| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLE |
From the Department of Epidemiology and Public Health (S.A.W., S.V.K., A.H., H.M.K.), and the Department of Medicine, Section of Cardiovascular Medicine (H.M.K.), Yale University School of Medicine, New Haven, CT; Center for Outcomes Research and Evaluation, Yale New Haven Hospital, New Haven, CT (H.M.K.); Department of Preventive Medicine, Griffin Hospital, Derby, CT (A.H.); and Department of Medicine (J.L.A., V.V.), Division of Cardiology, Emory University School of Medicine, Atlanta, GA.
Address reprint requests to: Viola Vaccarino, MD, PhD, Emory University School of Medicine, Department of Medicine, Division of Cardiology, 1256 Briarcliff Road NE, Suite 1-North, Atlanta, GA 30306. Email: lvaccar{at}emory.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: We examined the effect of depression (assessed by means of the Center for Epidemiological Studies Depression Scale (CES-D) with a cutoff point of 
21) on the incidence of heart failure in a community sample of persons aged 65 years who were participants in the New Haven cohort of the Established Populations for Epidemiological Studies in the Elderly.
RESULTS: At baseline 2501 individuals were free of heart failure. Of these, 188 (132 women and 56 men) scored as depressed. Depressed participants were significantly more likely to have hypertension, diabetes, and mobility-related functional limitations and were less likely to be male or married. During the 14-year follow-up period, 313 participants (146 men and 167 women) developed heart failure, defined as hospital admission for heart failure or mortality with heart failure as the underlying cause of death. After adjusting for baseline differences in demographic and comorbidity factors and functional status using Cox regression, depression tended to be associated with a greater risk of heart failure (hazard ratio (HR) = 1.52, 95% confidence interval (CI) = 0.94–2.43, p = .09). This effect was significant in women (HR = 1.96, 95% CI = 1.11–3.46, p = .02) but not in men (HR = 0.62, 95% CI = 0.23–1.71, p = .05 for the interaction term between sex and depression).
CONCLUSIONS: Depression is an independent risk factor for heart failure among elderly women but not elderly men.
Key Words: heart failure, • risk factors, • aging, • depression.
Abbreviations: BMI = body mass index;; CES-D = Center For Epidemiological Studies Depression Scale;; CI = confidence interval;; CVD = cardiovascular diseases;; DBP = diastolic blood pressure;; HF = heart failure;; HR = hazard ratio;; ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical Modification;; MI = myocardial infarction;; SBP = systolic blood pressure;; SPMSQ = Short Portable Mental Status Questionnaire;; YHAP = Yale Health and Aging Project.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
65 years of age. YHAP is one of the four cohorts of the Established Populations for Epidemiological Studies of the Elderly program conducted by the National Institute on Aging (20). The YHAP cohort was assembled in 1982 by obtaining a stratified probability sample of persons from three New Haven housing strata: public housing for the elderly, private housing for the elderly, and general housing in the community. Male participants were oversampled. Approximately 82% of eligible individuals participated in the baseline interview (N = 2812, 1169 men and 1643 women). The average age of women and men in the sample was 74.2 and 73.2 years, respectively. Of the 2812 participants, 175 who responded positively at baseline to two questions addressing orthopnea ("Do you get shortness of breath when you are lying down flat?" and "Does this shortness of breath improve when you sit up, or do you use extra pillows at night to prevent it?") and an additional 136 participants who used both loop diuretics and cardiac glycosides were considered to be potential HF cases and were excluded from the analyses.
Study Measures
Depressive symptoms at baseline were measured by means of the Center for Epidemiological Studies Depression Scale (CES-D), a 20-item scale of depressive symptoms with a potential total score ranging from 0 to 60 (21). If three or more items of the scale were missing, the sum score was set to missing. If one or two items were missing, the depressive symptom score was based on a weighted average of the nonmissing items. Although there is no standard cutoff point for use of this scale, a cutoff point of 21 was used to indicate symptom severity that approximates a clinical diagnosis of depression. When compared with a clinical diagnosis of depression based on DSM-III-R criteria, a score of 21 or more on the CES-D has a high sensitivity (92%) and specificity (87%) (22). The term "depression" in this report denotes a depressive symptom score of 21 or more on the CES-D.
