Psychosomatic Medicine
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Anstey, K. J.
Right arrow Articles by Luszcz, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Anstey, K. J.
Right arrow Articles by Luszcz, M. A.
Related Collections
Right arrow Other Epidemiology
Right arrow General Sexual Medicine Issues
Psychosomatic Medicine 64:880-888 (2002)
© 2002 American Psychosomatic Society

ORIGINAL ARTICLES

Mortality Risk Varies According to Gender and Change in Depressive Status in Very Old Adults

Kaarin J. Anstey, PhD and Mary A. Luszcz, PhD

From the Centre for Mental Health Research, Australian National University (K.J.A.), Canberra, Australian Capital Territory, Australia; and the School of Psychology and Centre for Ageing Studies, Flinders University (M.A.L.), Adelaide, South Australia.

Address reprint requests to: Kaarin Anstey, Centre for Mental Health Research, Australian National University, Canberra ACT 0200, Australia. Email: kaarin.anstey{at}anu.edu.au


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 
OBJECTIVE: We aimed to evaluate whether gender and different patterns of change in depressive status over 2 years were associated with different risks of mortality in the subsequent 6 years.

METHODS: Depression (CES-D) was assessed in 1947 participants in 1992 and a smaller proportion of the sample in 1994. The mortality risk at July 30, 2000, associated with depression and change in depression was estimated using proportional hazards models.

RESULTS: After controlling for demographic variables, smoking, alcohol, and medical conditions, depression was associated with mortality for men but not women. In men, incident depression was associated with mortality after controlling for all other variables. Chronic depression and remitted depression were also associated with mortality, but this effect was explained by medical conditions. In women, change in depressive status was not associated with mortality.

CONCLUSIONS: Depression confers a greater risk of mortality for men than women with incident depression in old age representing the greatest risk for men. The course of depressive illness must be considered when evaluating mortality risk.

Key Words: depression, • mortality, • longitudinal study, • aging, • sex differences.

Abbreviations: BMI = body mass index;; CES-D = Center for Epidemiological Studies Depression Scale;; CVD = cardiovascular disease;; IRR = incident rate ratio.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 
Many studies have shown that clinical depression or depressive symptoms are a risk factor for all-cause mortality (1–13) and mortality due to cardiovascular disease (CVD) (14, 15) at all stages of adulthood. These results are found in studies using a variety of measures of depression in clinical and community samples. In nonpsychiatric samples, factors shown to mediate the effect of depression on mortality include chronic physical illness, low body mass index (BMI), cancer, heart disease, smoking, and alcohol abuse (16).

In research on the elderly, depression has been found to be more prevalent in women, but studies vary with respect to their findings on gender differences in the mortality of depression (13, 17, 18). Studies have reported differing results, including 1) depression is a risk factor for mortality in men only (13); 2) minor depression is a risk factor for men but not women, whereas major depression is a risk factor for both men and women (11, 13); and 3) no differences between men and women (19).

Discrepancies in study findings may in part be due to differences in sample selection, composition or age, and the explanatory variables analyzed. It is possible that different explanatory variables are important at different ages. For example, depression predicts coronary heart disease more consistently in middle-aged than in elderly populations (20). Another possibility is that because depression is a fluctuating illness, a single occasion of measurement may not concur with a series of longitudinal measurements that more accurately reflects the chronicity of depression. One of the few studies that examined change in depressive status in relation to mortality (21) found that neither incident nor chronic depression (measured on the CES-D) at baseline and after 2 years was associated with mortality over a 12-month follow-up period. Although based on a large community sample, the follow-up period was relatively short. Over a 6-year interval, one study using an Established Populations for Epidemiologic Study of the Elderly (EPESE) sample (14) showed incident depression in men was associated with new coronary heart disease, mortality associated with CVD, and CVD events, but not with all-cause mortality. Chronic depression was not associated with mortality. However, in another study based on an EPESE sample, coronary heart disease in women was associated with depression, but the effect was no longer significant after controlling for impaired physical function (20). The Systolic Hypertension in the Elderly Program (SHEP) found that although baseline depressive symptoms did not predict mortality, emerging depressive symptoms were associated with risk of cardiovascular events and mortality in the subsequent 5 years (22). For women only, depression and stroke were associated.

