| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Department of Psychiatry (S.H., R.H.P., N.M.P.), University of Connecticut School of Medicine; Department of Oral Health and Diagnostic Sciences (J.W.), University of Connecticut School of Dental Medicine; and Pat and Jim Calhoun Cardiology Center (W.B.W.), Department of Medicine, University of Connecticut School of Medicine, Farmington, Connecticut.
Address correspondence and reprint requests to Nancy M. Petry, Department of Psychiatry, University of Connecticut Health Center, 263 Farmington Avenue, Farmington, CT 06030-3944. E-mail: petry{at}psychiatry.uchc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Methods: Data from the National Epidemiologic Survey on Alcohol and Related Conditions were analyzed with 10,573 adults aged 
60 years surveyed.
Results: A total of 13.30% of older adults reported diagnoses of CHD. Age (odds ratio (OR) = 1.04), morbid obesity (OR = 1.60), hypertension (OR = 2.29), lifetime nicotine dependence (OR = 1.41), and lifetime drug use disorders (OR = 2.19) were all significantly related to CHD. Being female (OR = 0.73) relative to male and a lifetime social drinker (OR = 0.71) compared with alcohol abstainers decreased the odds of CHD. After controlling for these characteristics, the presence of a lifetime major depressive episode was significantly associated with increased risk of CHD (OR = 2.05), but the lifetime anxiety disorders assessed were not. The association between lifetime mood disorders and CHD was similar for both genders, and single versus multiple depressive episodes conferred equal risk of CHD.
Conclusions: These data demonstrate that a lifetime major depressive episode increases the risk of CHD in older adults.
Key Words: coronary heart disease • mood disorders • anxiety disorders • older adults
Abbreviations: CHD = coronary heart disease; NIAAA = National Institute on Alcohol Abuse and Alcoholism; NESARC = National Epidemiologic Survey on Alcohol and Related Conditions; AUDADIS = Alcohol Use Disorder and Associated Disabilities Interview Schedule; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders—Fourth Edition; BMI = body mass index; OR = odds ratio; CI = confidence interval.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Traditional risk factors do not entirely explain the risks for developing CHD, and psychiatric disorders such as depression may play a role (11,12). Pratt et al. (13) conducted a 13-year follow up of 1551 respondents in the Epidemiologic Catchment Area study, who were free of medical illness at the initial assessment. A diagnosis of past-year major depression increased the risk of myocardial infarction more than four-fold, even after controlling for medical risk factors and other psychiatric diagnoses (13).
Although considerable progress linking psychiatric disorders and CHD has been made, much of the epidemiological research is limited. The vast majority of studies rely on symptom questionnaires and do not use systematic diagnoses (14), and control for confounders is often inadequate, inappropriate, or not well justified (15). Early studies were conducted in treatment-seeking samples that were not representative of the broader population (16) and used samples too small to examine gender effects. Many psychiatric disorders can remit, but most studies measure concomitant psychiatric symptoms and CHD risk. Consequently, there are scant data regarding lifetime history of psychiatric diagnoses (17). Few studies have examined whether a recurrent psychiatric disorder confers greater risk than a single episode (18), and little is known about specific psychiatric disorders other than depression and their relationship to CHD (19). Data regarding anxiety as an independent risk factor, for example, are inconsistent (20).
The purpose of this study was to examine the relationships between psychiatric disorders and CHD in data derived from a nationally representative sample of older adults. The National Institute on Alcohol Abuse and Alcoholism (NIAAA) National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) (21) produced a large database on psychiatric disorders and also assessed some medical conditions, including CHD. This dataset provides a unique opportunity to investigate the relationship between CHD and psychiatric disorders, paying particular attention to recency and chronicity of the disorders.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
18 years. Lay interviewers (with an average of 5 years of Census and health interviewing experience and extensive training) (22) conducted face-to-face interviews, with a response rate of 81%. African-American and Hispanic racial-ethnic groups were oversampled, with data weighted accordingly to account for oversampling and study design characteristics. Weighted data were then adjusted based on 2000 Decennial Census with respect to socioeconomic variables including age, gender, race-ethnicity, and region of the country (21). Of the 43,093 individuals surveyed, 10,573 (21.6%) were aged 60 years; they are the focus of this report because younger adults are unlikely to manifest CHD (2).
