Stress, loneliness, and depression in Taiwanese rural community-dwelling elders

International Journal of Nursing Studies 38 (2001) 339–347 Stress, loneliness, and depression in Taiwanese rural community-dwelling elders $ Jing-Jy Wanga,*, Mariah Snyderb, Merrie Kaasb b a FooYin Institute of Technology, 151 Chinh-hsuen Road, Ta-Liao Hsiang, Kaohsiung Hsien, 83101, Taiwan School of Nursing, University of Minnesota, 308 Harvard Street, 6-101WDH, Minneapolis, MN 55455, USA Received 1 February 2000; received in revised form 30 March 2000; accepted 14 April 2000 Abstract This study examined the stressors and stress levels, loneliness, and depression experienced by Taiwanese rural elders and determine if differences in the level of stress were associated with specific person-environmental characteristics. The relationship between stress and depression was also explored. A total of 201 older rural community residents participated in the study. Findings showed that rural elders experienced a number of stressors related to health and family issues. Differences in the degree of stressfulness of individuals differed by gender, educational level, living arrangement, and socioeconomic status (SES). Findings in this study also showed that there was a high association among stress level, depressive symptoms, and mood status. The findings of this study will assist community health policy-makers in determining the need for health-related services for rural elders experiencing these problems. # 2001 Elsevier Science Ltd. All rights reserved. Keywords: Taiwanese; Stress; Loneliness; Depression 1. Introduction The number of elders in countries throughout the world is increasing. This is true in Taiwan where the percentage of elders had increased to 8.2% in 1999 (Ministry of Interior, 1999). With recent modernization and industrialization of Taiwan, the extended family has gradually disappeared and has been replaced by the nuclear family. Providing unlimited filial responsibility will less likely exist in the Chinese society of the 21st century. However, social norms , financial concerns, and care quality issues are obstacles to the placement of elders in nursing home. Thus, many elders continue to live at home with a paucity of care resources. Statistics show that 98% of Taiwanese elderly live in the $ This study was supported by Zeta Chapter Sigma Theta Tau International Research Grant. *Corresponding author. Tel.: +886-7-7811151; fax: +886-77835112. community (Guan, 1996). Also, the number of Taiwanese elderly who live with their adult children has decreased dramatically during the past 10 years from 79.1 to 65.6% (Department of Budgets, Accounting & Statistics, 1998). These data indicate that there will be more elderly living apart from their adult children in the future, especially among the rural elderly, since rural elders are more reluctant to leave their farms, friends, and relatives to live with children who have moved to a city where better job opportunities exist. Numerous studies in the United States have reported that many rural elders do not have access to private or public transportation, which limits their access to health care (Gillanders and Buss, 1993). Allen (1990) reported that fewer health and social services are available for rural home elders, which can result in limited survival and safety needs. This problem is also a reality in Taiwan because the patient/physician ratio in rural areas is 3,000-10,000 to 1 (Hsu et al., 1997). Because of the geographic and physical isolation inherent in rural areas, 0020-7489/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 0 - 7 4 8 9 ( 0 0 ) 0 0 0 7 2 - 9 340 J.-J. Wang et al. / International Journal of Nursing Studies 38 (2001) 339–347 rural elders are believed to have greater physical impairment, more chronic illnesses, and poorer health behaviors then elders living in urban areas (Norton and McManus, 1989). Studies on rural elders’ health and health-related problems are needed to guide the practitioners in planning for health care for this group of people. Stress has often been cited as a health problem. High levels of stress have caused negative health outcomes for elders in the United States (Walters, 1993). It has been documented in the literature in the United States that high levels of stress can impair the health and well-being of elders (Phillips and Murrell, 1994). Moreover, evidence in literature in the United States also indicated that stressors experienced by rural elders may differ from those of urban elders. Studies in the United States have also suggested that older people may differ from younger people in either the appraisal of the stressfulness of a problem, the type of stressors reported, or the amount of stress they encounter (Aldwin et al., 1996). Few studies on stress in rural elders have been conducted either in the United States or in Taiwan. Only two studies regarding stress of Taiwanese elderly have been conducted, and these two were conducted in urban areas (Lin, 1998; Lin et al., 1996). Little information is available about strategies and interventions that could be planned and implemented to reduce stress in this group of elders. Identifying stressors and stress levels and the possible relationship that high levels of stress may have on health in rural elders is needed to guide plans of care. Loneliness and depression have been documented in the literature as common and distressing problems for many older people (Blazer, 1990; Lee, 1994). Loneliness and depression are often correlates of multiple losses and numerous life changes which are associated with the aging processes (Cohen-Sachs, 1993; Lee, 1994). In the literature on loneliness of elders in the United States as many as 62% of elderly people were found to be lonely (Johnson et al., 1993). In Taiwan, many elderly are confronted with loneliness and lack of family contact (Fu et al., 1988). However, data on the extent to which loneliness is experienced by rural elders in Taiwan is not available. Increased stress levels have been reported in many studies as affecting mood and psychosomatic symptoms of depression (Phifer and Murrell, 1986; Snow and Crapo, 1982). Clinical researchers and practitioners who focus on elders agree that depression is the most prevalent psychological disorder of the aged in both hospitalized and community settings (Blazer, 1990). Prevalence studies in the United States indicate that as many as 44% of all persons over 65 years of age living in the community show significant depressive symptomatology (Blazer, 1990). The prevalence of depression in Taiwanese rural elders is uncertain. Also, the connection between depression and stress has been supported by the finding of elevated blood cortisone in depressed persons; the cortisone is the main circulating steroid in persons with high levels of stress (Gershon and Rieder, 1993). Most studies of stress in elders in the United States have focused on coping strategies and the relationship of stress to health status as opposed to studies regarding the relationship between stress and depression. No studies about the relationship of these variables in Taiwanese elders were found. This lack of data mandates the need for research on stress, loneliness, and depression in Taiwanese elders so that if these problems are present, interventions can be implemented to alter these problems so as to provide higher quality health care for this group of people. The purpose of this study was to examine the stress level and stressors experienced by Taiwanese rural elders and to determine if differences in the level of stress exist among this group of people on a variety of personenvironmental characteristics. Secondly, the study explored the prevalence of loneliness and depression in this group. The relationship between stress and depression was examined. The psychological stress theories of Lazarus (1993) (cognitive appraisal theory) and Hobfoll (conservation of resources theory) (1989), served as the theoretical foundation for this study. 2. Methods A cross-sectional design utilizing a descriptive-correlational approach was used. Face-to-face administration of study instruments and direct observation of subject’s affect were used. 2.1. Sample and setting Subjects were required to meet the following criteria: (1) 65 years of age or older; (2) non-institutionalized; (3) able to speak in either Mandarin or Taiwanese; (4) capable of verbal communication; and (5) demonstrate no obvious cognitive impairments. The subjects were recruited from two districts located in the southern part of Taiwan. Study sites in two different districts were used to avoid sampling bias. Five villages from the two selected districts, which are rural in nature with a population of 1962–2064 in each village, were randomly selected as study sites. A village is a sub-unit of a district and there are approximately 13 villages in a Taiwanese district. 2.2. Measures The revised Taiwanese Elderly Stress Inventory (RTESI) consists of 62 stressor items was developed by Lin (1998) for use particularly with Taiwanese. Subjects were asked to check whether they have experienced any J.