Hostname: page-component-74656f9884-2dmjx Total loading time: 0 Render date: 2026-11-07T11:18:21.443Z Has data issue: false hasContentIssue false
  • English
  • Français

Age and risk for depression among the elderly: a meta-analysis of the published literature

Published online by Cambridge University Press:  15 August 2012

Ke-Xiang Zhao ,
Chang-Quan Huang ,
Qian Xiao ,
Yuan Gao ,
Qing-Xiu Liu ,
Zheng-Rong Wang ,
Yong-Hong Li  and
Yi-Zhou Xie
Ke-Xiang Zhao
Affiliation:
Department of Geriatrics, the First Affiliated Hospital, Chongqing Medical University, Chongqing, China
Chang-Quan Huang
Affiliation:
Department of Geriatrics, the Third Hospital of Mianyang, Mianyang, Sichuan, China
Qian Xiao*
Affiliation:
Department of Geriatrics, the First Affiliated Hospital, Chongqing Medical University, Chongqing, China
Yuan Gao
Affiliation:
Department of Geriatrics, the First Affiliated Hospital, Chongqing Medical University, Chongqing, China
Qing-Xiu Liu
Affiliation:
Department of Geriatrics, the Third Hospital of Mianyang, Mianyang, Sichuan, China
Zheng-Rong Wang
Affiliation:
Key Laboratory of Chronobiology of Health Ministry, Basic and Forensic School, Sichuan University, Chengdu, China
Yong-Hong Li
Affiliation:
Key Laboratory of Chronobiology of Health Ministry, Basic and Forensic School, Sichuan University, Chengdu, China
Yi-Zhou Xie
Affiliation:
Key Laboratory of Chronobiology of Health Ministry, Basic and Forensic School, Sichuan University, Chengdu, China
*
*Address for correspondence: Qian Xiao, Department of Geriatrics, the First Affiliated Hospital, Chongqing Medical University, Chongqing 400016, China. (Email XiaoQian1956@126.com)
Rights & Permissions [Opens in a new window]

Abstract

Objective

The goal of this study was to determine the relationship between age and risk for depression among the old and the oldest old.

Method

MEDLINE, EMBASE, and the Cochrane Library database were used to identify potential studies. The studies were divided into cross-sectional and longitudinal subsets. For each study, the numbers of the total participants, cases (for cross-sectional study), or incident cases (for longitudinal study) of depression in each age group were extracted and entered into Review Manager 4.2 software. Qualitative meta-analyses of cross-sectional studies and of longitudinal studies were performed. For prevalence and incidence rates of depression, odds risk (OR) and relative risk (RR) were calculated, respectively.

Results

The qualitative meta-analyses showed that, compared with younger participants (above vs. below 65 years, above vs. below 70 years, above vs. below 75 years, and above vs. below 80 years), older age groups had a significantly higher risk for depression. (All of the ORs and RRs were significant.) Compared with participants aged 55–89, those aged above 90 years had no higher risk for depression. (Neither the OR nor the RR was significant.)

Conclusions

Despite the methodological limitations of this meta-analysis, older age appears to be an important risk factor for depression in the general elderly population (aged below 80 years), but not in the oldest population (aged above 85 years).

Keywords

Agedepressionelderlymeta-analysis

Information

Type
Original Research
Information
CNS Spectrums , Volume 17 , Issue 3 , September 2012 , pp. 142 - 154
Copyright
Copyright © Cambridge University Press 2012

Introduction

Depression is a major contributor to healthcare costs associated with older populations, and is projected to be the leading cause of disease burden in older populations by the year 2020.Reference Goodwin1, Reference Katon, Lin, Russo and Unutzer2 The prevalence of depression in patients aged 65 and older may be as high as 40% in hospitalized and nursing home patients, and 8–15% in community settings.Reference Leon, Ashton, DMello and Dantz3 The prognosis of these depressive states is poor. A meta-analysis of outcomes at 24 months estimated that only 33% of subjects were well, 33% were depressed, and 21% had died.Reference Cole, Bellavance and Mansour4 Moreover, studies of depressed adults indicated that those with depressive symptoms, with or without depressive disorder, have poorer functioning that is comparable to or worse than that of people with chronic medical conditions such as heart and lung disease, arthritis, hypertension, and diabetes.Reference Gurland, Wilkder and Berkman5–Reference Wells and Burman7 In addition to poor functioning, depression increases the perception of poor health, the utilization of medical services, and healthcare costs.Reference Wells and Burman7–Reference Unutzer, Patrick and Simon9

Older age is commonly viewed as a risk factor for depression in the elderly, and this has been shown in many longitudinal and cross-sectional studies.Reference Harlow, Goldberg and Comstock10–Reference Sonnenberg, Beekman, Deeg and van Tilburg13 However, the converse conclusion was also reached by some studies.Reference Livingston, Watkin and Milne14, Reference Bruce, McAvay and Raue15 Moreover, a recent systematic review and meta-analysis showed that the odds ratio (OR) of being older as a function of increased depression was nonsignificant [OR = 1.2, 95% confidence intervals (95% CI) = 0.9–1.7].Reference Cole and Dendukuri16 However, this review included only two studies and could not conduct a definite conclusion. Therefore, the relationship between age and risk for depression among the old and the oldest old is still unclear in the literature.