It is noteworthy that no study has found remitted depression to be associated with increased risk of mortality. This finding suggests that a high depression score on a single occasion is not necessarily a risk factor for mortality and that improvement in depression may reduce morbidity and mortality. It may also partially explain the equivocal findings on the mortality of depression arising from studies using a single occasion of measurement. Longitudinal community-based studies including successive measurements of depression are required to clarify the general findings relating depression and gender to mortality.

There are several reasons why incident depression in the very old may be a risk factor for mortality. It is possible that depression is a prodrome of cardiovascular events or dementia (5, 23, 24) or that it occurs concurrently with these conditions. Depression may be an early sign of impending physical decline (25) or incur a physiological response that predisposes individuals to cardiovascular disease or other illness, such as cancer (14). It also seems that incident depression and gender interact in old age. For example, Penninx et al. (14) found that newly depressed mood in men was associated with impending cardiovascular events, whereas chronically depressed men were not at increased risk of cardiovascular events or mortality. The same study did not find that incident depression in women increased the risk of mortality. Hence the physiological impact of depression may be greater for men than for women despite the greater prevalence of depression among women. One difficulty inherent in longitudinal studies is that depression may be associated with nonresponse, leading to underestimation of chronic depression.

The present study investigated gender and change in depressive status in a community sample of very old adults who were assessed on the CES-D in 1992 and 1994 and for whom mortality data were available until 2000. The 8-year follow-up period is longer than that reported in most other studies on depression and mortality in this age group. Nonresponse and the effect of gender were identified as key factors that may have an impact on this relationship. The assumption that the negative effects of depression in very old adults are cumulative led to the following hypotheses. Individuals who remain depressed at both follow-ups would have an increased risk of mortality compared with individuals who were never depressed or individuals whose depression remitted during the study. Incident depression during the study would be associated with increased risk of mortality because it may indicate late-onset depression, which in turn has been associated with dementia (23), or because it may be associated with significant negative changes in health or personal well-being. Finally, individuals with remitted depression would not have higher mortality rates than individuals never depressed.

Confounding demographic variables (age, sex, marital status, education) identified in previous studies (14, 16) were included in analyses. Likewise because low weight (19) and medical conditions have been associated with both depression and mortality, as have medications, smoking, and alcohol use, these were tested as possible mediators of the link between depression and mortality.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 
Sample
The sample was drawn from participants in the Australian Longitudinal Study of Ageing (ALSA) and has been fully described elsewhere (26, 27). The South Australian Electoral Roll was used as a sampling frame to identify households with residents older than 70 years of age. The sample was stratified by age and sex into three 5-year cohorts (70–74, 75–79, and 80–84 years) and a fourth cohort of individuals older than 85 years of age. The study comprises six waves of data collection: the baseline collection (between September 1992 and March 1993) and five subsequent waves of data collected at approximately12-month intervals. Waves 2, 4, and 5 were telephone interviews and did not include a clinical or cognitive assessment. Waves 3 and 6 involved a clinical assessment. Data from waves 1 and 3 are used in this study. In these waves, a comprehensive 2-hour home interview was followed by an optional individual clinical assessment conducted approximately 2 weeks later. The home interviews included questions on demographic, health, medical psychosocial, and physical status. At wave 1, 1947 participants (1039 men) were interviewed, and 1500 (828 men) underwent the clinical assessment. At wave 3, 1557 participants (809 men) were interviewed, and 1311 (694 men) underwent the clinical assessment. Of the 1947 participants interviewed at wave 1, 1910 completed the CES-D. This subsample was used for the present study.

Measures
Depression.
Depression was measured using the CES-D, a 20-item questionnaire designed for use in community-based epidemiological studies (28). Responses are in reference to the way the individual felt in the last week and are answered on a four-point Likert-scale from rarely or none of the time (0) to most of the time (3). Scores range from 0 to 60.