Medical Diagnoses
The NESARC examined past-year prevalence of 10 medical conditions: cirrhosis, other liver disease, stomach ulcers, gastritis, arthritis, tachycardia, hypertension, angina pectoris, arteriosclerosis, and myocardial infarction. Respondents were asked if they had experienced any of these conditions within the past year. If the condition was endorsed, they were asked whether a physician or other health professional made the diagnosis. Only diagnoses reportedly made by a physician or other health professional were included as affirmative in this study. Angina, arteriosclerosis, and myocardial infarction were coded into a composite CHD outcome. Those reporting angina (8.77%, SE = 0.36%), arteriosclerosis (5.61%, SE = 0.29%), or myocardial infarction (2.82%, SE = 0.19%) were classified as having CHD. A total of 13.30% (SE = 0.45) of adults >60 years old had a past-year diagnosis of one or more of these diseases.
Body Mass Index (BMI)
Self-reports of height and weight were used to calculate BMI; weight in kilograms was divided by square of height in meters. Extreme obesity refers to BMI 40 kg/m2 (6).
DSM-IV Mood and Anxiety Disorder Assessment
The Alcohol Use Disorder and Associated Disabilities Interview Schedule-DSM-IV (AUDADIS-IV) assessed some lifetime and past-year Diagnostic and Statistical Manual of Mental Disorders—Fourth Edition (DSM-IV) (23) mood and anxiety disorders: major depressive episodes, dysthymia, manic episodes, hypomanic episodes, agoraphobia with panic, agoraphobia without panic, social phobias, specific phobias, and generalized anxiety. Data regarding the number of major depressive episodes experienced were also obtained.
All disorders required clinical significance, and only primary diagnoses were included in the analyses. Thus, substance-induced mood disorders or those resulting from a general medical condition (about 1.0% of all cases) were not included. Depressive episodes accounted for entirely by bereavement were also not included. Reliability and validity of the AUDADIS-IV measures of mood and anxiety disorder are adequate (24,25).
DSM-IV Alcohol and Drug Use Disorder Assessment
The AUDADIS-IV also included an assessment of nicotine dependence, alcohol use disorders, and drug abuse and dependence for ten classes of drugs: sedatives, cannabis, cocaine, heroin, opiates, stimulants, tranquilizers, hallucinogens, inhalants, and other. Respondents were classified as to whether or not they met lifetime nicotine dependence. Lifetime diagnoses of alcohol or other drug abuse required meeting at least one of four abuse criteria in the 12-month period preceding the interview and/or before the 12-month period. Lifetime dependence diagnoses required meeting at least three of seven dependence criteria during the past year and/or before the past year. Due to low prevalence rates, all drug use disorders other than nicotine and alcohol were classified together. To evaluate potentially protective effects of moderate alcohol consumption, respondents who reported never drinking alcohol were classified as lifetime abstainers. Those who were not lifetime abstainers and had no lifetime or past-year alcohol abuse or dependence were classified as "social drinkers." Test-retest reliability and validity of the AUDADIS-IV alcohol, nicotine, and drug diagnoses are established (24–29).
Statistical Analyses
Respondents with CHD were compared with those without CHD. Cross-tabulations evaluated differences in demographic characteristics and psychiatric disorders between the two groups.
Logistic regressions evaluated the relationships between the mood and anxiety disorders assessed and CHD after controlling for other variables that differed significantly between the CHD and non-CHD groups, as well as clinically relevant variables such as gender, ethnicity, and drug use disorders. Covariates included demographics (age, gender, ethnicity, education, marital status, and income), health variables (morbid obesity and hypertension), and substance use (alcohol use and disorders, nicotine dependence, and drug use disorders). In a subsequent analysis, lifetime anxiety disorders were removed from the model, and the relationship between lifetime mood disorders and CHD was examined for a gender interaction to ascertain if the effect differed for men and women. A parallel analysis was conducted to examine gender interactions with respect to anxiety disorders and CHD. In addition, associations between CHD and past-year mood and anxiety disorders, rather than solely lifetime diagnoses, were assessed; neither lifetime mood and anxiety disorders were included in this model, but demographic, health, and substance use variables noted above were controlled.