Depression is a critically important issue for the elderly and those working with the elderly. As the population of the elderly and the very old increases, the number of depression cases affecting the elderly and very old individuals can be expected to rise.Reference Harpole, Williams and Olsen17 Therefore, for the prevention and treatment of depression in the elderly, it is important to investigate its risk factors. So we decided to conduct a meta-analysis in order to measure the magnitude and shape of the association between age and depression in the elderly and in the very elderly.

Methods

Search method

We have conducted a meta-analysis according to the earlier systematic review and the guidelines for reporting meta-analyses of observational studies.Reference Biessels, Staekenborg and Brunner18, Reference Stroup, Berlin and Morton19 This was one part of a best-evidence research project on depression in the elderly. In the research, we collected literature through searching MEDLINE (from the beginning of 1966), EMBASE (from the beginning of 1980), and the Cochrane Library (1990 to August 2007). The search terms (provided by the Cochrane Center) included “depression,” “elderly patients” (55 years and above), and “clinical trials.” Four researchers selected literature that involved clinical trials, depression (diagnostic criteria in a formal depression scale), and elderly patients (55 years and above). Any articles in the literature that were not clinical trials, were unrelated to depression, or did not include elderly patients were rejected. The literature selection was completed in three stages: (1) We reviewed the article titles to reject the inappropriate articles and keep those that would be potentially included; (2) we reviewed the abstracts of the articles that remained after the first stage, rejected the inappropriate articles and kept those that would be potentially included; and (3) we read the full text of the articles that remained after the second stage, rejected the inappropriate literature, and kept those articles that would be included. After this review, 6420 articles remained. These articles were classified into four subgroups according to the objective of the research program: (1) etiology- or epidemiology-related, (2) diagnostics-related, (3) therapeutics-related, and (4) prognosis-related. The search terms, search results, and classification of literature were reported previously.Reference Huang, Zhang and Dong20 As stated earlier, the selection and classification of the articles were performed by the four researchers, then each article was selected and classified by two researchers independently, and discrepancies were addressed by discussion. In this meta-analysis, we measured the magnitude and shape of the association age and depression in the elderly, so only etiology- or epidemiology-related articles might be potentially included. The inclusion and exclusion criteria are listed as follows.

Inclusion criteria

The following inclusion criteria were used for the articles used in the meta-analysis:

  1. 1. Cross-sectional and longitudinal studies where all participants were 55 years and over (the age at the end of the follow-up for longitudinal study)

  2. 2. Original research reported in English

  3. 3. Includes complete information on the prevalence or incidence of depression in different age groups

  4. 4. Use of an acceptable definition of depression

We accepted the diagnostic category of depression as applied by the authors of each study, which included the following: (1) the presence of depressive disorder, depressive symptoms, or “psychological distress,” as defined by scores above a cut point for abnormality on a standard mood scale; (2) severity of depressive disorder, depressive symptoms, or psychological distress, as defined by scores on a standard mood scale; and (3) the presence of major depression or minor depression (or dysthymia) according to Diagnostic and Statistical Manual of Mental Disorders (DSM)-IIIR, DSM-IV, or other standard psychiatric diagnostic criterion.

Exclusion criteria

Studies were excluded if they were limited to specific patient characteristics, such as convenience sampling; were based on retrospective recruitment; or if only an unstructured assessment of mood was used.

Data extraction and checking

For the longitudinal study, information about the country of study, group size at baseline and follow-up, age, proportion of men relative to women, depression criteria, exclusion criteria at baseline, length of follow-up, and number of incident cases of depression in each age group was abstracted from each report. For the cross-sectional study, information about the country of study, group size, age, proportion of men relative to women, depression criteria, exclusion criteria, and number of cases of depression in each age group was abstracted from each report. Every article included in the meta-analysis was read, and the data were extracted and cross-checked independently by two authors; discrepancies were addressed with discussion.

Statistical analysis

Data were entered into the RevMan 4.2 meta-analysis program (Cochrane Collaboration, Oxford, UK; see http://ims.cochrane.org/revman). The meta-analysis of the cross-sectional studies had the advantages of huge sample size and the ability to easily show the association between age and prevalence of depression, and the meta-analysis of the longitudinal studies had the advantage of easily conducting a causality conclusion. We conducted the meta-analysess of the cross-sectional studies and that of the longitudinal studies separately. In the meta-analysis of the cross-sectional studies, for prevalence rates of depression, odds risk (OR) and 95% confidence intervals (95% CIs) were calculated. Results were summarized using conventional Forest plots and ORs, which were stratified by features of the studies included. In the meta-analysis of longitudinal studies, for incidence rates of depression, relative risk (RRs) and 95% CIs were calculated. Results were summarized using conventional Forest plots and RRs, which were stratified by features of the studies included. Summary ORs and RRs were estimated using a random effects model.