Smoking and alcohol.
Current smoking status (smoker vs. nonsmoker) and number of alcoholic drinks per week (one or less per week, two to three per week, four or more per week) were included as measures of self-care.

Medical conditions and medications.
Self-reported information on medical conditions was obtained from the interview, in which participants were asked if they had ever experienced individual medical problems (binary variables: 1 = yes, 0 = no), including diabetes, heart attack, heart disease, stroke, transient ischemic attacks, hypertension, and cancer. All medications were identified during the home interview, and the drug name was recorded from medication containers by the interviewer. Total number of medications was used as a continuous variable.

Body mass index.
BMI was calculated from height and weight measurements obtained during the clinical assessment, in which weight was recorded to the nearest 0.1 kg and height was measured to the nearest 0.1 cm.

Level of education.
Education level (<14 years or >=14 years) was recorded in the home interview.

Determination of Mortality Status
Information on mortality status was confirmed by searches of official death certificates conducted by the Epidemiology Branch of the Department of Health and Human Services in South Australia. Confirmation of deaths was obtained from the South Australian Cancer Registry, which by law has direct access to information held by births, deaths, and marriages. The cancer registry matched study participants’ identifying data (full name, date of birth, address) with deaths. If an obvious direct match was not easily identified, the electoral role was checked for names and addresses, errors with dates of birth, and misspellings of names. A minority of deaths occurring in other states or overseas could not be confirmed using this method; in these cases participant histories traced through friends or relatives listed as informants by study participants were relied on to provide final death status.

Statistical Analyses
For all analyses, participants were classed as alive or deceased at the censoring date of July 30, 2000. Differences between survivors and decedents were assessed with t tests and {chi}2 tests. To estimate the hazard ratios for mortality for each predictor variable, Cox proportional hazards regression models were fitted. Mortality risk was expressed as the incident rate ratio (IRR). An IRR of 1.1 suggests a 10% increase in the death rate with each unit increase in the raw test score compared with the reference category, adjusting for all covariates. The widely used clinical cutoff of 16 was used to divide the group into those with and those without significant levels of depressive symptoms (9, 29). Nonresponders and those deceased at wave 3 were identified and compared with responders at wave 3 on demographic variables and depression to enable an estimation of bias introduced by nonresponse.

For the longitudinal analyses, the sample was divided into four groups based on whether participants scored below or above the CES-D cutoff at wave 1 and wave 3. Those below the cutoff at both waves were classified in a group called "never depressed," those below the cutoff at wave 1 and above the cutoff at wave 3 were classified in a group called "incident depression," those above the cutoff at wave 1 and below it at wave 3 were classified in a group called "remitted depression," and those above the cutoff at both waves were classified in a group called "chronic depression." Mortality rates were compared for these four groups before and after the inclusion of confounding and explanatory variables.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 
Sample Characteristics
Those surviving (N = 1050) until July 30, 2000, with complete CES-D data comprised 55% of the sample at wave 1 and were followed for an average of 2802.68 days. Decedents (N = 860) were followed for an average of 1413.63 days. Table 1 shows the descriptive statistics for demographic variables, smoking, alcohol, medical conditions, medications, and CES-D for the group who survived until July 30, 2000, and the group who had died before this date. Unadjusted comparisons between groups showed that survivors were more likely to have been younger, female, and nonsmokers; less likely to have reported having had medical conditions; likely to be taking fewer medications; and likely to have lower CES-D scores at wave 1.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Descriptive Statistics of Wave 1 Variables According to Survival Group
 
Level of Depressive Symptoms as a Predictor of Mortality
A proportional hazards model of depressive symptoms as a risk factor for mortality was tested on the whole sample and then separately for males and females. It included potentially confounding variables (age, sex, marital status, education) and explanatory variables (BMI, smoking, alcohol, medical conditions, medications).

