Authors:
Jaafar Behbehania (1),
Nasra M. Shaha (1),
Hanan E. Badr (1)
Makhdoom A. Shahb (2)

(1) Department of Community Medicine and Behavioral Sciences,
Faculty of Medicine,
Kuwait University.

(2) Department of Health Information Administration,
Faculty of Allied Health Sciences,
Kuwait University

Correspondence:
Prof. Nasra M. Shah
Department of Community Medicine and Behavioral Sciences,
Faculty of Medicine, Kuwait University.
P.O. Box 24923 Safat
13110 Kuwait
Tel: (965) 498 6531
Fax: (965) 533 8948
Email: nasra@hsc.edu.kw

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

INTRODUCTION

With a life expectancy of 77 years and 75 years for Kuwaiti females and males in 2005, issues related to healthy aging have become especially important (MoH, 2005). Recognition of physical as well as psychological health deficiencies among older persons deserves a high priority in health planning since both are important ingredients of the overall quality of life. The occurrence and persistence of depressive symptoms among older persons may be indicative of long term psychological ailments that may also affect their physical health and functional ability. Depressive symptoms among older adults have been reported to have various health consequences. Such symptoms are positively associated with chronic illnesses and other medical conditions (Okwumabua, Baker, Wong & Pilgram, 1997; Baker, Okwumabua, Philipose, & Wong, 1996; Al-Shammari & Al-Subaie, 1999) increased risk of dementia (Fuhrer, Dufouil, & Dartigues, 2003), higher physical performance decline (Penninx, Guralnik, Ferrucci, Simonsick, Deeg, & Wallace, 1998), and increased mortality level (Schulz, Beach, Ives, Martire, Ariyo & Kop, 2000).

Co-residence with children and other relatives has been found to have strong negative associations with the prevalence of depressive symptoms in various developed as well as developing countries (Chen, Wei, Hu, Qin, Copeland, & Hemingway, 2006; Zunzunegui, Beland, & Otero, 2001; Harris, Cook, Victor, Rink, Mann, Shah et al., 2003). However, co-residence patterns are undergoing major changes. Kuwait has traditionally been a country with very strong family ties with rich networks of social support. Co-residence has declined in consort with rapid socio-economic development that the country has experienced during the last 3-4 decades. Once married, the children have an increasing preference for setting up their own nuclear units, resulting in a larger percentage of older persons residing without any co-habiting children. Comparison of data from the 2005/6 survey on which the present paper is based, with a previous household survey held in 1999, shows that the percentage of older persons age 50+ co-habiting with at least one child, declined from 92 % to 80 % in less than one decade.

Relatively little is known about the impact of the above changes on the psychological health of older Kuwaitis. The objective of this paper is to present a description of the self-reported depressive symptoms among Kuwaiti nationals aged 50+ and assess the role of family structure in such perceptions. The analysis controls for socio-demographic background, and experience of chronic illness, in order to gauge the net impact of co-residence. Kuwaiti men and women are compared in order to examine the differences in self-reported depressive symptoms and their correlates between the two sexes.

METHODS

Data for this paper were collected through a cross-sectional survey of Kuwaiti households conducted during April, 2005 to December, 2006 where 2,487 men and women aged 50 and over were interviewed. Non-Kuwaitis were not included in the study. The study was approved by the Ethics Committee of the Faculty of Medicine. Verbal consent was obtained from each respondent before the interview.

Our survey was based on two of Kuwait's six geographical regions, or governorates. One governorate (Capital) represented a relatively more, and the other (Ahmadi), a relatively less urban area. Within each governorate residential areas were randomly selected where all Kuwaiti households were approached for inclusion in the survey if they agreed and had a resident aged 50 or over. A sample representing the proportion of older persons aged 50+ in each of the governorates was chosen. About 800 persons were selected from each of the three following age groups: 50-59, 60-69, and 70+. All older men and women who agreed to participate in the study were interviewed. In January 2005, these two governorates had 41,205 Kuwaiti persons aged 50+. Our survey covered 6% of them.