Results

The search

Our search found 1027 potential etiology- or epidemiology-related articles. Of these, 896 articles were rejected as obviously unsuitable studies (unrelated to health status), which left 131 articles. Of these 131 articles, 92 were rejected for a variety of reasons, including lack of usable data or lack of a recognized instrument used for diagnosis. Thirty-nine studies remained and were included in the review.Reference Sonnenberg, Beekman, Deeg and van Tilburg13–Reference Bruce, McAvay and Raue15, Reference Al-Shammari and Al-Subaie21–Reference Turvey, Carney and Arndt56

Included studies

The characteristics of the 39 studies (including 10 longitudinalReference Livingston, Watkin and Milne14, Reference de Beurs, Beekman and Geerlings48–Reference Turvey, Carney and Arndt56 and 29 cross-sectionalReference Sonnenberg, Beekman, Deeg and van Tilburg13, Reference Bruce, McAvay and Raue15, Reference Al-Shammari and Al-Subaie21–Reference Woo, Ho and Lau47 studies available for meta-analysis) are summarized in Tables 1 and 2.

Table 1

Characteristics of the 29 cross-sectional studies included in the meta-analysis

CES-D Scale: Center for Epidemiologic Studies Depression Scale; DSM: Diagnostic and Statistical Manual of Mental Disorders; GMS-AGECAT: Geriatric Mental State Schedule Automated Geriatric Examination for Computer Assisted Taxonomy; Short CARE: Shortened Comprehensive Assessment and Referral Evaluation; GDS-15: Geriatric Depression Scale; SADS: Schedule for Affective Disorders and Schizophrenia; CIDI: Composite International Diagnostic Interview; MINI: Mini International Neuropsychiatric Interview; MMSE: Mini-Mental State Examination.

Table 2

Characteristics of the 10 prospective longitudinal studies included in the meta-analysis

CES-D Scale: Center for Epidemiologic Studies Depression Scale; DSM: Diagnostic and Statistical Manual of Mental Disorders; GMS-AGECAT: Geriatric Mental State Schedule Automated Geriatric Examination for Computer Assisted Taxonomy; Short CARE: Shortened Comprehensive Assessment and Referral Evaluation; GDS-15: Geriatric Depression Scale; SADS: Schedule for Affective Disorders and Schizophrenia; MMSE: Mini-Mental State Examination.

Data synthesis

We assessed this bias using a funnel plot (shown in Figure 1). The funnel plot of ORs (under a fixed-effects model) was taken from the 39 studies in Tables 1 and 2. In the absence of publication bias, the points should be symmetrical around the vertical line at the pooled ORs. The reasonably symmetrical plot suggests the absence of a publication bias.

Figure 1

Funnel plot of the 39 studies included in the meta-analysis.

Comparison of risk of depression between individuals aged 55–64 years and those aged 65 years and over

Six of the studies that were included compared the prevalence of depression between individuals aged 55–64 years and those aged 65 years and over.Reference Sonnenberg, Beekman, Deeg and van Tilburg13, Reference Al-Shammari and Al-Subaie21, Reference Blay, Andreoli, Fillenbaum and Gastal23, Reference Carvalhais, Lima-Costa and Peixoto25, Reference Chen, Li and Zhi26, Reference van der Wurff, Beekman and Dijkshoorn45 In the six studies, a total of 7004 individuals aged 55–65 years and 15017 aged 65 years and over were studied. There were 1313 and 3566 cases of depression in the groups aged 55–64 years and aged 65 years and over, respectively. After pooling these six studies, individuals aged 65 years had a higher prevalence of depression than those aged 55–64 years, OR: 1.36, 95% CI: 1.12–1.65 (Figure 2).

Figure 2

Comparison of risk of depression between individuals aged 65 years or over and those aged 55–64 years.

Comparison of risk of depression between individuals aged 55–69 years and those aged 70 years and over

Six studies compared the prevalence of depression between individuals aged 55–69 years and aged 70 years and over.Reference Al-Shammari and Al-Subaie21, Reference Blay, Andreoli, Fillenbaum and Gastal23, Reference Carvalhais, Lima-Costa and Peixoto25, Reference Chen, Li and Zhi26, Reference Kivela, Pahkala and Laippala33, Reference O'Hara, Kohout and Wallace39 In the six studies, there were 10,650 individuals aged 55–69 years and 11,875 individuals aged 70 years and over. There were 2326 and 2354 cases of depression in the groups aged 55–69 years and aged 70 years and over, respectively. After pooling these six studies the higher prevalence of depression among those aged above 70 years was borderline statistically significant, OR: 1.22, 95% CI: 0.97–1.53. Four studies compared the incidence of depression between groups aged 55–69 years and those aged 70 years or over.Reference de Beurs, Beekman and Geerlings48, Reference Giltay, Zitman and Kromhout50, Reference Harris, Cook and Victor51, Reference Roberts, Shema, Kaplan and Strawbridge54 After pooling these studies, the higher incidence of depression among those aged above 70 years was statistically significant, RR: 1.72, 95% CI: 1.37–2.16; RR: 1.43. (Figure 3).

