ORIGINAL INVESTIGATION. Depression in Older People in Rural China. the most common psychiatric

ORIGINAL INVESTIGATION Depression in Older People in Rural China Ruoling Chen, MD, PhD; Li Wei, MD, PhD; Zhi Hu, MD, MSc; Xia Qin, BSc; John R. M. Copeland, MD, FRCP; Harry Hemingway, MD, FRCP Background: In Western countries depression is the most common psychiatric condition in older persons and related to low socioeconomic status and low social supports. Along with social deprivation, the rural communities in China retain many Chinese traditions that involve high levels of social supports. Studying such a population might offer insights into the cause and prevention of depression that may be applicable in developing and developed countries. Methods: Using a cross-sectional, household-based, community survey in rural China, we aimed to determine the prevalence of and risk factors for depression among older people who had a low income (mean annual per capita income of about US $280) and high social support in 16 villages in Anhui Province. Participants included 754 men and 846 women aged 60 years or older. Depression was diagnosed using the Geriatric Mental State and the Automated Geriatric Examination for Computer Assisted Taxonomy. Risk factors, collected from the standard questionnaire and physical measurements, were examined in a stepwise multiple logistic regression model. Results: The prevalence of depression (world age standardized) was 6.0% (95% confidence interval [CI], 4.8%- 7.3%). Of all persons, 1374 older persons (85.9%) living with family members. Depression was significantly and independently associated with female gender, low family income, lack of social support, relationship problems, poor health status, and adverse life events in the past 2 years. The risk of depression increased in those eating meat less than once a week (multiple-adjusted odds ratio, 2.20; 95% CI, 1.20-4.03), not watching television (odds ratio, 1.76; 95% confidence interval, 1.03-3.00), and having undetected hypertension (odds ratio, 1.78; 95% CI, 1.05-3.01). Conclusions: Older people in rural China have a lower risk of depression than those in Western countries. Low socioeconomic status showed a dose-response relationship with depression, and social supports were much more common, which were protective for depression. Further exploration of Chinese culture and tradition may yield universal insights into preventive factors for depression in older people. Arch Intern Med. 2005;165:2019-2025 Author Affiliations: Department of Epidemiology and Public Health, Royal Free and University College Medical School, University College London, London (Drs Chen and Hemingway), and Department of Psychiatry, University of Liverpool, Liverpool (Dr Copeland), England; Medicines Monitoring Unit, Ninewells Hospital and Medical School, University of Dundee, Dundee, Scotland (Dr Wei); and School of Health Administration, Anhui Medical University, Anhui, Hefei, China (Drs Chen and Hu and Ms Qin). DEPRESSIVE DISORDERS ARE the most common psychiatric condition found in older people in the Western world. 1 The functional impairment, decreased quality of life, and increased mortality caused by depression impose an immense burden on individuals, communities, and health services. Globally, depression accounts for as much disease burden as ischemic heart disease, 2,3 and it has been projected to become the second most common cause of disability by 2020. 4 In Western countries, depression in older persons is related to low socioeconomic status (SES), low social support, poor health status, disability, adverse life events, and female gender. 5-7 In many Western populations, low social position is associated with low social support. 8 With an estimated population of 1.3 billion, China is the most populous country in the world. Since its reform in 1978, China has experienced rapid economic growth and an increase in life expectancy, and the population as a whole is aging. 9 In rural areas, where 900 million peasants live, the mean annual income (range, US $140-$340) is about 2 to 5 times lower than that in urban areas (range, US $412-$652). 