Age and sex were the only two demographic variables associated with mortality, with men having a greater risk than women. Alcohol and BMI were not associated with mortality; but smokers were at greater risk for mortality, and this effect was present for both men and women. Of the medical conditions, history of cancer and transient ischemic attacks were associated with mortality for all groups; diabetes was associated with mortality for women only. Higher levels of medication were associated with mortality for men and women. After controlling for all potentially confounding and explanatory variables, scores above the clinical cutoff for depression on the CES-D were associated with mortality in the full sample (Table 2). However, this effect was fully explained by depressed men, who had a significantly increased risk of mortality.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Associations of Risk Factors Measured at wave 1 With Mortality For Full Sample, Men, and Women
 
Nonresponse at Wave 3
Table 3 shows the demographic and depression statistics at wave 1 for four groups, including those who completed the interview at wave 3, those who did not participate in the interview at wave 3 (interview nonresponders), those who were deceased at wave 3, and those who completed part of the interview at wave 3 but for whom CES-D data were missing (CES-D nonresponders). Of those who were interview nonresponders, 8 could not be contacted, 29 had cognitive impairment or dementia severe enough to preclude participation, and 109 refused to participate.


View this table:
[in this window]
[in a new window]
  		TABLE 3. Group Differences in Demographics and Depressive Symptoms Between Responders, NonResponders, and Deceased Participants at Wave 3
 
Interview nonresponders were not significantly older than responders, whereas both those who were deceased at wave 3 (t(1909) = 4.75, p < .001) and CES-D nonresponders were significantly older (t(1909) = 1.76, p < .001) than responders. Gender distribution differed between groups ({chi}2 = 10.56, df = 3, p < .001) with a greater proportion of men having died. Those deceased by wave 3 also had higher levels of depressive symptoms at wave 1 (t(1909) = 3.29, p < .001), as did CES-D nonresponders (t(1909) = 1.66, p < .049), and had higher rates of classification of depression on the CES-D cutoff ({chi}2 = 31.35, df = 3, p < .001). Groups did not differ in education, but there was a difference in marital status ({chi}2 = 119.66, df = 9, p = .020). These results suggest that CES-D nonresponders at wave 3 were likely to have elevated CES-D scores and that there was a bias toward underestimating chronic depression in the sample.

Characteristics of Change in Depressive Status Groups
Table 4 shows the demographic characteristics and CES-D scores at wave 1 and wave 3 for the four groups classified on the basis of the CES-D cutoff (never depressed, incident depression, remitted depression, and chronic depression).


View this table:
[in this window]
[in a new window]
  		TABLE 4. Sample Characteristics at Wave 1 of the Change in Depressive Status Groups
 
The incident depression and chronic depression groups were older than the never depressed group (t(1447) = 1.89, p < .01 and t(1447) = 2.11, p < .01). Women were overrepresented in the incident depression and chronic depression groups ({chi}2(3), = 11.61, p < .01), and the groups differed significantly in marital status ({chi}2(9) = (31.70), p = < 0.01) with fewer widows in the never depressed group. The chronic depression group had an average score greater than 22 on the CES-D at both waves.

In the age-adjusted analyses of men shown in Figure 1, the incident depression group had a significant increased risk of mortality (IRR = 2.74, CI = 1.99–3.78), as did the chronic depression group (IRR = 2.12, CI = 1.42–3.17). In the age-adjusted analyses for women shown in Figure 2, the incident depression group had a significantly increased risk of mortality (IRR = 1.59, CI = 1.08–2.33), but the chronic depression and remitted depression groups did not have an increased risk of mortality.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 1. Survival curves for change in depressive status groups (men). Cum indicates cumulative.

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Survival curves for change in depressive status groups (women). Cum indicates cumulative.

 
Proportional Hazards Analysis of Change in Depressive Status Groups
Table 5 shows the unadjusted IRRs for the whole sample and for men and women separately and the results with successive control of covariates. In the unadjusted analyses of the whole sample, the incident depression and chronic depression groups had higher rates of mortality compared with the never depressed group, but there was no increased mortality rate in the remitted depression group and the never depressed group. This pattern held when the analysis was conducted on men. When the analyses was conducted for women, the incident depression group was the only group to have an increased risk of mortality before adjusting for any covariates.