Figure 3

Comparison of risk of depression between individuals aged 70 years or over and those aged 55–69 years.

Comparison of risk of depression between individuals aged 55–74 years and those aged 75 years and over

Nineteen studies compared the prevalence of depression between individuals aged 55–74 years and those aged 75 years or over.Reference Sonnenberg, Beekman, Deeg and van Tilburg13, Reference Bruce, McAvay and Raue15, Reference Blay, Andreoli, Fillenbaum and Gastal23–Reference Friedman, Conwell and Delavan29, Reference Gostynski, Ajdacic-Gross and Gutzwiller31, Reference Heok, Meng, Calvin and Li32, Reference Kulaksizoglu, Gurvit and Polat34–Reference McDougall, Matthews and Kvaal37, Reference O'Hara, Kohout and Wallace39–Reference Rokke and Klenow41, Reference Woo, Ho and Lau47 In the 19 studies, there were 20,534 subjects aged 55–74 years and 11,219 aged 75 years and over. There were 3767 and 2341 cases of depression in the groups aged 55–74 years and aged 75 years and over, respectively. After pooling these 19 studies, subjects aged 75 years had a higher prevalence of depression than those aged 55–74 years, OR: 1.35, 95% CI: 1.15–1.57. Two studies compared the incidence of depression between groups aged 55–74 years and aged 75 years or over. After pooling these studies, older age groups had a significantly higher incidence of depression than younger age groups, RR: 1.43, 95% CI: 1.00–2.06(Reference Harris, Cook and Victor51, Reference Schoevers, Beekman and Deeg55) (Figure 4).

Figure 4

Comparison of risk of depression between individuals aged 75 years or over and those aged 55–74 years.

Comparison of risk of depression between individuals aged 55–79 years and those aged 80 years and over

Nine studies compared the prevalence of depression between individuals aged 55–79 years and aged 80 years or over.Reference Al-Shammari and Al-Subaie21, Reference Blay, Andreoli, Fillenbaum and Gastal23, Reference Chen, Li and Zhi26, Reference Kulaksizoglu, Gurvit and Polat34, Reference O'Hara, Kohout and Wallace39, Reference Pitkälä, Kähönen-Väre and Valvanne40, Reference Valvanne, Juva, Erkinjuntti and Tilvis44, Reference Walters, Breeze and Wilkinson46, Reference Woo, Ho and Lau47 In the nine studies, there were 25,088 subjects aged 55–79 years and 10598 aged 80 years and over. There were 3928 and 1408 cases of depression in the groups aged 55–79 years and aged 80 years or over, respectively. After pooling these nine studies, subjects aged 80 years and over had a higher prevalence of depression than those aged 55–79 years, OR: 1.44, 95% CI: 1.28–1.62. Three studies compared the incidence of depression between groups aged 55–79 years and aged 80 years and over. After pooling these studies, older age groups had significantly higher incidence of depression than younger age groups, RR: 1.64, 95% CI: 1.36–1.98Reference Harris, Cook and Victor51, Reference Roberts, Shema, Kaplan and Strawbridge54, Reference Turvey, Carney and Arndt56 (Figure 5).

Figure 5

Comparison of risk of depression between individuals aged 80 years or over and those aged 55–79 years.

Comparison of risk of depression between individuals aged 55–84 years and those aged 85 years and over

Twelve of the studies included compared the prevalence of depression between individuals aged 55–84 years and those aged 85 years and over.Reference Bruce, McAvay and Raue15, Reference Chong, Tsang and Chen28–Reference Gostynski, Ajdacic-Gross and Gutzwiller31, Reference Kulaksizoglu, Gurvit and Polat34, Reference McDougall, Matthews and Kvaal37, Reference O'Hara, Kohout and Wallace39, Reference Saks, Kolk and Allev42, Reference Valvanne, Juva, Erkinjuntti and Tilvis44, Reference Walters, Breeze and Wilkinson46, Reference Woo, Ho and Lau47 In the 12 studies, there were 19,039 subjects aged 55–84 years and 4559 aged 85 years and over. There were 2072 and 658 cases of depression in the groups aged 55–84 years and 85 years and over, respectively. After pooling these 12 studies, subjects aged above 85 years had a higher prevalence of depression than those aged 55–84 years, OR: 1.52, 95% CI: 1.20–1.92. Two studies compared the incidence of depression between groups aged 55–84 years and aged 85 years and over. After pooling these studies, older age groups had a significantly higher incidence of depression than younger age groups, RR: 0.79, 95% CI: 0.23–2.77Reference Livingston, Watkin and Milne14, Reference Forsell49 (Figure 6).

Figure 6

Comparison of risk of depression between individuals aged 85 years or over and those aged 55–84 years.