-J. Wang et al. / International Journal of Nursing Studies 38 (2001) 339–347 of the listed stressors since they turned 65 years old. For those items they identified experiencing, they then rate how much stress it caused them. Stressfulness is rated on a seven-point scale ranging from (0) not stressful to (6) extremely stressful. Scoring of the R-TESI is based on the sum of the stressfulness for the stressors that the subject has experienced. It does not include stressors which have not occurred. One item ‘‘taking bus’’, which is not relevant to rural Taiwanese elders, was replaced with ‘‘transportation difficulty’’. Two stressor items, farming and multiple worry/concern about future, were added to the R-TESI. A Cronbach’s alpha coefficient of 0.80 was obtained in this study sample with 62 stressor items, indicating adequate internal consistency. The UCLA Loneliness Scale version 3 (RULS-V3) is a 20-item questionnaire which measures loneliness in adults (Russell, 1996). Each item is rated on a fourpoint Likert-type answer with a range from (1) ‘‘never feel this way’’ to (4) ‘‘often feel this way’’. Scores for the 20 items are summed. Positive worded items are reversed before scoring. The scores may range from 20 to 80 with a higher score indicating a higher level of loneliness. The RULS-V3 has high internal consistency (Cronbach alpha of 0.89 to 0.94). A high test-retest coefficient of 0.73 over 12 months was found (Russell, 1996). A Chinese version of the RULS-V3 was developed for this study and validation was provided by two Taiwanese persons. In this study, the internal consistency of the RULS-V3 was found to be 0.82. The Geriatric Depression Scale short form (GDS-SF) contains 15 items relating to depressed mood and psychophysiological indicators of depression (Sheikh and Yesavage, 1986). Respondents answer each of the statements with (1) ‘‘yes’’ or (0) ‘‘no’’ in relation to how they have felt over the past few weeks. Scores can range from 0 to 15 with a cut-off point of equal to or greater than seven suggesting a large number of depressive symptoms. The GDS has been translated into Chinese and validated with a psychiatric outpatient sample of 461 persons, aged 60 or above (Chan, 1996). Internal consistency reliability was 0.89, and concurrent validity was 0.96. In this study sample, a Cronbach’s alpha coefficient of 0.82 was found. The Apparent Emotion Rating Scale (AER) (Snyder et al., 1998) was used for assessing subjects’ mood and affect as another indicator of depression. The AER is an observation instrument used to rate the presence or absence of three positive affective states (pleasure, interest, and tranquility) and three negative affective states (anger, anxiety, and depression). The observer reviews the indicators for each of the six emotions, circles any of the indicators that were observed, marks ‘‘1’’ for any emotion which has one or more of the indicators circled, and marks ‘‘0’’ for an emotion where no indicators are circled. Scoring of the AER is done by assigning 15 points for each of the positive emotions that 341 is present and 15 points for each of the negative emotions that is absent. Total scores can range from 0 to 90 with a higher score indicating a more positive affective state. A Cronbach’s alpha of 0.56 was found in this study for the indication of internal consistency of AER. 2.3. Procedure Random selection of potential villages was initially made. The investigator obtained a printout of persons aged 65 years and older who resided in the five study sites (villages) and their home addresses and telephone numbers from each district population registration office. The investigator randomly selected several ‘‘Lins’’ in each village (a Lin is a sub-unit of a village). A convenience sample was obtained in each selected Lin. Potential subjects were contacted by telephone to determine if they were willing to learn more about the study. Subjects were recruited through home visits. At the beginning of the visit, the study was explained, questions answered, and the consent form signed if an elder decided to participate. As the investigator interacted with the potential subjects, she determined their cognitive ability based on her assessment skills and knowledge of frail elders. After subjects were recruited, data were collected. 3. Results A total of 201 subjects were enrolled in this study, with a participation rate of 93%. Demographic information is shown in Table 1. The mean age of the subjects was 73.4 (SD=6.36). The majority of the subjects were younger than 75 years old. The sample was quite evenly divided between males and females. A high percentage of the subjects had received no formal education. The majority of the subjects were married or widowed. Over half of the sample reported that their annual income was less than NT50,000 (US1,600). 3.1. Stressors and stress level The most frequent stressors experienced were: ‘‘decrease in individual physical strength or tolerance’’, experienced by 153 subjects (77.2%); ‘‘having grandchildren or great-grandchildren’’, experienced by 115 subjects (58%); ‘‘constant or recurring pain or discomfort’’, experienced by 114 subjects (57.5%); and ‘‘your own hospitalization’’, experienced by 107 subjects (54%). The scores for stressfulness ranged from 0 to 124 (possible score range=0–372) with a mean score of 39.61 (SD=23.7). The following stressors had the highest stressfulness mean scores (range=0–6): ‘‘bad habits such as gambling, alcoholism, drug abuse of 342 J.-J. Wang et al. / International Journal of Nursing Studies 38 (2001) 339–347 3.2. Loneliness and depression Table 1 Summary of demographic characteristics (N=201) Variable N % Domicile Original Taiwanese Mainland Taiwanese 172 29 85.6 14.4 66 57 43 24 7 4 32.8 28.4 21.4 11.9 3.5 2.0 Sex Male Female 86 115 42.8 57.2 Education No education Less than elementary school Elementary school Junior high school High school College graduate 138 12 31 9 7 4 68.7 6.0 15.4 4.5 3.5 2.0 Marital Status Married Widowed Never married Separated Divorced 123 69 5 2 2 61.2 34.3 2.5 1.0 1.0 38 72 86 2 3 18.9 35.8 42.8 1.0 1.5 102 62 20 9 4 1 2 1 51.0 31.0 10.0 4.5 2.0 0.5 1.0 0.5 Age 65–69 70–74 75–79 80–84 85–89 >90 Living arrangement Live alone Live with spouse Live with children Live with grandchildren Live with friends/neighbors Annual income status Less than NT50,000 Between NT50,000 and 100,000 Between NT100,000 and 200,000 Between NT200,000 and 300,000 Between NT300,000 and 400,000 Between NT400,000 and 500,000 More than NT500,00 Missing family members’’ (M=4.74; SD=1.63); ‘‘death of your spouse or children’’ (M=4.69; SD=1.53); ‘‘divorce or separation of your children’’ (M=4.10; SD=1.61); ‘‘unsuccessful in arranging meeting between your child and prospective mate’’ (M=4.07; SD=1.84); ‘‘constant or recurring pain’’ (M=4.07; SD=1.62); ‘‘decrease in individual physical strength or tolerance’’ (#40) (M=4.02; SD=1.83); and ‘‘multiple worry/concerns about future’’ (M=4.02; SD=1.92). The scores for loneliness ranged from 25 to 67 (possible score range=20–80), with a mean score of 43.59 (SD=9.30). Sixty-eight subjects (39.8%) experienced a low level of loneliness; 97 subjects (56.7%) experienced a medium level of loneliness; and only six subjects (3.5%) experienced a high level of loneliness according to the classification proposed by the investigator (low : 20–40; medium : 41–60; high : 61–80). ‘‘There are people who really understand you’’ and ‘‘Feel outgoing and friendly’’ (M=2.66 and 2.60; SD=0.93 and 1.05) (item range=0–4) were the items receiving the highest mean scores while ‘‘Feel in tune with people around you’’ had the lowest mean score (M=1.66; SD=0.80) as compared to the other items. A higher score indicates a higher level of loneliness. The scores on GDS-SF for depressive symptoms ranged from 0 to 15 (possible score range=0–15) with a mean of 7.58 (SD=3.89). One hundred and twelve subjects (55.7%) had scores equal to or greater than 7 (indicating higher level of depression) while 86 subjects (42.8%) had scores below 7 (indicating lower level of depression). The scores of AER for depressive mood status ranged from 45 to 90 (possible score range=0–90) with a mean of 83.50 (SD=13.32). Seventy percent of the subjects (n ¼ 139 ) had scores of 90 and only 6.6% of the subjects (n ¼ 13 ) had scores of 45. Forty-five of the subjects were ranked as having depressive symptoms on the GDS-SF were observed to have a depressed mood (GDS  7) while only three subjects were observed to have a depressed mood in the non-depressed group (GDS-SF57). The agreement between the GDS-SF score and depressive mood in the AER was statistically significant (w2 ¼ 31:74 ; p50:000 ). 3.3. Differences in stressfulness on selected demographic variables Analysis of variance with pairwise comparisons of groups was performed to compare the mean scores of stress level by domicile, age, sex, educational levels, marital status, living arrangement, and SES (socialeconomic status). A statistically significant difference was found between males and females ( p50:001; df ¼ 196 ) with females having higher stress scores. Likewise, a statistically significant difference in stressfulness scores was found among subjects with different educational levels. The mean score of stressfulness for those who received no education was much higher than that found in subjects who had received some education; a significant difference among groups was found (p50:001; df ¼ 2; 195). However, elders who were non-educated had significantly higher mean stressfulness score than those who had received little or greater level of education (see Table 2). 343 J.-J. Wang et al. / International Journal of Nursing Studies 38 (2001) 339–347 Table 2 Comparison of the mean stress level by demographic variables (N=198) Variable N Mean Domicile Original Taiwanese Mainland Taiwanese 169 29 39.73 38.82 65 55 43 35 35.95 43.67 35.53 45.00 Age 65–69 70–74 75–79 >80 Within group ( p-value) Between group ( p-value) 0.849 0.099 0.000c Sex Male Female 84 114 32.78 44.63 Educational level No education Little education 137 42 44.83 27.45 0.000c No education More education 137 19 44.83 28.78 0.007b Little education More education 42 19 27.45 28.78 0.78 120 78 37.62 43.20 38 70 49.13 36.70 0.013a Live alone Live with children and others 38 90 49.13 37.84 0.013a Live with spouse Live with children and others 70 90 36.70 37.84 0.75 Marital status Married Widow, single, separated Living arrangement Live alone Live with spouse 0.000c 0.085 0.020a 0.002b SES Low income Medium income 100 62 46.00 34.67 0.003b Low income High income 100 35 46.00 29.37 0.001c 62 35 34.67 29.37 0.15 Medium income High income a Significant at 0.05 level (2-tailed). Significant at 0.01 level (2-tailed). c Significant at 0.001 level (2-tailed). b A statistically significant difference was found among the stressfulness scores for persons with different living arrangements (p¼ 0:02; df ¼ 2; 195). Elders who lived alone experienced a higher level of stress as compared to elders who lived with their spouse or others (children, grandchildren, friends, or relatives). A significant difference was also found among subjects with different social economic backgrounds ( p50:01; df ¼ 2; 194Þ: The mean stressfulness score in the low income group was significantly higher than the mean stressfulness score in the medium and high income group (see Table 2). 344 J.-J. Wang et al. / International Journal of Nursing Studies 38 (2001) 339–347 3.4. Relationship between stress and depression To determine the relationship among stress, depressive symptoms, and depressed mood, a Pearson correlational analysis was carried out. The relationships of stress level (from the TESI), depressive symptoms (from the GDS-SF), and depressed mood status (from the AER) are shown in Table 3. The stress level was positively associated with depressive symptoms (r ¼ 0:55; p50:001) and negatively associated with depressed mood status (r ¼ 0:46; p50:001 ). Elders with higher levels of stress demonstrated more severe depressive symptoms and had lower mood status scores than elders with lower levels of stress. Depressive symptoms (from the GDS-SF) had a strong negative association with depressed mood status (from the AER) (r ¼ 0:44; p50:001 ). Elders with higher depressive symptom score had lower mood status scores than elders with lower depressive symptom scores. 4. Discussion Findings in this study showed that rural elders experienced a number of different stressors and had a higher level of stress than revealed in Lin’s (1998) study of urban elders. Scores on the loneliness scale indicated that relatively few respondents experienced a high level of loneliness. However, the scores on loneliness were higher than that those reported in community studies conducted in the United States. In addition, scores on the GDS-SF indicated that a majority of the elders had depressive symptoms. Differences in the degree of stressfulness of individuals were found to be related to the demographic variables of gender, educational level, living arrangement, and SES. Findings in this study also showed that there was a high association among stress level, depressive symptoms, and mood status. A decrease in individual physical strength and pain or discomfort were the two stressors most frequently experienced by subjects in this study. These two stressors Table 3 Correlation coefficients among stress, depressive symptoms, and depressed mood Stress level Depressive symptoms 0.546a (195) Depressive mood status  0.463a (195) a Significant at 0.001 level (2-tailed). Depressive symptoms  0.440a (193) experienced by rural Taiwanese elders were similar to those reported by urban Taiwanese elders and by elders in the United States (Backer, 1995; Lin, 1998). The similarity between urban and rural elders confirms that health problems are a common stressor experienced by elders. The presence of functional losses due to progressive chronic disease was supported by findings in a number of studies (Backer, 1995; Johnson et al., 1993); elders with functional deterioration and pain often experience more stressful feelings. The investigator made observations during the interviews. Difficulty in walking due to cardiovascular accidents, fractures, and hearing loss were found in a large number of the subjects in the study. Difficulty in walking may increase stress as a person may feel insecure in moving to reach objects. Some elders expressed a fear of being unable to call for help when a medical emergency would occur. Others were concerned about their lack of financial support as they were unable to work because of physical disabilities. Health professionals need to acknowledge and assess for these problems in rural elders and then plan care that will help them to increase their functional ability. The stressor of having grandchildren and great grandchildren may be specific to participants in this study who were primarily original Taiwanese and between 65 and 75 years old. Fourteen subjects during the interview pointed out that caring for grandchildren is stressful and that they have to take full responsibility for their grandchildren when their adult children are absent from home during the day time. Health professionals can assist elders by providing information about resources such as day care services for their grandchildren or great/grandchildren. The stressors most frequently experienced by the study subjects were not necessarily the ones that caused the highest stress level. It is not surprising that stressors that were deemed most stressful related to family problems such as bad habits of children, death of family members, divorce or separation of children, and lack of success arranging a meeting between a child and a prospective mate. Although it is not the responsibility of elderly people in pairing their children with a prospective mate, the family-oriented nature of the Chinese culture, children are often the primary concern of elders. Many family events become stressors that cause high stress levels for elderly parents. This could be discussed with elders’ family members because then they can have a better recognition and understanding of their aging parents’ concerns. A social support system could be established to compensate for the losses experienced by their parents. Rural Taiwanese elders had a higher level of stress than did urban elders. A mean score of 39.6 on the TESI was found in this study as compared to a mean of 22.4 in a study of urban elders (Lin, 1998). This higher level of J.-J. Wang et al. / International Journal of Nursing Studies 38 (2001) 339–347 stress could possibly be explained by rural elders having a lower educational level, poorer living arrangements, and a lower socioeconomic status than urban elders. Another explanation might be the lack of health and financial support inherent in rural areas as compared to urban areas. Rural areas are more isolated and thus resources are less available. Therefore, educational programs for elders which focus on stress-coping strategies is suggested. Educating Taiwanese rural elders to engage in healthier behaviors to help them achieve and strengthen their roles may help to decrease their feelings of stress. Findings indicated that gender, educational level, living arrangement, and SES were significantly associated with the level of stress experienced by rural elders. The perceived level of stress for females was significantly higher than that reported by males. This finding might be due to the fact that many elderly women in Taiwan are greatly influenced by decisions made by their husbands. When stressful situations are encountered, they may feel they have little control over events in their lives. The illiteracy rate for older Taiwanese women was about four times higher than that of the elderly men (Ministry of the Interior, 1998). Furthermore, the population profile of Taiwanese elders shows that women live longer than men. Thus, it is assumed that women must not only cope with the bereavement of the loss of their spouses but also with the economic decline, loss of independence, and loss of social status that often accompanies widowhood (Hansson and Carpenter, 1990). The other factor which contributed to a higher level of stress was educational level. Health professionals need to take educational level into consideration when assessing for community elders’ stress because a lower educational level can limit elders’ knowledge and ability of problem solving. Home health care services need to be planned to these illiterate elders with tasks such as making phone calls. Lecture materials for educational programs for community elders should consider their knowledge and ability of understanding. Living alone and low income were also found to be associated with higher levels of stress. This finding is consistent with the finding reported by Mumford et al. (1997). Elders who live alone are more likely to lack social and emotional support than elders who live with others. The absence of these resources can contribute to elders being isolated and could contribute to higher stress levels. The changing society in Taiwan is placing greater financial hardships on Taiwanese elders. Thus, financial strain is likely to continue to be a stressor experienced by Taiwanese elders. Government policy regarding financial support for those low income elders are needed. Higher levels of stress were significantly associated with depressive symptoms and a negative mood status. 345 These findings were further validated by the TESI and the AER instruments. When the researcher read the stressor items to the subjects, some cried or appeared sad in mood. When asked, ‘‘Have you experienced a death of your spouse or children since you turned 65?’’ a number of subjects started to cry. Comments included ‘‘I am useless, full of pain in my whole body, I should have been dead for long...,’’ and ‘‘I’m in discomfort all the time; I have no meaning to be alive’’. Facial expressions such as tearfulness, slow responses, and eyes downcast were indications of a depressed mood. The high association between the stress level and level of depression found in this study suggests that there is a need to examine and identify specific interventions which can reduce the stress level of rural elders. It will then be necessary to test and then implement these interventions. Although loneliness was reported by many subjects (60%), only 3.5% of the subjects experienced high level of loneliness. Studies on loneliness of rural elders in the United States reported a range of 35–64% (Holmen et al., 1992; Johnson et al., 1993); in Israel, 33% were lonely (Stessman et al., 1996). Rural people live in small communities tend to be more closely knit than urban people in Taiwan. Thus, even though elders did not have family living close by them, they may have received support from their neighbors. The characteristics of rural communities in Taiwan differ from those in the United States. Rural people live in villages and go out to their rice paddies. This living arrangement facilitates inter-connection among neighbors. This living arrangement may contribute to the differences in the findings in this study and those of rural elders in the United States. The percentage of Taiwanese elders who had depressive symptoms was higher than that reported in studies of American community-dwelling elders. Also, the prevalence of depressive symptoms in this study was higher than that found in previous studies of urban Taiwanese elders. The high prevalence of depression in this study suggests the importance of identifying early symptoms of depression. Because the GDS-SF consists of 15 items and takes less than 5 min to administer, it can be integrated into the routine assessment to screen for depressive symptoms in community elders. To recognize possible depression through observing mood status (such as non-verbal facial expression) is suggested because older people may deny feelings of depression. Findings from the AER were validated by the scores of the GDS in this study. This has clinical implications as community nurses can easily identify depression in elders by administering the AER while conducting routine assessments of their clients. Curricula for community health nurses should contain content about aging, loss, impact of family dynamics, social isolation in rural areas, and mental health problems encountered in elders. Findings from this 346 J.-J. Wang et al. / International Journal of Nursing Studies 38 (2001) 339–347 study also indicate that there is a desperate need for mental health services in rural Taiwan. District local health office should provide mental health care in addition to the routine physical health care to rural elders. Referral may be done for those in severe mental health needs. Current policy does not provide community elders with adequate health care, particularly with mental health services. Therefore, changes in Taiwanese health policy are needed so rural elders can receive mental health services that will help to diminish their stress and promote health. Findings of this study indicate the presence of problems that interfere with health and happiness. Plans are being initiated by the Taiwanese government to push policies that encourage more community-based care in the 21st century (Ministry of the Interior, 1999). Attention needs to be given to ways to provide elderly in their twilight years with happiness and dignity. 4.1. Limitations Aldwin, C.M., Sutton, K.J., Chiara, G., Spiro, A., 1996. Age differences in stress, coping, and appraisal: Findings from the normative aging study. Journal of Gerontology 51B (4), 179–188. Allen, M., 1990. Homebound aging women and the management of stress. Home Healthcare Nurse 8 (4), 30–33. Backer, J., 1995. Perceived stressors of financially secure, community-residing older women. Geriatric Nursing 16 (4), 155–159. Blazer, D., 1990. Emotional Problems in Later Life. Springer, New York. Chan, A.C., 1996. Clinical validation of the geriatric depression scale (GDS). Journal of Ageing and Health 8(2), 238–253. Cohen-Sachs, B., 1993. Coping with the stress of ageingcreatively. Stress Medicine 9, 45–49. Department of Budgets, Accounting & Statistics, 1998. Statistics of Taiwanese Elderly, from 1987 to 1995. ROC. Fu, C., Tseng, L., Lee, C., 1988. The recognition and perspectives of comprehensive insurance plan for public health nurse. Nursing Research 4 (4), 375–386. Gershon, E.S., Rieder, R.O., 1993. Major disorders of mind and brain. In: Readings from Scientific American: Mind and Brain. Freeman and Co, New York, pp. 91–100. Gillanders, W.R., Buss, T.F., 1993. Access to medical care among elderly in rural Northeastern Ohio. Journal of Family Practice 34, 37. Guan, H., 1996. Types of living arrangement and its associated problems for taiwanese elderly. Annual Elderly Education Conference, Taipei. Hansson, R.O., Carpenter, B.N., 1990. Relationship competence and adjustment in older adults: implications for the demands of aging. In: Stephens, M., Crowther, J., Hobfoll, S., Tennenbaum, D. (Eds.), Stress and Coping in Later-life Families. Hemisphere Publishing Corporation, New York, pp. 131–151. Hobfoll, S.E., 1989. Conservation of resources: a new attempt at conceptualizing stress. American Psychologist 44 (3), 513–524. Holmen, K., Ericsson, K., Andersson, L., Winblad, B., 1992. Loneliness among elderly people living in Stockholm: a population study. Journal of Advanced Nursing 17, 43–51. Hsu, R.C., Lin, M., Chou, M., Lin, M., 1997. Medication use characteristics in an ambulatory elderly population in Taiwan. The Annals of Pharmacotherapy 31, 308–313. Johnson, J.E., Waldo, M., Johnson, R.G., 1993. Research considerations: stress and perceived health status in the rural elderly. Journal of Gerontological Nursing 19 (10), 24–29. The researcher acknowledges limitations relevant to methodological and sampling issues. First, subjects in this study were from the southern part of Taiwan and had a lower level of education, a higher rate of living alone or with spouse, and a lower socioeconomic status than had been found in studies of urban Taiwanese elders. Caution needs to be exercised in generalizing the findings of this study to elders living in urban areas and in the northern part of Taiwan or to elders who are institutionalized. Second, the data in this study were gathered at one point in time. The subject’ perception may have been influenced by covariate factors. Thus, the interpretation of the results is constrained by the crosssectional nature of the data. Lastly, problems such as difficulty in understanding the Likert-type scale and tiredness of the subjects were encountered in the administration of the R-TESI and the RULS-V3 instruments. These problems may have affected the scores obtained and thus weaken the validity of the study. Also, the low alpha level (r ¼ 0:56 ) for internal consistency of the AER placed further concern for the reliability of using this instrument. 4.2. Recommendations for future research Based on the results of this study, the recommendations for future research include: (1) the high incidence of stress and depression in rural Taiwanese elders found in this study suggest the need for studies on the efficacy of interventions for treating and preventing stress and depression in older rural community-dwelling elders; (2) further research is needed to explore the dynamics of the person-environment characteristics and how these influence the stress perception of rural elders; (3) subjects in this study had an extremely high prevalence of depressive symptoms. Correlational studies are needed to examine the relationship between depression and quality of life in rural community elders; and (4) difficulties in administering the TESI warrant the development of a shorter instrument to identify stressor and measure the stress level in elders. References J.-J. Wang et al. / International Journal of Nursing Studies 38 (2001) 339–347 Lazarus, R.S., 1993. From psychological stress to the emotions: a history of changing outlooks. Annual Review of Psychology 44, 1–21. Lee, R.N., 1994. Passage from the homeland. Canadian Nurse 90 (9), 27–32. Lin, L., 1998. Revised Taiwanese elderly stressor inventory. Journal of Kaohsiung Medical Science 14 (3), 150–160. Lin, L., Snyder, M., Egan, E., 1996. The development of Taiwanese elderly stressor inventory. International Journal of Nursing Studies 33 (1), 29–36. Ministry of the Interior, 1999. Bureau of Health Statistics Report, ROC. Mumford, D.B., Saeed, K., Ahmad, I., Latif, S., Mubbashar, M.H., 1997. Stress and psychiatric disorder in rural Punjab. British Journal of Psychiatry 170, 473–478. Norton, C.H., McManus, M.A., 1989. Background tables on demographic characteristics, health status, and health services utilization. Health Service Research 23, 725–755. Phifer, J.F., Murrell, S.A., 1986. Etiologic factors in the onset of depressive symptoms in older adults. Journal of Abnormal Psychology 95, 282–297. Phillips, M.A., Murrell, S.A., 1994. Impact of psychological and physical health, stressful events, and social support on subsequent mental health help seeking among older 347 adults. Journal of Consulting and Clinical Psychology 62, 270–275. Russell, D., 1996. UCLA Loneliness Scale (version 3): reliability, validity, and factor structure. Journal of Personality Assessment 66 (1), 20–40. Sheikh, V.I., Yesavage, V.A., 1986. Geriatric Depression Scale (GDS): recent evidence and development of a shorter version. In: Brink, T.L. (Ed.), Clinical Gerontology: A Guide to Assessment and Intervention. Haworth Press, New York, pp. 165–174. Snow, R., Crapo, L., 1982. Emotional bondedness, subjective well-being, and health in elderly medical patients. Journal of Gerontology 37, 609–615. Snyder, M., Ryden, M.B., Shaver, P., Wang, J., Savik, K., Gross, C., 1998. The apparent emotion rating instrument: assessing affect in cognitively impaired elders. Journal of Gerontologist 18 (4), 17–29. Stessman, J., Ginsberg, G., Klein, M., Hammerman-Rozeberg, R., Cohen, A., 1996. Determinants of loneliness in Jerusalem’s 70-year-old population. Israel Journal of Medical Science 32, 639–648. Walters, V., 1993. Stress, anxiety and depression: women’s accounts of their health problems. Social Science Medicine 36 (4), 393–402.