Of the total households with at least one older person aged 50+, 75.3 % agreed to participate. These households had 2,605 potentially eligible persons, 2,487 (95 %) of whom participated. A total of 1,451 individuals from the Capital and 1,036 from Ahmadi governorates were successfully interviewed. A proxy respondent, usually a close relative, was used in 5.4% of the cases where the respondent was not able to answer, primarily due to old age or disability. A questionnaire was developed in English and then translated into Arabic. Trained male and female Arab interviewers collected the data on psychological health, physical health, socio-demographic background and several other aspects.
The present paper focuses on psychological health as measured by a scale of depressive symptom experience adapted from the Mexican Health and Aging Study (Soldo, Wong, & Palloni, 2003). This scale consists of 10 items measured on a Likert scale ranging from "Never" to "Most of the time", shown in Table 2. A reference period of 7 days prior to the interview was used. For example, a question was asked on whether the respondent had poor appetite during the previous 7 days, and whether this occurred most of the time, sometimes, seldom or never. A weight of 3 was given to the item if it occurred most of the time and a weight of 0 was given if it never occurred. The ten items were summed to arrive at a composite score of depressive symptom experience, with a range between 0 and 30. Comparisons between males and females were made by calculating the mean score of depressive symptoms by sex according to major socio-demographic characteristics. Associations between explanatory variables and depressive symptoms were tested by using ANOVA. Multivariate analysis using binary logistic regression was then conducted in order to assess the factors that significantly predicted the prevalence of higher than median depressive symptom score. The p values of <0.05 were considered to be statistically significant.

The association of co-residence with depressive symptoms was examined by using three variables, namely, the type of respondent's family (nuclear vs. extended), the number of children who were living with the respondent and the number of children living away from him/her. In order to control for socio-demographic background the following variables were included: age, marital status, ethnic background (Bedouin vs. non-Bedouin, the former representing relatively more traditional socio-demographic characteristics and attitudes), years of schooling, work status, and per capita family monthly income. Also, the presence of chronic illnesses was assessed in terms of whether the respondent had been diagnosed to have hypertension, diabetes or heart disease. A variable indicating the intensity of illnesses was developed in terms of respondents who reported none of the above illnesses, or reported 1, 2, or all 3 of these illnesses.

RESULTS

A profile of the socio-demographic characteristics of the 2,487 Kuwaiti older persons aged 50+ is shown in Table 1. About 61% of the sample comprised women. Men were significantly older than women (Mean= 66.3 and 62.3 years, respectively, p < .001). The percentage of widowed persons was significantly higher among women than men (32.5% and 3.6%). About 45% of men as well as women belonged to a Bedouin family. About 14% of men and 18% of women had less than 5 years of education, while almost a quarter of each had beyond high school education. Men and women differed very significantly in terms of their work history. About 84% of the women had never worked while 87% of the men had worked earlier but were now retired. The distribution of per capita monthly income was similar for men and women.

A larger percentage of men (72%) were residing in nuclear families (defined as a family where the husband and/or wife is living alone or with unmarried children) compared with women (63%). Kuwait is a relatively high fertility society and co-residence with children is common. We found that none of the children were living with the respondent among 19% of men and 20% of women, while 3 or more children were living with the respondent among 47% of men and 41% of women. About 70% of men as well as women were suffering from at least one chronic illness, namely hypertension, diabetes or heart disease, while 14% of men and 10% of women were suffering from all three of these diseases.
(Table 1).

Table 2. Respondents' self-reported depressive symptoms during the 7 days preceding the survey among older Kuwaiti males and females (% is row wise)
b These items were phrased in a positive manner and the percentages reflect a negative response to the given item. For example, the question asked whether the respondent felt happy.
*p value indicates the differences between males and females.

The ten items used as indicators of depressive symptoms are shown in Table 2. About 21% of men and 23% of women reported that during the previous week they did not enjoy doing anything most of the time. Other items that 7-12% of the respondents reported experiencing most of the time consisted of having poor appetite, restless sleep, feeling unhappy, and not feeling proud of children's accomplishment. A consistently larger percentage of women expressed the experience of depressive symptoms compared with men for most items. The differences between the two sexes were statistically significant for seven of the ten items. The mean score on depressive symptoms was significantly higher for women than men (11.2 and 10.6; p<.001), as shown in Table 3.