Comparison of risk of depression between individuals aged 55–89 years and those ages 90 years and over, and between individuals aged aged 80–89 years and those aged 90 years and over

Five studies compared the prevalence of depression between individuals aged 55–89 years and those aged 90 years and over.Reference Al-Shammari and Al-Subaie21, Reference Girling, Barkley and Paykel30, Reference Kulaksizoglu, Gurvit and Polat34, Reference O'Hara, Kohout and Wallace39, Reference Teresi, Abrams and Holmes43 In the five studies, there were 12,439 subjects aged 55–89 years and 360 aged 90 years or over. There were 1151 and 46 cases of depression in the groups aged 55–89 years and aged 90 years or over, respectively. After pooling these studies, the difference in the prevalence of depression between the populations above and below 90 years old was not statistically significant, OR: 1.17, 95% CI: 0.67–2.07. These studies also provided the prevalence of depression in subjects aged 80–89 years old and 90 years and above. After pooling these studies, the difference in the prevalence of depression between the two age groups was not statistically significant, OR: 0.92, 95% CI: 0.60–1.41. Two studies compared the incidence of depression between groups aged 55–89 years and those aged 90 years or over, and between groups aged 80–89 years and aged 90 years or over. After pooling these studies, there were no statistically significant differences in the incidence of depression among these age groups, RR: 0.90, 95% CI: 0.59–1.37 and RR: 0.80, 95% CI: 0.55–1.18 for groups aged above 90 years old vs. below and above 90 years old vs. aged 80–89 years old, respectivelyReference Meller, Fichter and Schröppel53, Reference Turvey, Carney and Arndt56 (Figure 7).

Figure 7

Comparison of risk of depression between individuals aged 90 years or over and those aged 55–89 years, and between individuals aged 90 or over and those aged 80–89 years.

Discussion

We conducted the meta-analyses of the cross-sectional studies and the prospective longitudinal studies, respectively. The results are clear: Older age is a risk factor for depression in the general elderly population, but is not in the oldest population (over 90 years old). This is a robust finding concerning the relationship between age and risk for depression among the elderly and the oldest population.

There were some interesting findings in the present the meta-analysis. First, the meta-analysis showed the magnitude and shape of the association age and depression in the elderly as an “S” shape (see Figure 8). In the elderly population aged below 85 years the risk of depression increased along with the increase of age, but in the population aged above 85 years, the relationship between age and risk of depression was not significant. Being older is a risk factor for late life depression, and this may be explained by more disability, worse social support, worse health status, more new medical illness, more unmarried status, and lower cognitive function in the older age population, as all of these are commonly viewed as risk factors for depression. It follows that older individuals should be screened for depression since their risk is higher. Subsequently, these individuals could be targeted for interventions to abate the potentially modifiable risk factors, such as disability, social support, health status, and cognitive function.

Figure 8

Odds risks and relative risks of depression and age.

Secondly, between the general elderly and the oldest populations, there was a discrepancy in the relationship between age and risk for depression. As was shown, the relationship between age and risk of depression was not significant in the oldest population, and the reason for this is unclear. We know that the prognosis of late-life depression is poor, and there is a high mortality rate in the depressed population. Mortality in the oldest population might tend to remove those with depression and leave those without depression. Therefore, it is reasonable to infer that mortality in the oldest population might weaken the association of age with risk for depression. This should be further confirmed.

However, the meta-analysis of prospective longitudinal studies also showed that the incidence of depression was related to age in the general elderly, but not in the oldest population, so for the relationship between age and risk for depression, mortality could not completely interprete the discrepancy between the elderly and the oldest population. However, in the oldest population, the level of risk of depression was higher than that in the general population. The prevalence and incidence rates were still very high and might be extreme in certain populations. But this was only a hypothesis, and needs to be confirmed. Meanwhile, for the oldest population, there were no special diagnostic tools for depression, and diagnostic tools for depression, generally used in clinical studies, might not be available for this age group. Therefore, in the oldest population, the relationship between age and risk of depression needs to be further investigated.

Although we attempted to adhere to the guidelines for reporting meta-analyses of observational studies, this review does have four limitations.Reference Stroup, Berlin and Morton19 First, we did not hand search journals and made no attempt to identify unpublished studies, raising the possibility that some studies have been missed. Second, despite our extensive literature search, we only included MEDLINE, EMBASE, and the Cochrane Library in our search; other databases such as CINAHL and PsycINFO were not included. Moreover, we screened the articles by reading abstracts, rather than the full texts, which was also a limitation. Third, the search was limited to articles published in English. Finally, there was heterogeneity in the results, perhaps related to different definitions of depression in different studies and small study groups in some studies. Therefore, a random-effects model, which is less precise than a fixed-effects model, was used in the review. Consequently, the results of the meta-analysis for these risk factors must be interpreted cautiously.

Conclusion

After an extensive literature search on the risk of depression in older adults, we conducted both cross-sectional and longitudinal meta-analyses of the articles we found that were related to our topic and met the criteria for inclusion. We found that though being elderly is a risk factor for depression in the general elderly population (55–89 years old), the risk was not statistically significant in the oldest population (90 years old and older). Further research is needed to determine the cause of this discrepancy.