9 Along with low SES and deprivation, rural communities retain many Chinese traditions that involve high levels of social support. Living with family members, having contact with neighbors and family outside the home, and welcoming community participation and help when needed are important aspects of social support. However, it is not known how common depression is or which risk factors operate among the rural older population in China. Studying such a low-income population that have high social support might offer insights into the cause and prevention of depression that are applicable in developing and developed coun- 2019

Beijing Yingshang County, Anhui Province Shanghai Hong Kong Figure. Location of Yingshang County. tries. We, therefore, investigated the prevalence of depression among older people in rural China, examined its risk factors, and explored differences between China and Western countries. METHODS PARTICIPATING RURAL VILLAGES In 2003, we selected all 16 villages in the Tangdian district of Yingshang County, Anhui Province (mideastern China), as our field of study (Figure), representing a typical rural community. The district included 34 700 residents within an area of about 600 km 2. More than 95% of the population were agriculturalists; the main work of peasants was producing rice (paddy) and wheat. The mean annual per capita income was approximately 2300 Renminbi ([RMB] about US $280). More than 20% of the households lived in mud houses, about 75% had no telephone, and approximately 80% had no television set. The district had roads constructed with sand and stone, but it had no railway or industry. Within each village, there were usually 4 small subvillages, where 100 to 150 households were clustered, with a mean distance of 5 m between households. Our target population included persons 60 years or older who had lived in the villages 5 years or longer. Based on the sample size in a previous study 10 of urban China and on an estimated mean ± SE prevalence of depression of 5.5% ± 1.1% (at 95% confidence intervals [CIs]), we opted to investigate 1700 older persons in the community. We randomly chose 2 subvillages from each village for the survey. PARTICIPANTS From a residency registration list in the district, we identified 1709 eligible older persons and successively visited the households accompanied by local subvillage leaders. After explaining the aims and methods of the study, 1624 persons agreed to participate in the investigation. Beginning on April 1, 2003, our 8-member trained survey team interviewed them. Permission for interview was obtained from each person or from an adult guardian, and 1600 participants (754 men and 846 women) were successfully interviewed at home. The response rate was 93.6% (1600/1709) after excluding the 24 participants with incomplete questionnaires. The survey team completed all data collection by April 30, 2003. The study was approved by Anhui Medical University, Hefei, China, the district government, and village residency committees. 2020

ASSESSMENT OF DEPRESSION We defined depression using the Geriatric Mental State (GMS) and the Automated Geriatric Examination for Computer Assisted Taxonomy (AGECAT), 11,12 which have been widely used in western European countries 1 and in parts of Asia, South America, and Africa. 11 The Chinese version of the GMS has been validated among older Chinese in Singapore, Hong Kong, Taiwan, and mainland China. 10,11 Systematic training of the survey team in the use of the GMS was done before the interviews, as previously described. 10 In brief, 2 research workers were trained in 2001 in the use of the GMS at the Institute of Psychiatry, Beijing Medical University, Beijing, China, by an investigator who had attended a GMS-AGECAT course in Liverpool, England. They then trained the 8 members of our survey team. To identify persons with depression at a level of severity warranting intervention, 12 the GMS data were analyzed using the AGECAT, which has been validated in different countries, 11 including China. 