View this table:
[in this window]
[in a new window]
  		TABLE 5. Mortality Rate Ratios for Change in Depressive Status Groups for Men and Women With Successive Adjustment for Mediating Variables
 
Successive adjustment for demographic variables, smoking, alcohol, BMI, and finally medical conditions did not change the pattern observed among the groups in the analysis of the whole sample. That is, the incident depression and the chronic depression groups had a higher risk of mortality than the never depressed group, but the remitted depression group did not. This pattern became more complicated when men and women were analyzed separately. For men, all three comparison groups had a greater risk of mortality than the never depressed group once demographic variables were controlled. Because age was the only significant demographic variable, this result suggests that age was confounding the effect of the remitted depression group in the unadjusted analyses. A similar pattern occurred when alcohol, smoking, and BMI were controlled; however, once medical conditions were also controlled, the remitted depression and chronic depression groups were no longer associated with an increased risk for mortality.

For women, the incident depression group had a higher risk for mortality that was not obtained once demographic variables were controlled. Together these results suggest that the overall effect of depression on mortality observed in the total sample is largely accounted for by the male gender.

Post hoc analyses were conducted to determine whether becoming widowed between wave 1 and wave 3 was a significant predictor of mortality and a possible explanation for the differing relationship between the change in depressive status groups and mortality. One hundred fifteen participants were widowed between wave 1 and wave 2, 82 of whom did not score above the cutoff of 16 on the CES-D at wave 1 or wave 3. Becoming widowed was not a significant predictor of mortality in any analysis.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
ACKNOWLEDGMENTS
 REFERENCES
 
Consistent with previous studies, we found a significant independent effect of depression on mortality for men but not women. The present study controlled for seven medical conditions that may be associated with depression and mortality. Of these, self-reported transient ischemic attacks, followed by cancer, were associated with the largest risk for mortality, whereas stroke, heart disease, hypertension, and heart conditions were not. This result differs from other studies that have found CVD to be associated with mortality. Our finding may have been because of our measure relating to the history of CVD whereas other studies have measured prevalent CVD (30, 31). Nevertheless, the present study included a wide range of medical conditions that have been associated with depression and that may have overlapping symptoms. Control of this range of factors ensured that the significant yet small effect of depression on mortality is a robust finding, suggesting that depressed mood may be causally related to changes in physical status in men. The lack of an effect of alcohol consumption on mortality in the present study may also be due to the lack of sensitivity of the measure used.

A difficulty of longitudinal studies of depression and mortality is the bias introduced by nonresponse and sample attrition (32). Evaluation of nonresponders at wave 3 in the present study showed that nonresponders had higher rates of depression at wave 1 of the study, suggesting that our observations underestimate the number of chronically depressed participants. This may explain why our hypothesis of increased mortality among the chronic depression group was not supported.

Although women were overrepresented in the incident depression and chronic depression groups, analyses conducted on women showed no effect of change in depressive status group after demographics were controlled. In comparison, men in the chronic depression and remitted depression groups had higher rates of mortality, but these were explained by medical conditions. Men in the incident depression group had a higher rate of mortality than men in the never depressed group that was independent of all other factors. Penninx et al. (14) also found that incident depression in men was associated with an increased risk of mortality due to cardiovascular disease. The possibility that onset of depression in men may reflect widowhood was explored but was not significant.

Even though medical conditions explained the association in some subgroups, a consistent independent effect of depression on mortality in men with incident depression was observed. This is consistent with findings from Penninx et al. (14), who also found that incident depression in men was associated with poorer outcomes, although Thomas et al. (21) did not find this result over a 12-month follow-up interval. The analysis of change in depressive status provided a more complex picture of the nature of the association between depression and mortality than analyses based on a single occasion of measurement and allowed for further investigation of gender differences. Overall the results suggest that the course of depressive illness is more reliable for determining the mortality of depression than a single occasion of measurement.