Information on a number of other baseline factors was collected and considered in our analyses. Baseline sociodemographic variables of interest included age, sex, years of education, race (white vs. nonwhite), and marital status. Self-reported history of physician-diagnosed myocardial infarction (MI), stroke, hip fracture, cancer, and diabetes were also assessed, as were smoking (current, past, never) and body mass index (BMI) (coded as follows: low, 
23 kg/m2; intermediate, 24–27 kg/m2; or high, 28 kg/m2). An additional dummy variable was created for participants with missing BMI values (N = 199). Information on baseline functional status was measured by means of four items addressing mobility-related function, which is considered to be most closely associated with cardiovascular diseases (CVDs) (8). Two items were selected from an instrument for gross mobility (ability to climb one flight of stairs and walk half a mile) (23), and two were from the Katz index of basic activities of daily living (ability to walk across a room and to transfer from bed to chair) (24). Functional limitation was defined as the inability to perform at least one of these four functions. Cognitive function was also taken into consideration as a potentially confounding variable, given its association with depressive symptoms and CVD. The Short Portable Mental Status Questionnaire (SPMSQ) (25) was used to measure cognitive function. The SPMSQ score was used as a continuous variable.
In addition to the baseline factors noted above, we also collected information on the occurrence of incident MIs during follow-up. Incident MIs were based on the occurrence of hospitalizations for MI and were validated according to accepted criteria for MI diagnosis (26).
Study Outcome
Follow-up time started from the date of baseline interview in 1982 to the date of first hospitalization for HF, date of death due to HF as the underlying cause, or the end of follow-up period (December 31, 1996), whichever came first (27). In secondary analyses, we also examined survival time (for HF-specific mortality) and time to first hospitalization as separate outcomes. Information on hospital admissions for study participants was obtained from the Yale New Haven Hospital and Hospital of Saint Raphael, the two New Haven community hospitals. This information was collected through a surveillance system that conducted weekly reviews of all admissions to these two local hospitals. Additional information on hospitalizations in this cohort was obtained from federal data (Medicare Part A Beneficiary Bill History data from the Health Care Financing Administration [HCFA]). Only about 5% of hospitalizations identified through the HCFA data were missed from our surveillance and not included in the study.
Nonfatal HF events during follow-up were initially identified by reviewing the hospitalization records of study participants who were hospitalized between 1982 and 1986. When a given hospitalization record indicated that the primary discharge diagnosis or one of the first three secondary discharge diagnoses was HF (ICD-9-CM codes 428, 402.01, 402.11, 402.91, 404.01, 404.13, 404.91, and 404.93), the record was reviewed in detail. To qualify as a HF event, the hospitalization record had to indicate 1) the presence of shortness of breath accompanied by either physical signs or radiographic evidence of HF, 2) HF as one of the three major precipitating factors for hospitalization, and 3) HF-related symptoms occurring before or within 24 hours of hospital admission. Fatal HF events were identified by reviewing death certificates of all participants who died. A single nosologist coded all death certificates according to ICD-9-CM criteria. Deaths due to HF were identified by the same codes as those used for hospitalizations. Information on vital status was virtually complete (>99% of all cohort members).
We also considered an alternative method of defining HF events during follow-up. This alternative method defined HF on the basis of medication use during follow-up (ie, new use of both loop diuretics and cardiac glycosides) as determined by direct examination of medication bottles in the participant’s home at any of the follow-up interviews. This was done to shed light on HF events occurring outside the hospital.
Statistical Analysis
As the first step in our analysis, we examined bivariate associations between depression and other baseline factors. Next, we ran a Cox proportional hazards regression model, which assessed the unadjusted association between depression and HF. Similar Cox models were run to examine the unadjusted relationship between other baseline factors and risk of HF. Finally, we ran multivariable Cox models that assessed the association between depression and HF risk after adjustment for other factors. Initially, these multivariable models adjusted for baseline factors only. However, we also examined a multivariable model that adjusted for baseline factors as well as the occurrence of MI during follow-up as a time-dependent covariate. In all multivariable models, a backward elimination procedure was used. This procedure retained only those variables that were significant (p < .10) predictors of the outcome or that changed the point estimate for depression by 10% (depression, however, was always forced into the model).
To account for the complex nature of the YHAP sampling design, we used Survey Data Analysis (SUDAAN) software. SUDAAN applies the Taylor linearization procedure to estimate standard errors for regression coefficients. SUDAAN was used for all statistical analyses except the multivariable model that included a time-dependent covariate for MI during follow-up. For this model, we used SAS software because SUDAAN does not handle time-dependent covariates.
