisory boards of Appliance Computing, BrainCells, Eli Lilly, and Takeda; is a stock shareholder of Appliance Computing and BrainCells; and, through Massachusetts General Hospital, owns copyrights to the Clinical Positive Affect Scale and the MGH Structured Clinical Interview for the Montgomery-Asberg Depression Rating Scale exclusively licensed to the MGH Clinical Trials Network and Institute.

References

  1. Top of page
  2. Abstract
  3. Significant outcomes
  4. Limitations
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interests
  11. References