Table 3. Mean and Standard Deviation (SD) of depressive symptoms score according to major socio-demographic characteristics of older Kuwaiti males and females (n=2487)
***p <0.001, **p<0.01, *p<0.05

The depressive symptoms total score was divided according to percentiles into three categories, mild (<50th percentile), moderate (50th-75th percentiles) and severe (>75th percentile). Severe depressive symptoms were prevalent among 16.6% of the sample (18.1% of the females and 14.2% of the males, p<0.05). Moderate symptoms were reported among 38.2% of the older men and women with almost equal gender distribution (data not shown).

Bivariate analysis showed a strong inverse association between co-residence and depressive symptoms. Men as well as women who had children living with them had a lower depressive symptom score (Table 3). For example, women who had 3 or more children living with them had a mean score of 10.1 compared with 12.7 among those who had no children living with them (p< 0.001). On the other hand, the association between the number of children living away from the house and the depressive symptom score was generally positive for men (p< 0.001), as well as women (p< 0.01). The presence and intensity of chronic illnesses had a strong positive association with depressive symptoms (Table 3). Among women with no illness, for example, the mean score was 9.9 compared with 13.3 among those who were suffering from all three chronic illnesses (p< 0.001 for each sex).

The bivariate association of depressive symptom mean scores with socio-demographic and other predictors showed a very similar pattern for males and females, even though the score was consistently higher for females. The mean score of depressive symptoms increased significantly by age among both sexes (p< 0.001). Marital status did not show a notable difference in the depressive symptoms score. Bedouins had significantly higher scores than non-Bedouins among men (p< 0.01) as well as women (p< 0.05). There was a linear, inverse association between educational level and depressive symptom score among men as well as women (p< 0.001). Those who were currently employed had significantly lower scores than those who had never worked or were retired, among both sexes (p< 0.001). Those with relatively higher per capita monthly income reported significantly higher mean scores on depressive symptoms within each sex. Those living in extended families generally had lower scores than those living in nuclear families.

Table 4. Stepwise binary logistic regression of significant predictors of depressive symptoms among older Kuwaiti males and females (n=2487)
(a) RG: Reference Group


Logistic regression analysis for the total sample as well as men and women is shown in Table 4. Respondents at or below the median score were coded as 0 (54%) and those above the median (46%) were coded as 1. All ten variables shown in Table 3 were included as predictors. Also, sex was included as a variable in the analysis for the combined sample. We found that in the total sample, sex did not appear as a significant variable, indicating that men were not significantly different from women once the socio-demographic and other characteristics were controlled.

In terms of the associations of various predictors with higher depressive symptoms, the findings were very similar for the two sexes. The presence and intensity of chronic illnesses was the most important predictor in case of men and the second most important one in case of women. For example, among males, those with two chronic illnesses were 2.6 (p< .001) times more likely and the ones with all three illnesses were 3.6 (p < .001) times more likely to report higher depressive symptoms compared with those who had no illnesses.

All three variables that measured co-residence patterns were significantly associated with higher depressive symptoms. Among men, those who had no children living with them were 2.2 times more likely to report higher depressive symptoms than those who had 3 or more children living with them (p < .001). A similar association was present for women, among whom presence of children in the house was the most important variable. Men as well as women who had larger numbers of children living away from the house were significantly more likely to report higher depressive symptoms than those who had no children living away. Among men, those living in a nuclear family were almost twice as likely to report higher depressive symptoms than those living in an extended family with married and unmarried children as well as other relatives (p < .05). Among women, the association with type of family was different than men; those living only with married and unmarried children reported a significantly lower risk (odds ratio = 0.66) of higher depressive symptoms than those living with their children plus other relatives.

Of the six predictors included to measure the socio-demographic background of respondents, educational level was significant among both sexes and work status was significant among men. Educational level had a linear, inverse association with higher depressive symptoms. For example, compared to women with above secondary level education, the odds of reported depressive symptoms were 1.9 times higher among those with 9-12 grades of education (p < .05), 2.3 times higher among those with 5-8 grades of education (p , < .01) and 2.8 times higher among those with 0-4 grades of education (p < .001). A roughly similar pattern was reported by men. Current employment appeared to reduce the odds of higher depressive symptoms among men while this variable did not appear as a significant one for women. Men who were currently employed were about one-third less likely to report depressive symptoms than the retired men (p < .01).

DISCUSSION

Our study on 2,487 Kuwaitis aged 50+ showed that on a ten item scale of depressive symptoms ranging from 0-30 the mean and median scores of respondents were 10.97 and 11.0, respectively. A previous study conducted in Kuwait indicated a prevalence rate of 9.1% among persons aged 45 years and above (Al-Otaibi, Al Weqayyan, Taher, Sarkhou, Gloom, Aseeri et al., 2007). Women had a significantly higher mean score than men, 11.2 and 10.6, respectively. A higher level of depressive symptoms among women than men has been reported in several previous studies (Al-Shammari & Subaie, 1999; Fuhrer, Dufouil, & Dartigues, 2003; Zunzunegui, Beland, Llacer, & Leon, 1998; Minicuci, Maggi, Pavan, Enzi, & Crepaldi, 2002).

After controlling for the presence of chronic illness and socio-demographic background, co-residence emerged as a very significant factor in the multivariate analysis. Absence of children in the house was positively associated with depressive symptoms in our study. The odds of depressive symptoms among men as well as women were 2.2 times higher among those with no children in the house compared with those who had 3 or more co-resident children. In Kuwaiti society, children play a pivotal role in the lives of families, especially women. In a national survey of married women in reproductive ages in 1999, it was found that the desired number of children was 5.5 per woman. Children are perceived as a form of social, economic and political capital that help an individual to enlarge and extend a family's network (Shah & Nathanson, 2004). The number of children borne by Kuwaiti women has declined from almost 7 in 1965 to 4.1 in 2005 (MoH, 2005). However, Kuwaiti fertility is much higher than world fertility (2.7), as well as the fertility of developing countries (2.9) (PRB, 2007). Thus, the country is pro-fertility and encourages the growth of the national population. For older persons, it appears that the presence of co-resident children as well as their larger number act as factors that protect against depressive symptoms. It is culturally expected that children would take care of parents and co-residing with them is one way of fulfilling that expectation.

We presume that co-residence with a larger number of children provides a great deal of social support and personal satisfaction. Consistent with our findings, several previous studies have reported an inverse association between social support and depressive symptom experience among older persons in diverse locations such as rural China (Chen, Wei, Hu, Qin, Copeland & Hemingway, 2006), Spain (Zunzunegui, Beland & Otero, 2001), and London (Harris, Cook, Victor, Rink, Mann, Shah et al., 2003). Social isolation, loss of close social contacts and low emotional support from children were found to be potential risk factors for the onset of depressive symptoms and depressive episodes in older people (Brilman & Ormel, 2001; Zunzunegui, Beland, Llacer & Leon, 1998). Furthermore, the size of the social network was found to be inversely associated with depressive symptoms in community-dwelling persons in urban Canada (St John, Blandford & Strain, 2006), and men and women aged 65+ in an American community (Palinkas, Wingard & Barrettconnor, 1990).

In addition to family structure, experience of chronic illnesses (hypertension, diabetes, and heart disease) appeared as a very important variable among both sexes. Those with all three illnesses were 3.6 times more likely among men and 2.1 times more likely among women to report above median depressive symptoms compared with their counterparts without any illness. A similar association was reported among Saudi elderly where depression was associated positively with the number of medical diagnoses and medications received (Al-Shammari & Al-Subaie, 1999). Among older African Americans, depression symptoms were significantly higher among those with six or more chronic illnesses (Okwumabua, Baker, Wong & Pilgram, 1997). Similar findings were reported among Taiwanese older residents at nursing homes (Lin, Wang & Huang, 2007), community dwelling older adults (Mojtabai & Olfson, 2004) and among Canadian older population (Ostbye, Steenhuis, Walton & Cairney, 2000). The presence of multiple chronic illnesses is likely to seriously affect functional ability, interfere with sleep, and have a negative impact on self-perception of well being, resulting in the increase of depressive symptoms.

Similar to our results, an inverse association between educational level and depressive symptoms was also reported among older community residing Taiwanese (Wang, 2001) and older Finnish persons (Pahkala, Kesti, Kongassaviaro, Laippala & Kivela, 1995). The risk of mental disorders decreased with increasing educational level in an Iranian study (Noorbala, Yazdi, Yasamy & Mohammad, 2004). Higher educational level may affect depressive symptom experience through several pathways. It may enable an older person to appreciate and develop a healthier lifestyle. It may also provide greater self confidence and ability to control one's health. A similar situation may exist in case of employed men who were found in our study to be about one-third less likely to report high depressive symptoms compared to the retired men. Our study has two important implications, one, regarding the protective role of social networks and the other regarding the positive association between chronic illness and depressive symptoms. The presence of social networks, as well as their larger size, seems to play a very important role in acting as a buffer against depressive symptoms in older persons. However, co-residence patterns are changing and a larger percentage of older persons now live without any children as they did about a decade ago. This change seems to have had some negative implications for the emotional health of older persons, as judged from the presence of depressive symptoms in our study. While health planners cannot intervene in residential arrangements of families, health care providers should be sensitive to the type of social network support available to the older patients under their care, especially in cases where depressive symptoms as well as chronic illnesses are especially high.

About 70% of the older respondents in our study were suffering from at least one of the three specified chronic illnesses, while 12% had all three diseases. Data from the same survey used for this paper also show a marked increase in chronic illnesses in the country compared to the past (Shah, Behbehani & Shah, 2008). The increasing prevalence of chronic illnesses poses a serious concern not only for the physical but also the emotional health of older persons and needs to be addressed on a priority basis by health planners.

One of the limitations of the present study is its cross-sectional nature that does not allow an assessment of the causal direction of association between chronic illnesses and depressive symptoms, or between changes in co-residence patterns and depressive symptoms. Longitudinal studies could contribute greatly to a better understanding of the above relationships.

Finally, the two implications identified above may form the basis for the following interventions by the Ministry of Health. Firstly, efforts at reducing the incidence of chronic illnesses may be intensified through encouraging lifestyle changes such as weight control, exercise and improved dietary habits. Secondly, health care workers may be provided training to identify older persons at a higher risk of developing depressive symptoms by asking specific questions about residential patterns and the availability of social support. Programs to provide social support for those living alone may then be devised. The above may help not only in improving the emotional health of older persons but may also reduce the impact of depressive symptoms as a risk factor for chronic illnesses.

REFERENCES

Al Otaibi, B., Al Weqayyan, A., Taher, H., Sarkhou, E., Gloom, A., Aseeri, F., Al Mousa, E., Al Zoubi, H., & Habeeba, S. (2007). Depressive symptoms among Kuwaiti population attending primary healthcare setting: Prevalence and influence of sociodemographic factors. Medical Principles and Practice, 16, 384-388.
Al Shammari, S. A. & Al Subaie, A. (1999). Prevalence and correlates of depression among Saudi elderly. International Journal of Geriatric Psychiatry, 14, 739-747.
Baker, F. M., Okwumabua, J., Philipose, V., & Wong, S. (1996). Screening African-American elderly for the presence of depressive symptoms: A preliminary investigation. Journal of Geriatric Psychiatry and Neurology, 9, 127-132.
Brilman, E. I. & Ormel, J. (2001). Life events, difficulties and onset of depressive episodes in later life. Psychological Medicine, 31, 859-869.
Chen, R., Wei, L., Hu, Z., Qin, X., Copeland, J. R., & Hemingway, H. (2005). Depression in older people in rural China. Archives of International Medicine, 165(17), 2019-2025.
Fuhrer, R., Dufouil, C., & Dartigues, J. F. (2003). Exploring sex differences in the relationship between depressive symptoms and dementia incidence: Prospective results from the PAQUID study. Journal of the American Geriatrics Society, 51, 1055-1063.
Harris, T., Cook, D. G., Victor, C., Rink, E., Mann, A. H., Shah, S., DeWilde, S., & Beighton, C. (2003). Predictors of depressive symptoms in older people - a survey of two general practice populations. Age and Ageing, 32, 510-518.
Lin, P. C., Wang, H. H., & Huang, H. T. (2007). Depressive symptoms among older residents at nursing homes in Taiwan. Journal of Clinical Nursing, 16, 1719-1725.
Minicuci, N., Maggi, S., Pavan, M., Enzi, G., & Crepaldi, G. (2002). Prevalence rate and correlates of depressive symptoms in older individuals: The Veneto study. Journals of Gerontology Series A-Biological Sciences and Medical Sciences, 57, M155-M161.
Ministry of Health (MoH) (2005). Health Kuwait. Department of Statistics. Kuwait.
Mojtabai, R. & Olfson, M. (2004). Major depression in community-dwelling middle-aged and older adults: prevalence and 2-and 4-year follow-up symptoms. Psychological Medicine, 34, 623-634.
Noorbala, A. A., Yazdi, S. A. B., Yasamy, M. T., & Mohammad, K. (2004). Mental health survey of the adult population in Iran. British Journal of Psychiatry, 184, 70-73.
Okwumabua, J. O., Baker, F. M., Wong, S. P., & Pilgram, B. O. (1997). Characteristics of depressive symptoms in elderly urban and rural African Americans. Journals of Gerontology Series A-Biological Sciences and Medical Sciences, 52, M241-M246.
Ostbye, T., Steenhuis, R., Walton, R., & Cairney, J. (2000). Correlates of dysphoria in Canadian seniors: The Canadian study of health and aging. Canadian Journal of Public Health-Revue Canadienne de Sante Publique, 91, 313-317.
Ostir, G. V., Markides, K. S., Black, S. A., & Goodwin, J. S. (2000). Emotional well-being predicts subsequent functional independence and survival. Journal of the American Geriatrics Society, 48, 473-478.
Pahkala, K., Kesti, E., Kongassaviaro, P., Laippala, P., & Kivela, S. L. (1995). Prevalence of Depression in An Aged Population in Finland. Social Psychiatry and Psychiatric Epidemiology, 30, 99-106.
Palinkas, L. A., Wingard, D. L., & Barrettconnor, E. (1990). The Biocultural Context of Social Networks and Depression Among the Elderly. Social Science & Medicine, 30, 441-447.
Penninx, B. W., Guralnik, J. M., Ferrucci, L., Simonsick, E. M., Deeg, D. J., & Wallace, R. B. (1998). Depressive symptoms and physical decline in community-dwelling older persons. The Journal of American Medical Association, 279(21), 1720-1726.
Population Reference Bureau (PRB). (2007). 2007 World Population Data Sheet. Washington DC.
Shah, N., Behbehani, J., & Shah, M. (2008). Prevalence and correlates of major chronic illness among older Kuwaiti nationals. Presented at the annual meeting of Population Association of America, New Orleans, USA.
Schulz, R., Beach, S.R., Ives, D.G., Martire, L.M., Ariyo, A.A., & Kop, W.J. (2000). Association between depression and mortality in older adults: the cardiovascular health study. Archives of Internal Medicine, 160(12), 1731-1732.
Soldo, B. J., Wong, R., & Palloni, A. (2003). Mexican Health and Aging Study.
St John, P. D., Blandford, A. A., & Strain, L. A. (2006). Depressive symptoms among older adults in urban and rural areas. International Journal of Geriatric Psychiatry, 21(12), 1175-1180.
Wang, J. J. (2001). Prevalence and correlates of depressive symptoms in the elderly of rural communities in southern Taiwan. Journal of Nursing Research, 9(3), 1-12.
Zunzunegui, M. V., Beland, F., Llacer, A., & Leon, V. (1998). Gender differences in depressive symptoms among Spanish elderly. Social Psychiatry and Psychiatric Epidemiology, 33, 195-205.
Zunzunegui, M. V., Beland, F., & Otero, A. (2001). Support from children, living arrangements, self-rated health and depressive symptoms of older people in Spain. American Journal of Epidemiology, 153, S22.