Previous literature suggests that the association between psychosocial factors and cardiovascular outcomes (7–9), as well as the association between biological risk factors and HF incidence (28), varies by gender. Therefore, analyses were conducted in the entire sample as well as separately by gender, and gender-related interactions were explored.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Bivariate associations between baseline depression and other baseline factors under study are shown in Table 1. Depressed persons were significantly more likely to be hypertensive, diabetic, and to have mobility-related functional limitations. Depressed persons were significantly less likely to be male; among men, 56 (5.3%) were depressed at baseline, whereas among women, 132 (9.0%) were depressed. Depressed persons were also significantly less likely to be married and had significantly fewer years of education.
|
21 was a significant risk factor for incident HF (HR = 1.69, 95% CI 1.19–2.39). Other univariate Cox models indicated that age, male gender, history of MI, history of diabetes, hypertension, pulse pressure, and cognitive impairment were significantly associated with increased risk of HF. Smoking status and obesity were not associated with HF risk in univariate models. Next we conducted analyses to determine whether depression was associated with HF after adjustment for other factors. Figures 1 and 2 show that, compared with those who were not depressed, depressed participants at baseline had a lower probability of survival free of HF after adjustment for age. Table 2 presents the results of the multivariable Cox proportional hazards model, which assessed the relationship between depression and HF after adjustment for age and other baseline factors. The model selected depression, age, gender, diabetes, pulse pressure, hypertension, and functional status as independent predictors of increased HF risk. In this model, the adjusted relative risk of incident HF among depressed individuals, compared with those who were not depressed, was marginally significant (hazard ratio (HR) = 1.52, 95% confidence interval (CI) = 0.94–2.43, p = .086). Because BMI and smoking were found to be risk factors for HF in previous studies, we reran the model shown in Table 2 with additional adjustment for these two factors. We found, however, that this model yielded results that were nearly identical to those shown in Table 2. We also refitted this model with additional adjustment for MI occurring during follow-up as a time-dependent covariate, but again the results were very similar to those from the model shown in Table 2.
|
|
|
Despite the relatively small number of minority participants, analyses were done to see if race showed a major influence on risk. We did not detect any such effects. Race was neither a significant predictor of HF incidence (p = .15) nor a confounder of the association between depression and HF incidence. We also explored the interaction between race and depression in the multivariable model, but we found that this interaction was not significant (p = .97).
In additional secondary analyses, we examined whether perceived social and instrumental support, which were measured in this cohort (29), buffered the relationship between depression and HF, by examining the interaction between each type of support and depression. However, we found that neither of these interaction terms was significant in the multivariable model.
Depressed patients are more likely to have diminished health status and increased healthcare utilization (30). We therefore examined mortality and hospitalization separately. When analyzed separately, participants with a depressive symptom score 21 were more likely to die of HF (HR = 1.78, 95% CI = 0.34–9.24) than those who did not score as depressed, although the confidence interval was quite large and included 1. Similarly, those who had a depressive symptom score 21 were more likely to be hospitalized for HF (HR = 1.65, 95% CI = 1.02–2.67). The point estimate for the association between depression and HF was greater when mortality was considered as the outcome than when hospitalization was considered as the outcome. It is therefore unlikely that the observed association between depression and HF hospitalization is biased by the greater propensity of depressed individuals to seek medical attention. When HF medication use during follow-up was considered as a measure of HF incidence, the association with baseline depression showed similar trends, but these trends were less strong than those described above: for the total population, HR = 1.21 (95% CI = 0.78–1.88); for women, HR = 1.26 (95% CI = 0.73–2.19); and for men, HR = 0.95 (95% CI = 0.44–2.05).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Instead, depressed mood may affect HF risk through biological pathways possibly involving the neuroendocrine system. Emotional factors have long been hypothesized to precipitate hospitalization for HF patients (31). Psychological stress is believed to impair left ventricular function through various mechanisms, including the induction of transient changes of the electrophysiological properties of the myocardium and subsequent ventricular arrhythmia (32) and the occurrence of wall-motion abnormalities and a subsequent drop in ejection fraction (33, 34). Furthermore, patients with dilated cardiomyopathy and HF have shown impairment in left ventricular function during psychological stress (35). Emotional factors may increase the risk of HF through activation of the sympathetic nervous system (12, 13), which may contribute to progression of left ventricular dysfunction and renal sodium retention. Given the limited data available, however, these biological mechanisms remain speculative.
It is also possible that depression may act to increase risk through behavioral mechanisms. For example, investigators have reported that depressed persons are less likely to adhere to medical therapies (30), and failure to adhere to medical therapies could, presumably, increase the risk of developing HF. Unfortunately, this study did not have information on adherence to medical therapies; therefore, we were unable to determine if this factor explained the observed relationship between depression and increased HF risk.
It is unclear why depression was associated with increased HF risk among women but not men. The lack of association between depression and HF among men may be due to the underrepresentation of men who might have been susceptible to the adverse effects of depression, because these men might have died before the inception of the study. This may be true when dealing with an older cohort as in the present study. This hypothesis is supported by previous studies, which indicated a greater association between depression and CVD in younger men compared with older men (1, 3, 4) and a greater association between depression and CVD in older women compared with older men (8). It may also be that depressed women are simply more likely to seek hospitalization for HF than are men. Most of the HF events we observed in this study were hospitalizations. If women were more likely than men to go to the hospital for their HF symptoms, then the gender difference we observed in the association between depression and HF events may simply have been indicative of a difference in medical attention–seeking behavior between depressed men and women. However, when we defined HF based on medication use instead of hospitalization, we still found that depression increased HF risk in women but not men. This argues against the claim that the gender difference we observed simply reflects a greater willingness of depressed women to go to the hospital. One could speculate that that the gender difference we observed has a pathophysiological basis. If depression increases HF risk through sympathoadrenal activation, it may be that depression leads to a greater degree of sympathoadrenal activation in women compared with men or that the effects of such activation are more deleterious in women as opposed to men.
The current analysis has a number of limitations. First, we did not have information on serum lipoprotein levels. However, serum lipoprotein levels are not likely to have changed the association of interest because they do not seem to differ markedly according to the presence or absence of depressed mood (7), and if they do, they are lower rather than higher in depressed individuals (36). Second, the identification of incident cases by hospitalization and mortality in the present study may have excluded less severe cases of disease. Only persons who were hospitalized or those who were symptomatic (ie, in whom HF had undergone decompensation) were included as cases. However, decompensated HF is likely to have a greater impact on the clinical course and quality of life of patients than HF at an earlier stage. Therefore, this is a reasonable method of outcome ascertainment. Nevertheless, because information on medication intake was available at follow-up, we also considered this as a measure of HF incidence. The association between baseline depression and HF incidence as measured by medication use showed similar trends but was less strong, suggesting that depression may affect primarily incidence of HF defined by hospitalization and mortality rather than by medication use. However, measuring the outcome by medication use is less accurate. Although we have the exact date for mortality and hospital admission, we do not know the exact date for the initiation of drug therapy during the 3-year follow-up intervals. This may dilute the association between depression and HF when the outcome is measured by drug use. Third, we were unable to assess how treatment for depression during follow-up, and adherence to medical treatment recommendations, may have altered the relationship between depression and HF.
Despite its limitations, this study has several advantages. This is the first longitudinal study to examine the effect of depression on HF incidence among the elderly, who comprise the majority of HF patients. The strengths of this study include complete and validated ascertainment of end points, prospective design, and control for various established risk factors for HF. Lastly, the present study is based on a random sample of community-dwelling elderly persons. The New Haven population is ethnically and socioeconomically diverse and as such should be representative of the elderly population in many communities of the United States.
Our findings indicate that the presence of depressed mood should alert clinicians to a higher risk of not only ischemic heart disease but also HF in their patients. Early diagnosis and treatment of depression may reduce the burden of morbidity and mortality due to HF in the elderly and reduce the public health costs associated with HF treatment. The results of this study affirm the importance of current clinical guidelines from the Agency for Health Care Policy and Research for diagnosis and treatment of depression (37), including routine use of screening for depression in the elderly.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication January 9, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. S. Rathore, Y. Wang, B. G. Druss, F. A. Masoudi, and H. M. Krumholz Mental Disorders, Quality of Care, and Outcomes Among Older Patients Hospitalized With Heart Failure: An Analysis of the National Heart Failure Project Arch Gen Psychiatry, December 1, 2008; 65(12): 1402 - 1408. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Whooley, P. de Jonge, E. Vittinghoff, C. Otte, R. Moos, R. M. Carney, S. Ali, S. Dowray, B. Na, M. D. Feldman, et al. Depressive Symptoms, Health Behaviors, and Risk of Cardiovascular Events in Patients With Coronary Heart Disease JAMA, November 26, 2008; 300(20): 2379 - 2388. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. D. Schocken, E. J. Benjamin, G. C. Fonarow, H. M. Krumholz, D. Levy, G. A. Mensah, J. Narula, E. S. Shor, J. B. Young, and Y. Hong Prevention of Heart Failure: A Scientific Statement From the American Heart Association Councils on Epidemiology and Prevention, Clinical Cardiology, Cardiovascular Nursing, and High Blood Pressure Research; Quality of Care and Outcomes Research Interdisciplinary Working Group; and Functional Genomics and Translational Biology Interdisciplinary Working Group Circulation, May 13, 2008; 117(19): 2544 - 2565. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Strain and M. Blumenfield Challenges for Consultation-Liaison Psychiatry in the 21st Century Psychosomatics, April 1, 2008; 49(2): 93 - 96. [Full Text] [PDF]
|
|
|
|
|
|
A. Azevedo, P. Bettencourt, F. Frioes, M. Alvelos, C. Abreu-Lima, H.-W. Hense, and H. Barros Depressive Symptoms and Heart Failure Stages Psychosomatics, February 1, 2008; 49(1): 42 - 48. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Fraguas Jr., D. V. Iosifescu, J. Alpert, S. R. Wisniewski, J. L. Barkin, M. H. Trivedi, A J. Rush, and M. Fava Major Depressive Disorder and Comorbid Cardiac Disease: Is There a Depressive Subtype With Greater Cardiovascular Morbidity? Results From the STAR*D Study Psychosomatics, October 1, 2007; 48(5): 418 - 425. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. C.T.F. Alves, J. Rays, R. M.S. Telles, R. F. Junior, M. Wajngarten, B. W. Romano, C. Watanabe, and G. F. Busatto Effects of Antidepressant Treatment on Cognitive Performance in Elderly Subjects With Heart Failure and Comorbid Major Depression: An Exploratory Study Psychosomatics, February 1, 2007; 48(1): 22 - 30. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Rutledge, V. A. Reis, S. E. Linke, B. H. Greenberg, and P. J. Mills Depression in Heart Failure: A Meta-Analytic Review of Prevalence, Intervention Effects, and Associations With Clinical Outcomes J. Am. Coll. Cardiol., October 17, 2006; 48(8): 1527 - 1537. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Whooley Depression and cardiovascular disease: healing the broken-hearted. JAMA, June 28, 2006; 295(24): 2874 - 2881. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. P. van Melle, P. de Jonge, J. Ormel, H. J.G.M. Crijns, D. J. van Veldhuisen, A. Honig, A. H. Schene, M. P. van den Berg, and for the MIND-IT investigators Relationship between left ventricular dysfunction and depression following myocardial infarction: data from the MIND-IT Eur. Heart J., December 2, 2005; 26(24): 2650 - 2656. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. G. Kahl, M. Bester, W. Greggersen, S. Rudolf, L. Dibbelt, B. M. Stoeckelhuber, H.-B. Gehl, V. Sipos, F. Hohagen, and U. Schweiger Visceral Fat Deposition and Insulin Sensitivity in Depressed Women With and Without Comorbid Borderline Personality Disorder Psychosom Med, May 1, 2005; 67(3): 407 - 412. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Frasure-Smith and F. Lesperance Reflections on Depression as a Cardiac Risk Factor Psychosom Med, May 1, 2005; 67(Supplement_1): S19 - S25. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Mallik, H. M. Krumholz, Z. Q. Lin, S. V. Kasl, J. A. Mattera, S. A. Roumains, and V. Vaccarino Patients With Depressive Symptoms Have Lower Health Status Benefits After Coronary Artery Bypass Surgery Circulation, January 25, 2005; 111(3): 271 - 277. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. G. Schneider The Intersection of Mental and Physical Health in Older Mexican Americans Hispanic Journal of Behavioral Sciences, August 1, 2004; 26(3): 333 - 355. [Abstract] [PDF]
|
|
|
|
|
|
J. S. Rumsfeld, E. Havranek, F. A. Masoudi, E. D. Peterson, P. Jones, J. F. Tooley, H. M. Krumholz, J. A. Spertus, and Cardiovascular Outcomes Research Consortium (CORC) Depressive symptoms are the strongest predictors of short-term declines in health status in patients with heart failure J. Am. Coll. Cardiol., November 19, 2003; 42(10): 1811 - 1817. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Wilson, J. L. Bienias, C. F. Mendes de Leon, D. A. Evans, and D. A. Bennett Negative Affect and Mortality in Older Persons Am. J. Epidemiol., November 1, 2003; 158(9): 827 - 835. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. G. Blazer Depression in Late Life: Review and Commentary J. Gerontol. A Biol. Sci. Med. Sci., March 1, 2003; 58(3): M249 - 265. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Depression: A Risk Factor for Onset of Heart Failure in the Elderly? Journal Watch Psychiatry, March 6, 2002; 2002(306): 2 - 2. [Full Text]
|
|
|
|
|
|
R. Shabetai Depression and Heart Failure Psychosom Med, January 1, 2002; 64(1): 13 - 14. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