The present study is limited by the lack of a clinical measure of depression and the lack of data on cause of mortality despite having a large sample and using an instrument designed specifically for community samples that has been used in a number of similar studies. Countering these limitations, this study drew from a population-based sample that included a large number of very old adults and had an 8-year follow-up period. Another strength of the present study is the evaluation of nonresponse in relation to the measurement of depression over time, showing that chronic depression is likely to be underestimated. Our findings generally confirm previous studies showing that late-life depression occurs more often in women but has greater negative outcomes for men. We also replicated the finding that remitted depression was not associated with mortality and suggest that this finding has both clinical and methodological implications. Clinically it suggests that treating depression in very old adults may reduce the risk of mortality; methodologically it suggests that depression should be assessed more than once when evaluating associated morbidity and mortality. Further research is required to determine the benefits of remitted depression and the factors underlying the poorer outcomes of incident depression in men.


   ACKNOWLEDGMENTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
REFERENCES
 
This study was partly funded by a National Health and Medical Research Council of Australia Clinical Research Fellowship (Number 987100), the South Australian Health Commission, the Australian Rotary Health Research Fund, and by a grant from the US National Institutes of Health (Grant AG 08523-02). We gratefully acknowledge the men and women who participated in this study, the assistance of Sabine Schreiber, and the Epidemiology Branch of the Department of Health and Human Services in South Australia.

Received for publication June 22, 2001.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGMENTS
 REFERENCES
 