To further assess the relationships between mood disorders and CHD, additional analyses examined a) whether lifetime or past-year mood disorders were more closely associated with CHD when both were entered simultaneously into the model; b) whether chronicity of depressive episodes (0, 1, or >1) predicted CHD; and c) which specific mood disorders were independently related to CHD. Presence or absence of anxiety disorders was not included in these models, as it was not significantly related to CHD, but other covariates were.
A total of 9913 older adults had complete data for all variables and were included in these analyses. Odds ratios (ORs) and 95% confidence intervals (CIs) are presented for each independent variable. Analyses were conducted using SUDAAN, a software package that utilizes Taylor series linearization to adjust for design effects of complex sample surveys (30).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
|
Likelihood of CHD Predicted by Any Lifetime Mood and Anxiety Disorder
Table 2 displays ORs and 95% CIs from the logistic regression with covariates included, and several variables remained significant predictors of CHD. Older age, morbid obesity, hypertension, and nicotine and drug use diagnoses were associated with CHD. Women and social drinkers had decreased odds of CHD relative to men and alcohol abstainers. Any lifetime mood disorder was significantly associated with increased odds of CHD. Any lifetime anxiety disorder (of the five anxiety disorders assessed) was not significantly associated with CHD in these multivariate analyses. The interaction between any lifetime mood disorder and gender was not significantly associated with CHD (Wald F(1,65) = 0.62; p = .43), nor was the interaction of lifetime anxiety disorders and gender (Wald F(1,65) = 2.05; p =.16).
|
Likelihood of CHD Predicted by Any Past-Year Mood and Anxiety Disorder
Analyses also evaluated the relationships between more recent mood and anxiety disorders and CHD with all covariates included. Any lifetime mood and anxiety disorders were removed from the model and replaced by any past-year mood and anxiety disorders. Any past-year mood disorder was a significant predictor of CHD, with an OR of 2.12 (95% CI = 1.61–2.80). The presence of any of the past-year anxiety disorders assessed in this survey was not associated with CHD (OR = 1.22; 95% CI = 0.95–1.57).
Likelihood of CHD Predicted by Recency or Chronicity of Mood Disorder
Both any lifetime and any past-year mood disorder were entered into an analysis simultaneously to ascertain whether recency of mood disorders was more closely associated with CHD than a lifetime history. Any lifetime mood disorder remained significant (OR = 1.73; 95% CI = 1.38–2.18), but any past-year mood disorder was not (OR = 1.37; 95% CI = 0.96–1.95).
The impact of chronicity of depression was assessed by entering the number of depressive episodes into the analysis along with covariates. Having one lifetime depressive episode (OR = 2.10; 95% CI = 1.70–2.60) and more than one depressive episode (OR = 2.26; 95% CI = 1.75–2.91) were both significantly associated with increased odds of CHD relative to no lifetime depressive episodes. However, those with more than one depressive episode did not have increased odds of CHD relative to those with only one episode (OR = 1.07; 95% CI = 0.78–1.48).
Likelihood of CHD Predicted by Specific Mood Disorders
Table 3 lists specific mood disorders and their prevalence in relation to the presence or absence of CHD. In logistic regressions controlling for covariates, lifetime and past-year major depression, dysthymia, and manic disorders were associated with CHD, but hypomania was not.
|
To determine whether the association between dysthymia and CHD was due to its overlap with depression, another regression analysis evaluated the effect of lifetime dysthymia plus major depression and dysthymia alone. Dysthymia plus depression was a significant predictor of CHD (OR = 1.79; 95% CI = 1.46–2.20), but dysthymia alone was not (OR = 1.43; 95% CI = 0.99–2.05). When lifetime manic episode plus major depression and mania alone were entered as independent variables, neither were significantly associated with CHD (OR = 1.89; 95% CI = 0.92–3.90; and OR = 1.78; 95% CI = 0.66–4.82, respectively).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This study expands on prior reports by examining the relationships between CDH and specific psychiatric disorders. Major depression, dysthymia, and manic disorders increased the likelihood of CHD after controlling for demographic, substance use, and behavioral risk factors. However, most individuals with dysthmia and manic disorder also have major depression; associations between these mood disorders and CHD did not remain when only those with the single mood disorder (dysthmia or mania without major depression) were analyzed separately.
In contrast, the anxiety disorders assessed in this study were not significantly associated with CHD in multivariate analyses, although other studies suggested a possible role of anxiety in the development of CHD (32). Studies finding effects for anxiety generally used symptom questionnaires, which may confound anxiety symptoms with other problems. The present study used diagnoses, although not all anxiety disorders were assessed.
Of note, lifetime mood disorders remained significant when both lifetime and past-year disorders were analyzed simultaneously, suggesting that depression is associated with long-term sequelae independent of proximal effects. Given that the average age of onset for mood disorders is early to middle adulthood (33), an individual >60 years old who has experienced a mood episode in the past year is likely to have experienced one or more episodes earlier in life. Thus, past-year depression in older adults may be a marker for lifetime mood disorder. Alternatively, there may be two subgroups for whom depression and CHD covary. One subgroup may experience a mood disorder early in life, which increases the risk for subsequent CHD; another subgroup may experience their first episode at the time of their CHD diagnosis or cardiac event (34). Those falling into this second category would endorse past-year depression only and their experiences would not support, or even speak to, depression as a risk factor for CHD. Unfortunately, this cross-sectional study design does not permit the elucidation of complex temporal relationships.
These data extend prior research by investigating chronicity of depressive episodes. A single lifetime depressive episode increased the risk of CHD, and additional episodes did not further elevate risk. Some studies showed that the number of depressive episodes correlates with more impaired markers for CHD (18), perhaps due to the sensitive measurement of CHD risk. To our knowledge, the number of mood episodes and incident overt CHD has not been studied.
The finding that depression confers equivalent risk for CHD in men and women differs from an earlier report suggesting depression is a stronger risk factor in men (35). However, our results are consistent with another recent study showing similar associations across genders (36).
Demographic, health, and substance use risk factors associated with CHD in this study are generally consistent with those reported elsewhere (1). Gender was not related to CHD in the univariate analyses, perhaps because the largest proportion of participants was in the youngest age category (60–69 years of age). When age was covaried in the multivariate analyses, the female gender was significantly associated with decreased odds of CHD.
Although this study points to novel findings about mood disorders and CHD, limitations exist. First, not all mood and anxiety disorders were assessed, and the CHD category consisted of only angina, arteriosclerosis, and myocardial infarction. Second, CHD was based on the respondents' self-reports of physician or health professionals' diagnoses, and there was no objective confirmation of CHD. Finally, this was a cross-sectional study so directionality between mood disorders and heart disease could not be determined. Further research is needed to ascertain mechanisms by which mood disorders influence the development and course of CHD so that treatments to ameliorate the effects of depression can be evaluated.
Despite limitations, the reliability of the findings is enhanced by the large randomly selected sample, high response rate, in-person interviews, and use of psychometrically sound psychiatric diagnostic instruments. Confounders were statistically controlled, and both lifetime and past-year mood and anxiety disorders were evaluated, as was chronicity of depressive episodes.
These data suggest that efforts to prevent depression in susceptible individuals, by virtue of genetic predisposition or life stressors, may be beneficial. In light of data showing depression increases cardiac mortality in persons with and without established cardiovascular disease (37), efforts to ameliorate depression both before and after CHD incidence may be important.
In sum, a lifetime history of a major depressive episode is associated with increased likelihood of CHD in both men and women, after controlling for demographic, substance use, and behavioral risk factors. These data suggest that practitioners should assess for past as well as current major depressive episode, as both are strongly and consistently associated with heart disease in older adults.
We thank NIAAA and the US Census Bureau field representatives who administered the NESARC interview.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Preparation of this report was supported in part by National Institutes of Health Grants GCRC M01-RR06192, R01-MH60417, R01-MH60417-Supp, R01-DA13444, R01-DA018883, R01-DA14618, R01-DA016855, P50-AA03510, and P50-DA09241.
DOI:10.1097/PSY.0b013e3181574977
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  H. M. Gonzalez, T. Croghan, B. West, D. Williams, R. Nesse, W. Tarraf, R. Taylor, L. Hinton, H. Neighbors, and J. Jackson Antidepressant Use in Black and White Populations in the United States Psychiatr Serv, October 1, 2008; 59(10): 1131 - 1138. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