Footnotes

This work was supported by funding from National Foundation of Natural Science of China (No.81170752), the government of Sichuan province of china (2010FZ0061), health department in the government of Sichuan province, the government of MianYang City (09s001), and the Third Hospital of Mianyang. The authors thank the staff of Key Laboratory of Chronobiology of Health Ministry in Basic and Forensic School of Sichuan University, the Department of Geriatrics of the Third Hospital of Mianyang, and all participants (as well as their legal proxies) for their great contributions. Mr. Rebbo is also acknowledged for his language assistance.

Ke-Xiang Zhao and Chang-Quan Huang were equal contributors to this article.

References

1.Goodwin, RD. Association between physical activity and mental disorder among adults in the united states. Prev Med. 2003; 36: 698–703.CrossRefGoogle ScholarPubMed
2.Katon, WJ, Lin, E, Russo, J, Unutzer, J. Increased medical costs of a population-based sample of depressed elderly patients. Arch Gen Psychiatry. 2003; 60(9): 897–903.CrossRefGoogle ScholarPubMed
3.Leon, FG, Ashton, AK, DMello, DA, Dantz, B. Depression and comorbid medical illness: therapeutic and diagnostic challenges. J Fam Pract. 2003; Suppl: 19–23.Google ScholarPubMed
4.Cole, MG, Bellavance, F, Mansour, A. Prognosis of depression in elderly community and primary care populations: a systematic review and meta-analysis. Am J Psychiatry. 1999; 156: 1182–1189.CrossRefGoogle ScholarPubMed
5.Gurland, BJ, Wilkder, DE, Berkman, C. Depression and disability in the elderly: reciprocal relations and changes with age. Int J Geriatr Psychiatry. 1988; 3: 163–179.CrossRefGoogle Scholar
6.von Korff, M, Ormel, J, Katon, W, Lin, EH. Disability and depression among high utilizers of health care: a longitudinal analysis. Arch Gen Psychiatry. 1992; 49: 91–100.CrossRefGoogle ScholarPubMed
7.Wells, KB, Burman, MA. Caring for depression in America: lessons learned from early findings of the Medical Outcomes Study. Acta Psychiatr Scand. 1991; 9: 503–519.Google ScholarPubMed
8.Katon, W, von Korff, M, Lin, E, et al. . Adequacy and duration of antidepressant treatment in primary care. Med Care. 1992; 30: 67–76.CrossRefGoogle ScholarPubMed
9.Unutzer, J, Patrick, DL, Simon, G, et al. . Depressive symptoms and the cost of health services in HMO patients aged 65 and over: a 4-year prospective study. JAMA. 1997; 277: 1618–1623.CrossRefGoogle ScholarPubMed
10.Harlow, SD, Goldberg, EL, Comstock, GW. A longitudinal study of risk factors for depressive symptomatology in elderly widowed and married women. Am J Epidemiol. 1991; 134: 526–536.CrossRefGoogle ScholarPubMed
11.Green, BH, Copeland, JRM, Dewey, ME, et al. . Risk factors for depression in elderly people: a prospective study. Acta Psychiatr Scand. 1992; 86: 213–221.CrossRefGoogle ScholarPubMed
12.Gallo, JJ, Bogner, HR, Morales, KH, et al. . Depression, cardiovascular disease, diabetes, and two-year mortality among older, primary-care patients. Am J Geriatr Psychiatry. 2005; 13(9): 748–755.CrossRefGoogle ScholarPubMed
13.Sonnenberg, CM, Beekman, ATF, Deeg, DJH, van Tilburg, W. Sex differences in late-life depression. Acta Psychiatr Scand. 2000; 101: 286–292.Google ScholarPubMed
14.Livingston, G, Watkin, V, Milne, B, et al. . Who becomes depressed? The Islington study of older people. J Affect Disord. 2000; 58: 125–133.CrossRefGoogle ScholarPubMed
15.Bruce, ML, McAvay, GJ, Raue, PJ, et al. . Major depression in elderly home health care patients. Am J Psychiatry. 2002; 159: 1367–1374.CrossRefGoogle ScholarPubMed
16.Cole, MG, Dendukuri, N. Risk factors for depression among elderly community subjects: a systematic review and meta-analysis. Am J Psychiatry. 2003; 160: 1147–1156.CrossRefGoogle ScholarPubMed
17.Harpole, LH, Williams, JW Jr, Olsen, MK. Improving depression outcomes in older adults with comorbid medical illness. Gen Hosp Psychiatry. 2005; 27(1): 4–12.CrossRefGoogle ScholarPubMed
18.Biessels, GJ, Staekenborg, S, Brunner, E, et al. . Risk of dementia in diabetes mellitus: a systematic review. Lancet Neurol. 2006; 5(1): 64–74.CrossRefGoogle ScholarPubMed
19.Stroup, DF, Berlin, JA, Morton, SC, et al. . Meta-analysis of Observational Studies in Epidemiology (MOOSE): a proposal for reporting. JAMA. 2000; 283: 2008–2012.CrossRefGoogle Scholar