  1. Ariyo AA, Haan M, Tangen CM, Rutledge JC, Cushman M, Dobs A, Furberg CD. Depressive symptoms and risks of coronary heart disease and mortality in elderly Americans. Cardiovascular Health Study Collaborative Research Group. Circulation 2000; 102: 1773–9.[Abstract/Free Full Text]
  2. Black SA, Markides KS. Depressive symptoms and mortality in older Mexican Americans. Ann Epidemiol 1999; 9: 45–52.[CrossRef][Medline]
  3. Fawcett J. The morbidity and mortality of clinical depression. Int Clin Psychopharmacol 1993; 8: 217–20.[Medline]
  4. Herrmann C, Brand-Driehorst S, Buss U, Ruger U. Effects of anxiety and depression on 5-year mortality in 5,057 patients referred for exercise testing. J Psychosom Res 2000; 48: 455–62.[CrossRef][Medline]
  5. Janzing JG, Bouwens JM, Teunisse RJ, Van’t Hof MA, Zitman FG. The relationship between depression and mortality in elderly subjects with less severe dementia. Psychol Med 1999; 29: 979–83.[CrossRef][Medline]
  6. Kimmel PL, Peterson RA, Weihs KL, Simmens SJ, Alleyne S, Cruz I, Veis JH. Multiple measurements of depression predict mortality in a longitudinal study of chronic hemodialysis outpatients. Kidney Int 2000; 57: 2093–8.[CrossRef][Medline]
  7. Kripke DF. Mortality risk of major depression [letter]. Am J Psychiatry 1995; 152: 962.[Medline]
  8. Murphy E, Smith R, Lindesay J, Slattery J. Increased mortality rates in late-life depression. Br J Psychiatry 1998; 152: 347–53.[Abstract/Free Full Text]
  9. Penninx BW, Geerlings SW, Deeg DJ, van Eijk JT, van Tilburg W, Beekman AT. Minor and major depression and the risk of death in older persons. Arch Gen Psychiatry 1999; 56: 889–95.[Abstract/Free Full Text]
  10. Schofield PW, Jacobs D, Marder K, Sano M, Stern Y. The validity of new memory complaints in the elderly. Arch Neurol 1997; 54: 756–9.[Abstract/Free Full Text]
  11. Schoevers RA, Geerlings MI, Beekman AT, Penninx BW, Deeg DJ, Jonker C, Van Tilburg W. Association of depression and gender with mortality in old age: results from the Amsterdam Study of the Elderly (AMSTEL). Br J Psychiatry 2000; 177: 336–42.[Abstract/Free Full Text]
  12. Schulz R, Beach SR, Ives DG, Martire LM, Ariyo AA, Kop WJ. Association between depression and mortality in older adults: the Cardiovascular Health Study. Arch Intern Med 2000; 160: 1761–8.[Abstract/Free Full Text]
  13. Zheng D, Macera CA, Croft JB, Giles WH, Davis D, Scott WK. Major depression and all-cause mortality among white adults in the United States. Ann Epidemiol 1997; 7: 213–8.[CrossRef][Medline]
  14. Penninx BW, Guralnik JM, Mendes de Leon CF, Pahor M, Visser M, Corti MC, Wallace RB. Cardiovascular events and mortality in newly and chronically depressed persons over 70 years of age. Am J Cardiol 1998; 81: 988–94.[CrossRef][Medline]
  15. Whooley MA, Browner WS. Association between depressive symptoms and mortality in older women. Study of Osteoporotic Fractures Research Group. Arch Intern Med 1998; 158: 2129–35.[Abstract/Free Full Text]
  16. Wulsin LR, Vaillant GE, Wells VE. A systematic review of the mortality of depression. Psychosom Med 1999; 61: 6–17.[Abstract/Free Full Text]
  17. Beekman AT, Copeland JR, Prince MJ. Review of community prevalence of depression in later life. Br J Psychiatry 1999; 174: 307–11.[Abstract/Free Full Text]
  18. Sonnenberg CM, Beekman AT, Deeg DJ, van Tilburg W. Sex differences in late-life depression. Acta Psychiatr Scand 2000; 101: 286–92.[CrossRef][Medline]
  19. Pulska T, Pahkala K, Laippala P, Kivela SL. Depressive symptoms predicting six-year mortality in depressed elderly Finns. Int J Geriatr Psychiatry 2000; 15: 940–6.[CrossRef][Medline]
  20. Mendes de Leon CF, Krumholz HM, Seeman TS, Vaccarino V, Williams CS, Kasl SV, Berkman LF. Depression and risk of coronary heart disease in elderly men and women: New Haven EPESE, 1982–1991. Established Populations for the Epidemiologic Studies of the Elderly. Arch Intern Med 1998; 158: 2341–8.[Abstract/Free Full Text]
  21. Thomas C, Kelman HR, Kennedy GJ, Ahn C, Yang CY. Depressive symptoms and mortality in elderly persons. J Gerontol 1992; 47: S80–7.[Abstract]
  22. Wassertheil-Smoller S, Applegate WB, Berge K, Chang CJ, Davis BR, Grimm R Jr, Kostis J, Pressel S, Schron E. Change in depression as a precursor of cardiovascular events. SHEP Cooperative Research Group (Systolic Hypertension in the Elderly Program). Arch Intern Med 1996; 156: 553–61.[Abstract/Free Full Text]
  23. Jorm AF. Is depression a risk factor for dementia or cognitive decline? A review. Gerontology 2000; 46: 219–27.[CrossRef][Medline]
  24. Geerlings MI, Deeg DJ, Penninx BW, Schmand B, Jonker C, Bouter LM, van Tilburg W. Cognitive reserve and mortality in dementia: the role of cognition, functional ability and depression. Psychol Med 1999; 29: 1219–26.[CrossRef][Medline]
  25. Penninx BW, Guralnik JM, Ferrucci L, Simonsick EM, Deeg DJ, Wallace RB. Depressive symptoms and physical decline in community-dwelling older persons. JAMA 1998; 279: 1720–6.[Abstract/Free Full Text]
  26. Anstey KJ, Luszcz MA, Giles LC, Andrews GR. Demographic, health, cognitive, and sensory variables as predictors of mortality in very old adults. Psychol Aging 2001; 16: 3–11.[CrossRef][Medline]
  27. Luszcz MA, Bryan J, Kent P. Predicting episodic memory performance of very old men and women: contributions from age, depression, activity, cognitive ability, and speed. Psychol Aging 1997; 12: 340–51.[CrossRef][Medline]
  28. Radloff L. The CES-D scale: a self-report depression scale for research in the general population. J Appl Psychol Measure 1977; 3: 385–401.
  29. Weissman MM, Sholomskas D, Pottenger M, Prusoff BA, Locke BZ. Assessing depressive symptoms in five psychiatric populations: a validation study. Am J Epidemiol 1977; 106: 203–14.[Abstract/Free Full Text]
  30. Frasure-Smith N, Lesperance F. Coronary artery disease, depression and social support only the beginning. Eur Heart J 2000; 21: 1043–5.[Free Full Text]
  31. Stein PK, Carney RM, Freedland KE, Skala JA, Jaffe AS, Kleiger RE, Rottman JN. Severe depression is associated with markedly reduced heart rate variability in patients with stable coronary heart disease. J Psychosom Res 2000; 48: 493–500.[CrossRef][Medline]
  32. Fitzmaurice GM, Lipsitz SR, Molenberghs G, Ibrahim JG. Bias in estimating association parameters for longitudinal binary responses with drop-outs. Biometrics 2001; 57: 15–21.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Psychosom. Med.Home page
K. J. Anstey, R. Burns, P. Butterworth, T. D. Windsor, H. Christensen, and P. Sachdev
Cardiovascular Risk Factors and Life Events as Antecedents of Depressive Symptoms in Middle and Early-Old Age: Path Through Life Study
Psychosom Med, November 1, 2009; 71(9): 937 - 943.
[Abstract] [Full Text] [PDF]