10 To ensure the reliability of the GMS-AGECAT used in the rural community, 2 consultant psychiatrists from Hefei Psychiatric Hospital, a teaching hospital of Anhui Medical University, reexamined the depression cases identified by the GMS-AGECAT and a matched number of control subjects without mental illness who were randomly selected according to proximity to the cases home addresses. Double-blind methods were used for the validation of the GMS-AGECAT depression diagnoses. RISK FACTORS FOR DEPRESSION The interview materials, apart from the GMS, included a general health record that was derived partly from a previous health and risk factor questionnaire, 13 the Minimum Data Set in the Medical Research Council Ageing in Liverpool Project-Health Aspects (MRC-ALPHA) Study, 7 and the Scottish WHO MONICA surveys. 14 Data obtained included the following: (1) SES (including educational level, occupation, and self-assessed family income), (2) social support, (3) relationships, (4) histories of disease and medications and self-assessed physical health, (5) adverse life events occurring in the past 2 years, and (6) other information (including dietary intakes 13 and hobbies). We measured social support in the 3 domains of quality (good relationships with neighbors, parents, or others; ease in acquiring friends; and available help when needed), quantity (marital status, residence with family members, frequency of visiting children or other relatives, and contact with neighbors or friends in the village), and community participation (having any religious belief and taking part in activities and participating in community activities for seniors). Structure refers to the number of persons in the social network and the frequency of contacts with them. 7 According to standard procedures, 14 we measured systolic and diastolic blood pressure, as well as height, weight, and waist circumference, for all participants. STATISTICAL ANALYSIS The SPSS statistical package (Windows version 11.0; SPSS Inc, Chicago, Ill) was used for data analysis. The prevalence rates of depression with 95% CIs among men and women were calculated, and differences in gender and age (in 5-year increments to 80 years) groups were assessed by 2 test. The prevalence rate was age standardized using the world population in 2003 (available at: http://www.census.gov/ipc/www/idbagg.html, accessed May 2004). Agreement on depression diagnoses between the GMS-AGECAT and the psychiatrists was validated by test. The association of depression with risk factors was explored using a logistic regression model, with adjustment for age and gender. 15 A stepwise multiple logistic regression model was used to calculate odds ratios (ORs) and 95% CIs to assess the independent effects of risk factors for depression. For comparison with the MRC-ALPHA Study, 7 a prevalence relative risk of depression and some risk factors among the Chinese older population were estimated using a Poisson regression model. RESULTS We found that 1395 participants (87.2%) were illiterate, 134 (8.4%) had a primary school education, 52 (3.3%) had attended secondary school, 17 (1.1%) had a high school diploma, and 2 (0.1%) had a college education. Ninetyfive depressed cases (26 men and 69 women) were diagnosed. The crude prevalence rate of depression was 5.9% (95% CI, 4.8%-7.2%), with a 3.5% (95% CI, 2.3%-5.0%) prevalence among men and an 8.2% (95% CI, 6.4%- 10.2%) prevalence among women (P.001). Prevalences of depression among persons in the age groups 60 to 64, 65 to 69, 70 to 74, 75 to 79, and 80 years or older were 4.5% (95% CI, 2.8%-7.0%), 8.3% (95% CI, 5.5%-11.9%), 5.5% (95% CI, 3.4%-8.5%), 5.6% (95% CI, 3.3%-8.9%), and 6.4% (95% CI, 3.6%-10.3%), respectively (P=.30). Using the world population in 2003, the age-standardized prevalence of depression was 6.0% (95% CI, 4.8%-7.3%), with a 3.4% (95% CI, 2.0%-4.7%) prevalence among men and an 8.5% (95% CI, 6.5%-10.5%) prevalence among women. There was good agreement on depression diagnoses between the GMS-AGECAT and the psychiatrists. Of 95 cases and their 95 controls, 86 cases and 87 controls were reexamined by the 2 psychiatrists. Agreement was 82.1% (142/173), with a statistic of 0.64 (P.001). The sensitivity and specificity of the GMS-AGECAT depression diagnoses were 82.6% (71/86) among cases and 81.6% (71/87) among controls. Table 1 gives the frequencies of risk factors and the age- and sex-adjusted ORs for depression in our total population of older persons in the Tangdian district. One thousand three hundred seventy-four older persons (85.9%) lived with family members. Depression was significantly related to female gender, age between 65 and 69 years, and lower weight (but not waist circumference). It was further significantly related to low SES, lack of social support, problematic relationships, poor health, and adverse life events occurring in the past 2 years. Regarding SES indicators, the percentages of those eating meat once a week or more among participants with very satisfactory or satisfactory, average, or poor family incomes were 59.5%, 32.7%, and 15.1%, respectively; the percentages of those watching television were 60.8%, 47.2%, or 31.0%; and the percentages of those watching performances were 44.8%, 42.9%, and 24.6% (all trend P.001). All social support and relationship measures, except having any religious belief, were significantly related to low family income (data available from the author); for example, 76.9%, 50.9%, and 38.8% of participants in the 3 categories of family income, respectively, had good relationships with others and ease in acquiring friends. In a stepwise multiple logistic regression model, depression was significantly associated with female gender, low family income (in a dose-response manner), less 2021

Table 1. Risk Factor Frequencies and Odds Ratios (ORs) for Depression Among Older People in the Tangdian District* Depressed (n = 95) Nondepressed (n = 1505) OR (95% Confidence Interval) P Value Basic Characteristics Gender Male 26 (27.4) 728 (48.4) 1.00 Female 69 (72.6) 777 (51.6) 2.51 (1.58-3.99).001 Age, y 60-64 19 (20.0) 400 (26.6) 1.00 65-69 26 (27.4) 287 (19.1) 1.96 (1.06-3.61).05 70-74 19 (20.0) 326 (21.7) 1.22 (0.63-2.34) 75-79 16 (16.8) 271 (18.0) 1.26 (0.63-2.50) 80 15 (15.8) 221 (14.7) 1.43 (0.71-2.88) Weight, by quartile, kg 61.5-100.0 13 (13.7) 390 (25.9) 1.00 55.0-61.0 20 (21.1) 408 (27.1) 1.42 (0.69-2.90) 49.5-54.5 26 (27.4) 329 (21.9) 1.93 (0.96-3.88) 32.0-49.0 36 (37.9) 378 (25.1) 2.10 (1.05-4.21).05 Waist circumference, by quartile, cm 85.0-114.0 23 (24.2) 405 (26.9) 1.00 79.5-84.0 17 (17.9) 342 (22.7) 0.94 (0.49-1.81) 74.0-79.0 26 (27.4) 405 (26.9) 1.21 (0.68-2.17) 55.0-73.5 29 (30.5) 353 (23.5) 1.46 (0.82-2.59) Socioeconomic Status Indicators Current family income (estimated mean, US $) Very satisfactory or satisfactory ($402) 12 (12.6) 521 (34.6) 1.00 Average ($256) 47 (49.5) 788 (52.4) 2.49 (1.30-4.75).01 Poor ($98) 36 (37.9) 196 (13.0) 8.14 (4.13-16.06).001 Consumption of meat, including fresh and salted meat and fish, during the past year Once a week 16 (16.8) 609 (40.5) 1.00 Once a week 69 (72.6) 828 (55.0) 3.14 (1.80-5.49).001 Never 10 (10.5) 68 (4.5) 4.87 (2.10-11.25).001 Watching television Yes 23 (24.2) 767 (51.0) 1.00 No 72 (75.8) 738 (49.0) 2.85 (1.75-4.66).001 Social Support and Relationships Quality Good relationships with neighbors Yes 48 (50.5) 1107 (73.6) 1.00 No 47 (49.5) 398 (26.4) 2.77 (1.82-4.24).001 Good relationships with parents in the past year Yes 88 (92.6) 1457 (96.8) 1.00 No 7 (7.4) 48 (3.2) 2.33 (1.01-5.38).05 Good relationships with others and ease in acquiring friends Yes 40 (42.1) 885 (58.8) 1.00 No 55 (57.9) 620 (41.2) 1.92 (1.25-2.93).01 Available help when needed Yes 74 (77.9) 1367 (90.8) 1.00 No 21 (22.1) 138 (9.2) 3.07 (1.81-5.19).001 Quantity Marital status Married 54 (56.8) 1046 (69.5) 1.00 Not married, including divorced 5 (5.3) 105 (7.0) 1.44 (0.54-3.85) Widowed 36 (37.9) 354 (23.5) 1.80 (1.15-2.84).05 Living with Spouse, children, or grandchildren 66 (69.5) 1279 (85.0) 1.00 Parents 4 (4.2) 25 (1.7) 4.00 (1.31-12.2).05 No one 25 (26.3) 201 (13.4) 2.78 (1.70-4.57).001 Frequency of visiting children or other relatives Daily 68 (71.6) 1143 (75.9) 1.00 At least weekly 11 (11.6) 191 (12.7) 1.06 (0.55-2.06) At least monthly or less often 11 (11.6) 138 (9.2) 1.49 (0.76-2.92) Never 5 (5.3) 33 (2.2) 3.34 (1.23-9.05).05 Community participation Having any religious belief No 63 (66.3) 1076 (71.5) 1.00 Yes, taking part in activities 26 (27.4) 403 (26.8) 0.80 (0.49-1.31) Yes, not taking part in activities 6 (6.3) 26 (1.7) 3.17 (1.23-8.17).05 Participating in senior community activities during leisure time Yes 56 (58.9) 1113 (74.0) 1.00 No 39 (41.1) 392 (26.0) 1.92 (1.25-2.95).01 (continued) 2022

Table 1. Risk Factor Frequencies and Odds Ratios (ORs) for Depression Among Older People in the Tangdian District* (cont) Depressed (n = 95) Health Status Nondepressed (n = 1505) OR (95% Confidence Interval) P Value Self-assessed physical health status Very good or good 10 (10.5) 521 (34.6) 1.00 Average 41 (43.2) 670 (44.5) 3.03 (1.50-6.11).01 Poor 44 (46.3) 314 (20.9) 6.46 (3.19-13.08).001 Hypertension# No 50 (52.6) 982 (65.2) 1.00 Yes, detected 15 (15.8) 220 (14.6) 1.21 (0.66-2.21) Yes, undetected 30 (31.6) 303 (20.1) 1.87 (1.17-3.02).01 Diabetes mellitus No 91 (95.8) 1493 (99.2) 1.00 Yes 4 (4.2) 12 (0.8) 5.08 (1.56-16.53).01 Vision problems No 69 (72.6) 1243 (82.6) 1.00 Yes 26 (27.4) 262 (17.4) 1.77 (1.10-2.85).05 Adverse Life Events Occurring in the Past 2 Years Serious financial problems No 74 (77.9) 1392 (92.5) 1.00 Yes 21 (22.1) 113 (7.5) 3.82 (2.24-6.49).001 Death of closely related person No 82 (86.3) 1414 (94.0) 1.00 Yes 13 (13.7) 91 (6.0) 2.30 (1.22-4.33).01 Unpleasantness with relatives, friends, or neighbors No 86 (90.5) 1446 (96.1) 1.00 Yes 9 (9.5) 59 (3.9) 2.81 (1.33-5.94).01 Anything else severely upsetting No 46 (48.4) 1278 (84.9) 1.00 Yes 49 (51.6) 227 (15.1) 5.61 (3.64-8.63).001 Horrifying experience, including accident, fire, physical attack, etc No 76 (80.0) 1453 (96.5) 1.00 Yes 19 (20.0) 52 (3.5) 5.87 (3.25-10.60).001 *Data are given as number (percentage) unless otherwise indicated. Adjusted for age and gender. Only statistically significant values are reported. Data shown are similar for Watching performances on stage (local opera, etc). Data shown are similar for Frequency of contacting and speaking to friends in the village and Frequency of contacting and speaking to neighbors. Data shown are similar for Health obviously deteriorating in the last 2 years. #Blood pressure 160/95 mm Hg or higher or taking antihypertensive drugs. consumption of meat, absence of watching television, residence with parents or alone, bad relationships with neighbors, poor health status, undetected hypertension, and adverse life events in the past 2 years (Table 2). COMMENT Among a poor rural older population in China, we found a lower prevalence of depression than in Western countries. Within this population (94.9% of whom were peasants), the risk of depression was associated with low family income and other indicators of social position. Despite their low SES, high levels of social support in this community may contribute to the lower risk of depression. STRENGTHS AND WEAKNESSES OF THE STUDY The main strengths of the study were a typical rural community setting in China with low SES and high social support, use of the GMS-AGECAT to diagnose depression (validated by Chinese psychiatrists as having a similar statistic as in previous studies 10,11,16 ), and a high response rate to face-to-face interviews using standard questionnaires to collect data on risk factors (including blood pressure and physical measurements). We selected participants from 16 villages in one region of the country. Although the study region had levels of economic development and modernization comparable to those in other provinces of rural China, caution should be exercised in generalizing our findings to China s 90 million rural older inhabitants. A further limitation of our study was the cross-sectional design, limiting inference on causal direction. However, the risk factor associations we observed were consistent with those in longitudinal studies 5,6,15 among Western populations. Most of the putative risk factors were self reported (as in many Western studies), and this may lead to bias if depressed participants are more likely to report adverse data. In Western populations, assessments of risk factors (eg, median household income) are available from computerized records, which are nonexistent in poor rural populations. However, we sought multiple indicators of measuring deprivation (eg, less consumption of meat) and obtained objective measures of health status (eg, blood pressure, weight, and waist circumfer- 2023

Table 2. Stepwise Multiple Logistic Regression Analysis of Odds Ratios (ORs) for Depression Among Older People in the Tangdian District OR (95% Confidence Interval)* Basic Characteristics ence). Consumption of meat and blood pressure measurement were significant variables in the multiple logistic regression analysis. SOCIOECONOMIC STATUS P Value Gender Male 1.00 Female 2.19 (1.31-3.67).01 Socioeconomic Status Indicators Current family income (estimated mean, US $) Very satisfactory 1.00 or satisfactory ($402) Average ($256) 1.33 (0.67-2.64) Poor ($98) 2.49 (1.17-5.28).05 Consumption of meat, including fresh and salted meat and fish, during the past year Once a week 1.00 Once a week 2.20 (1.20-4.03).05 Never 1.25 (0.47-3.29) Watching television Yes 1.00 No 1.76 (1.03-3.00).05 Social Support and Relationships Good relationships with neighbors Yes 1.00 No 1.66 (1.04-2.66).05 Living with Spouse, children, or 1.00 grandchildren Parents 4.31 (1.27-14.70).05 No one 1.98 (1.15-3.42).05 Health Status Self-assessed physical health status Very good or good 1.00 Average 2.49 (1.19-5.22).05 Poor 3.11 (1.45-6.70).01 Hypertension No 1.00 Yes, detected 0.85 (0.44-1.66) Yes, undetected 1.78 (1.05-3.01).05 Adverse Life Events Occurring in the Past 2 Years Anything else severely upsetting No 1.00 Yes 3.54 (2.19-5.73).001 Horrifying experience, including accident, fire, physical attack, etc No 1.00 Yes 3.31 (1.70-6.43).001 * Adjusted for the multivariate factors listed in the table. Only statistically significant values are reported. Blood pressure 160/95 mm Hg or higher or taking antihypertensive drugs. The population that we studied was poor. Despite this, 33.3% of the participants rated their family income as satisfactory or very satisfactory. Within this population, deprivation was associated in a dose-response manner with depression, suggesting the importance of SES. This supports a causal role for SES in the etiology of depression in older people within a rural Chinese population. 5,15 In a study 10 of urban China, there was less poverty and, consistent with this, a lower 2.2% prevalence of depression. Social factors may have various effects at different life stages. British data show that low SES in middle age is an important predictor of morbidity, including depression, 17 and our participants in middle age would have been poorer still before the reform in China. One wonders why our much poorer population had a lower prevalence of depression than populations in western Europe (range across 7 countries,8.8%-23.6%,accordingtothegms-agecat 1 ).Three possible reasons may be considered. First, there were higher levels of social support and positive life values among older people in China. Second, working and living environments in rural areas were more relaxed (eg, less stressful work and more physical farming activity). Third, the causes of diseases within populations may differ from the factors that explain differences between populations. Nevertheless, the finding of a dose-response relationship between deprivation and depression in our population suggests that strategies for addressing depression in older people are needed in developed and developing countries. OTHER INDICATORS OF SOCIAL POSITION AND SOCIAL PARTICIPATION A higher risk of depression was associated with eating less meat and not watching television, which were related to low family income and were common in rural China. Less consumption of meat may be linked to low family income, a waning appetite, or a low intake of fish. 18 The absence of television watching could be due to insufficient money to buy a television set or a lack of interest in entertainment or newscasts, further isolating one from society and exacerbating depression. For example, in our study population 63.1% of those not participating in senior community activities during leisure time did not watch television, compared with 45.7% of those participating (P.001). Nevertheless, low family income, consumption of less meat, and absence of television watching were significant in the stepwise multiple logistic regression model, suggesting that all 3 variables are operative in depression. SOCIAL SUPPORT We speculate that aspects of Chinese culture and tradition may prevent depression in older people. For example, Chinese adages include Bringing up a son will prevent aging and One should not go far away from his or her parents. Althoughgeneticsmaycontributetothelowprevalence of depression among rural Chinese, it is unlikely to explain the prevalence differences between Chinese and Western older populations. Older Chinese persons living in the United Kingdom have a much higher 13.0% prevalence of depression, 19 consistent with the observation that Chinese emigrants adopt some aspects Western lifestyles, including less social support. Table 3 compares older populations in Yingshang, China, and in Liverpool, 7,16 reveal- 2024

Table 3. Comparison of Older People in Yingshang County and in Liverpool* Yingshang County (n = 1600) Liverpool (n = 5222) 7,16 ing large differences in social support measures. Fewer older people in China lived alone, and more of them visited children or other relatives and participated in community activities. Although the social and cultural significance of these factors is complex and cross-cultural comparisons should be made with caution, our data are consistent with a protective role of social support in rural China. The high level of social support and the low prevalence of depression imply that environmental factors may be operative in the cause of depression among older Chinese. HEALTH STATUS Prevalence Rate Ratio (95% Confidence Interval) Age-standardized 95 (6.0) 483 (10.0) 0.58 (0.46-.074) prevalence of depression Age, mean ± SD, y 70.9 ± 7.6 78.6 ± 8.4... Female gender 864 (52.9) 2760 (52.9) 1.06 (0.98-1.16) Living alone 226 (14.1) 1983 (38.0) 0.40 (0.35-0.46) Visiting children or 226 (75.7) 2043 (39.1) 1.80 (1.67-1.95) other relatives daily Participating in senior community activities 1169 (73.1) 1805 (34.6) 1.85 (1.70-2.00) *Data are given as the number (percentage) unless otherwise indicated. Adjusted for age and gender. Age-standardized prevalence of depression. Statistically significant at P.001. In Western older populations, poor physical health has been found to be a strong risk factor for depression. 6,7 Our study identified a relationship between poor health status and depression among Chinese older populations. In rural China, the primary care system is mainly carried out at clinics staffed by minimally trained medical personnel ( barefoot doctors ). In our community, there was no medical insurance for agriculturalists, and the peasants had to pay 100% of their medical fees. Poor health status was reported in 22.4% in our rural population, compared with 11.6% in urban China. 10 Interestingly, we found that depression was significantly associated with undetected hypertension. The reason is unclear and needs further investigation. Undetected hypertension may be associated with deprivation, minimal access to health care, or vascular complications in the brain. Hypertensionscreeningamongolderpopulationsmaybewarranted for preventing depression and cardiovascular disease. The lower prevalence of depression among our population compared with Western populations is unlikely to result from chance or bias. Large prospective studies are required to investigate geographic variations, environmental and genetic risk factors, and management of depression among older people across China. Exploration of Chinese culture and tradition may yield insights into preventive factors for depression in older people, of interest to Chinese and Western populations. Accepted for Publication: May 1, 2005. Correspondence: Ruoling Chen, MD, PhD, Department of Epidemiology and Public Health, Royal Free and University College Medical School, University College London, 1-19 Torrington Pl, London WC1E 6BT, England (ruoling.chen@ucl.ac.uk). Financial Disclosure: None. Funding/Support: This study was supported in part by grant 574006.G603/22085 from The Royal Society and the Universities China Committee, both in London. Dr Hemingway is supported by a National Public Health Career Scientist Award from the Department of Health, London. Previous Presentation: This study was presented at the 132nd Annual Meeting of the American Public Health Association; November 10, 2004; Washington, DC. Acknowledgment: We thank the participants and all who took part in the survey, especially Yi Dong, MD, and Haitao Xia, MD, for their validation of depression diagnoses. REFERENCES 1. Copeland JR, Beekman AT, Dewey ME, et al. Depression in Europe: geographical distribution among older people. Br J Psychiatry. 1999;174:312-321. 2. World Health Organization. The World Health Report 2002: Reducing Risks, Promoting Healthy Life. Geneva, Switzerland: World Health Organization; 2002. 3. Üstün TB, Ayuso-Mateos JL, Chatterji S, Mathers C, Murray CJL. Global burden of depressive disorders in the year 2000. Br J Psychiatry. 2004;184:386-393. 4. Murray C, Lopez A. The Global Burden of Diseases: A Comprehensive Assessment of Mortality and Disability From Diseases, Injuries and Risk Factors in 1990 and Projected to 2020. Boston, Mass: Harvard School of Public Health, World Health Organization, and World Bank; 1996. 5. Wilson KC, Chen R, Taylor S, McCracken CF, Copeland JR. Socio-economic deprivation and the prevalence and prediction of depression in older community residents: the MRC-ALPHA Study. Br J Psychiatry. 1999;175:549-553. 6. 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. 7. Copeland JR, Chen R, Dewey ME, et al. Community-based case-control study of depression in older people: cases and sub-cases from the MRC-ALPHA Study. Br J Psychiatry. 1999;175:340-347. 8. Marmot MG, Smith GD, Stansfeld S, et al. Health inequalities among British civil servants: the Whitehall II study. Lancet. 1991;337:1387-1393. 9. Woo J, Kwok T, Sze FKH, Yuan HJ. Ageing in China: health and social consequences and responses. Int J Epidemiol. 2002;31:772-775. 10. Chen R, Hu Z, Qin X, Xu XC,Copeland JR. A community-based study of depression in older people in Hefei, China: the GMS-AGECAT prevalence, case validation and socio-economic correlates. Int J Geriatr Psychiatry. 2004;19:407-413. 11. Copeland JR, Prince M, Wilson KC, Dewey ME, Payne J, Gurland B. The Geriatric Mental State examination in the 21st century. Int J Geriatr Psychiatry. 2002; 17:729-732. 12. Copeland JR, Dewey ME, Griffiths-Jones HM. A computerized psychiatric diagnostic system and case nomenclature for elderly subjects: GMS and AGECAT. Psychol Med. 1986;16:89-99. 13. Chen R, Hu Z, Seaton A. Eating more vegetables might explain reduced asthma symptoms [letter]. BMJ. 2004;328:1380. 14. Chen R, Tunstall-Pedoe H, Morrison C, Connaghan J, A Brook R. Trends and social factors in blood pressure control in Scottish MONICA surveys 1986-1995: the rule of halves revisited [published correction appears in J Hum Hypertens. 2004;18:753-754]. J Hum Hypertens. 2003;17:751-759. 15. Lorant V, Deliege D, Eaton W, Robert A, Philippot P, Ansseau M. Socioeconomic inequalities in depression: a meta-analysis. Am J Epidemiol. 2003;157: 98-112. 16. Saunders PA, Copeland JR, Dewey ME, et al. The prevalence of dementia, depression and neurosis in later life: the Liverpool MRC-ALPHA study. Int J Epidemiol. 1993;22:838-847. 17. Marmot M, Shipley M, Brunner E, Hemingway H. Relative contribution of early life and adult socioeconomic factors to adult morbidity in the Whitehall II study. J Epidemiol Community Health. 2001;55:301-307. 18. Peet M. International variations in the outcome of schizophrenia and the prevalence of depression in relation to national dietary practices: an ecological analysis. Br J Psychiatry. 2004;184:404-408. 19. McCracken CF, Boneham MA, Copeland JR, et al. Prevalence of dementia and depression among elderly people in black and ethnic minorities. Br J Psychiatry. 1997;171:269-273. 2025