20.Huang, CQ, Zhang, XM, Dong, BR, et al. . Health status and risk for depression among the elderly: a meta-analysis of published literature. Age Ageing. 2009; 38(11): 1–8.Google Scholar
21.Al-Shammari, SA, Al-Subaie, A. Prevalence and correlates of depression among Saudi elderly. Int J Geriatr Psychiatry. 1999; 14: 739–747.3.0.CO;2-1>CrossRefGoogle ScholarPubMed
22.Barry, KL, Fleming, MF, Manwell, BL, et al. . Prevalence of and factors associated with current and lifetime depression in older adult primary care patients. Fam Med. 1998; 30(5): 366–371.Google ScholarPubMed
23.Blay, SL, Andreoli, SB, Fillenbaum, GG, Gastal, FL. Depression morbidity in later life: prevalence and correlates in a developing country. Am J Geriatr Psychiatry. 2007; 15(9): 790–799.CrossRefGoogle ScholarPubMed
24.Blazer, D, Williams, CD. Epidemiology of dysphoria and depression in an elderly population. Am J Psychiatry. 1980; 137: 439–444.Google Scholar
25.Carvalhais, SMM, Lima-Costa, MF, Peixoto, SV, et al. . The influence of socio-economic conditions on the prevalence of depressive symptoms and its covariates in an elderly population with slight income differences: the Bambuí Health and Aging Study (Bhas). Int J Soc Psychiatry. 2008; 54: 447–456.CrossRefGoogle Scholar
26.Chen, R, Li, W, Zhi, H, et al. . Depression in older people in rural China. Arch Intern Med. 2005; 165: 2019–2025.CrossRefGoogle ScholarPubMed
27.Chi, I, Yip, PS, Chiu, HF, et al. . Prevalence of depression and its correlates in Hong Kong's Chinese older adults. Am J Geriatr Psychiatry. 2005; 13(5): 409–416.CrossRefGoogle Scholar
28.Chong, MY, Tsang, HY, Chen, CS, et al. . Community study of depression in old age in Taiwan: prevalence, life events and socio-demographic correlates. Br J Psychiatry. 2001; 178(1): 29–35.CrossRefGoogle ScholarPubMed
29.Friedman, B, Conwell, Y, Delavan, RL. Correlates of late-life major depression: a comparison of urban and rural primary care patients. Am J Geriatr Psychiatry. 2007; 15(1): 28–41.CrossRefGoogle Scholar
30.Girling, DM, Barkley, C, Paykel, ES, et al. . The prevalence of depression in a cohort of the very elderly. J Affect Disord. 1995; 34(4): 319–329.CrossRefGoogle Scholar
31.Gostynski, M, Ajdacic-Gross, V, Gutzwiller, F, et al. . Depression among the elderly in Switzerland. Nervenarzt. 2002; 73(9): 851–860.CrossRefGoogle ScholarPubMed
32.Heok, KE, Meng, KS, Calvin, FSL, Li, TS. Comorbidity of depression in the elderly—an epidemiological study in a Chinese community. Int. J. Geriatr. Psychiatry. 1996; 11: 699–704.3.0.CO;2-8>CrossRefGoogle Scholar
33.Kivela, SL, Pahkala, K, Laippala, P. Prevalence of depression in an elderly population in Finland. Acta Psychiatr Scand. 1988; 78(4): 401–413.CrossRefGoogle Scholar
34.Kulaksizoglu, IB, Gurvit, H, Polat, A, et al. . Unrecognized depression in community-dwelling elderly persons in Istanbul. Int Psychogeriatr. 2005; 17(2): 303–312.CrossRefGoogle ScholarPubMed
35.Licht-Strunk, E, van der Kooij, KG, van Schaik, DJ, et al. . Prevalence of depression in older patients consulting their general practitioner in the Netherlands. Int J Geriatr Psychiatry. 2005; 20(11): 1013–1019.CrossRefGoogle ScholarPubMed
36.Madianos, MG, Gournas, G, Stefanis, CN. Depressive symptoms and depression among elderly people in Athens. Acta Psychiatr Scand. 1992; 86(4): 320–326.CrossRefGoogle ScholarPubMed
37.McDougall, FA, Matthews, FE, Kvaal, K, et al. . Prevalence and symptomatology of depression in older people living in institutions in England and Wales. Age Ageing. 2007; 36: 562–568.CrossRefGoogle ScholarPubMed
38.Naarding, P, de Koning, I, dan Kooten, F, et al. . Depression in vascular dementia. Int J Geriatr Psychiatry. 2003; 18(4): 325–330.CrossRefGoogle ScholarPubMed
39.O'Hara, MW, Kohout, FJ, Wallace, RB. Depression among the rural elderly: a study of prevalence and correlates. J Nerv Ment Dis. 1985; 173(10): 582–589.CrossRefGoogle ScholarPubMed
40.Pitkälä, K, Kähönen-Väre, M, Valvanne, J, et al. . Long-term changes in mood of an aged population: repeated Zung-tests during a 10-year follow-up. Arch Gerontol Geriatr. 2003; 36(2): 185–195.CrossRefGoogle ScholarPubMed
41.Rokke, PD, Klenow, DJ. Prevalence of depressive symptoms among rural elderly: examining the need for mental health services. Psychotherapy. 1998; 35(4): 545–558.CrossRefGoogle Scholar
42.Saks, K, Kolk, H, Allev, R, et al. . Health status of the older population in Estonia. Croat Med J. 2001; 42(6): 663–668.Google ScholarPubMed
43.Teresi, J, Abrams, R, Holmes, D, et al. . Prevalence of depression and depression recognition in nursing homes. Soc Psychiatry Psychiatr Epidemiol. 2001; 36(12): 613–620.CrossRefGoogle ScholarPubMed
44.Valvanne, J, Juva, K, Erkinjuntti, T, Tilvis, R. Community studies: major depression in the elderly: a population study in Helsinki. Int Psychogeriatr. 1996; 8: 437–443.CrossRefGoogle Scholar
45.van der Wurff, FB, Beekman, AT, Dijkshoorn, H, et al. . Prevalence and risk-factors for depression in elderly Turkish and Moroccan migrants in the Netherlands. J Affect Disord. 2004; 41: 33–41.CrossRefGoogle Scholar
46.Walters, K, Breeze, E, Wilkinson, P, et al. . Local area deprivation and urban-rural differences in anxiety and depression among people older than 75 years in Britain. Am J Public Health. 2004; 94(10): 1768–1774.CrossRefGoogle ScholarPubMed
47.Woo, J, Ho, SC, Lau, J, et al. . The prevalence of depressive symptoms and predisposing factors in an elderly Chinese population. Acta Psychiatr Scand. 1994; 89(1): 8–13.CrossRefGoogle Scholar
48.de Beurs, E, Beekman, A, Geerlings, S, et al. . On becoming depressed or anxious in late life: similar vulnerability factors but different effects of stressful life events. Br J Psychiatry. 2001; 179: 426–431.CrossRefGoogle ScholarPubMed
49.Forsell, Y. Predictors for depression, anxiety and psychotic symptoms in a very elderly population: data from a 3-year follow-up study. Soc Psychiatry Psychiatr Epidemiol. 2000; 35(6): 259–263.CrossRefGoogle Scholar
50.Giltay, EJ, Zitman, FG, Kromhout, D. Dispositional optimism and the risk of depressive symptoms during 15 years of follow-up: the Zutphen Elderly Study. J Affect Disord. 2006; 91: 45–52.CrossRefGoogle ScholarPubMed
51.Harris, T, Cook, DG, Victor, C, et al. . Onset and persistence of depression in older people—results from a 2-year community follow-up study. Age Ageing. 2006; 35: 25–32.CrossRefGoogle ScholarPubMed
52.Kivela, SL, Kongas-Saviard, P, Kimmo, P, et al. . Health, health behaviour and functional ability predicting depression in old age: a longitudinal study. Int J Geriatr. 2006; 21: 237–245.Google Scholar
53.Meller, I, Fichter, MM, Schröppel, H. Risk factors and psychosocial consequences in depression of octo- and nonagenerians: results of an epidemiological study. Eur Arch Psychiatry Clin Neurosci. 1997; 17: 278–287.CrossRefGoogle Scholar
54.Roberts, RE, Shema, SJ, Kaplan, GA, Strawbridge, WJ. Sleep complaints and depression in an aging cohort: a prospective perspective. Am J Psychiatry. 2000; 157(1): 81–88.CrossRefGoogle Scholar
55.Schoevers, RA, Beekman, AT, Deeg, DJ, et al. . Risk factors for depression in later life: results of a prospective community based study (AMSTEL). J Affect Disord. 2000; 59(2): 127–137.CrossRefGoogle ScholarPubMed
56.Turvey, C, Carney, S, Arndt, R, et al. . Conjugal loss and syndromal depression in a sample of elders aged 70 years or older. Am J Psychiatry. 1999; 156: 1596–1601.CrossRefGoogle ScholarPubMed
Figure 0

Table 1 Characteristics of the 29 cross-sectional studies included in the meta-analysis

Figure 1

Table 2 Characteristics of the 10 prospective longitudinal studies included in the meta-analysis

Figure 2

Figure 1 Funnel plot of the 39 studies included in the meta-analysis.

Figure 3

Figure 2 Comparison of risk of depression between individuals aged 65 years or over and those aged 55–64 years.

Figure 4

Figure 3 Comparison of risk of depression between individuals aged 70 years or over and those aged 55–69 years.

Figure 5

Figure 4 Comparison of risk of depression between individuals aged 75 years or over and those aged 55–74 years.

Figure 6

Figure 5 Comparison of risk of depression between individuals aged 80 years or over and those aged 55–79 years.

Figure 7

Figure 6 Comparison of risk of depression between individuals aged 85 years or over and those aged 55–84 years.

Figure 8

Figure 7 Comparison of risk of depression between individuals aged 90 years or over and those aged 55–89 years, and between individuals aged 90 or over and those aged 80–89 years.

Figure 9

Figure 8 Odds risks and relative risks of depression and age.