Home page
 Psychosom. Med.Home page
J.-P. Zhang, B. Kahana, E. Kahana, B. Hu, and L. Pozuelo
Joint Modeling of Longitudinal Changes in Depressive Symptoms and Mortality in a Sample of Community-Dwelling Elderly People
Psychosom Med, September 1, 2009; 71(7): 704 - 714.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. PsychiatryHome page
J. Ryan, I. Carriere, K. Ritchie, R. Stewart, G. Toulemonde, J.-F. Dartigues, C. Tzourio, and M.-L. Ancelin
Late-life depression and mortality: influence of gender and antidepressant use
The British Journal of Psychiatry, January 1, 2008; 192(1): 12 - 18.
[Abstract] [Full Text] [PDF]

Home page
 J. Epidemiol. Community HealthHome page
L. C Giles, G. F V Glonek, M. A Luszcz, and G. R Andrews
Effect of social networks on 10 year survival in very old Australians: the Australian longitudinal study of aging
J Epidemiol Community Health, July 1, 2005; 59(7): 574 - 579.
[Abstract] [Full Text] [PDF]

Home page
 Psychosom. Med.Home page
S. C. Matthews, R. A. Nelesen, and J. E. Dimsdale
Depressive Symptoms Are Associated With Increased Systemic Vascular Resistance to Stress
Psychosom Med, July 1, 2005; 67(4): 509 - 513.
[Abstract] [Full Text] [PDF]

Home page
 Psychosom. Med.Home page
A. Nicholson, R. Fuhrer, and M. Marmot
Psychological Distress as a Predictor of CHD Events in Men: The Effect of Persistence and Components of Risk
Psychosom Med, July 1, 2005; 67(4): 522 - 530.
[Abstract] [Full Text] [PDF]

Home page
 Psychosom. Med.Home page
S. A. Everson-Rose, J. S. House, and R. P. Mero
Depressive Symptoms and Mortality Risk in a National Sample: Confounding Effects of Health Status
Psychosom Med, November 1, 2004; 66(6): 823 - 830.
[Abstract] [Full Text] [PDF]

Home page
 J Aging HealthHome page
N. A. McKeen, J. G. Chipperfield, and D. W. Campbell
A Longitudinal Analysis of Discrete Negative Emotions and Health-Services Use in Elderly Individuals
J Aging Health, April 1, 2004; 16(2): 204 - 227.
[Abstract] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Anstey, K. J.
Right arrow Articles by Luszcz, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Anstey, K. J.
Right arrow Articles by Luszcz, M. A.
Related Collections
Right arrow Other Epidemiology
Right arrow General Sexual Medicine Issues

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS