Influence of Demographic Characteristics on Subjective Well-Being of Older Adults

Research into well-being of older adults is at advanced stage the world over. Most of the studies are conducted in the west. A few studies done in Kenya have focused on older adults’ abuse and vulnerability. Furthermore, these studies have used younger populations, thus lack self-reporting by the older adults themselves. This causal-comparative study sought to determine the level of subjective well-being of older adults and how demographic data related to the population and within groups influence the subjective well-being of older adults. Data was collected from older persons (n=140, >65yrs) participating in the Older Person’s Cash Transfer programme that serves the non-pensionable and aged Kenyans. Findings revealed that the older adults experienced low levels of subjective well-being, low levels of positive affect and low levels of negative affect and were dissatisfied with their life. The findings further revealed that being married, having own source of income, attaining secondary school education and poor self-perceived health, significantly influenced subjective well-being. Relative absence of negative affect strongly predicted subjective well-being, followed by presence of positive affect and finally the dimension of satisfaction with life. The demographic characteristics did not reveal themselves as predictive variables in this study. Understanding the dynamics, emotional and cognitive processes of older adults may be useful in designing interventions, strategies and policy programs that could enhance subjective well-being of older adults.


Introduction
The 21 st century older adults are living longer lives than the previous generations (Population Pyramid of the World, 2015a). It is estimated that 80% of older persons in the world will be living in low-income and middle-income countries by the year 2050 (United Nations Population Division, 2013). It is envisioned that these older persons will be faced with more concerns and worries about ageing (Pew Research Center Survey, 2014). This demographic gains towards the later years of life present challenges that necessitate healthy psychosocial and economic improvement. These statistical assessment is professionally supported by Grundy (2006) and Buki (2014) observations that older people's capacity to live a longer life of quality is a pressing issue. Moreover, Heppner, Casas, Carter and Stone (2000 p. 7) proposed that ageing concerns should go beyond biological changes; to consider the contextual factors that influence adaptive subjective well-being across the world. In context, subjective well-being represents life experiences that translate into personal happiness within the general normal daily life tasks and activities (Ryff, 2014). The emphasis is on healthy psychological experiences within the dimensions of satisfaction with life, presence of positive affect and relative absence of negative affect. The satisfaction with life dimension denotes the cognitive appraisal of specific domains in a person's life over a relatively long period of recall. The emotional appraisals reflect on an individual feelings experienced within a shortterm frame that are relatively less stable. In this regard, well-being is a self-report measure of both emotional and cognitive appraisals relating to the there-and-then and here-and-now experiences respectively. According to Suzuki, Fujii, Gärling, Ettema, Olsson and Friman (2013) frequency of feelings is strongly related to global well-being measure than the emotional intensity. In line with the theory of Reasoned Action (Fishbein, & Ajzen, 1975) and the Theory of Planned Behaviour (Ajzen, & Fishbein, 1980) adopted for this study, subjective emotional experience serves as a signal that helps an older adult to engage in adaptive voluntary behaviour once the initial involuntary emotional surge has passed. Individuals high in negative affect exhibit on average low levels of subjective well-being and higher levels of life dissatisfaction. Positive affect is linked with an increase in longevity, a decrease in stress and a high subjective well-being (Paterson, Yeung, & Thornton, 2015). Forgas (2013) posits that experiencing negative affect is a normal part of life and human nature. Further evidence shows that cognitive and affective component have distinct findings and that one component may not be a reflective of the other component.
Studies into subjective well-being seeks to underscores the importance of subjectivity in assessing what makes life good and desirable according to an individualized criterion in relation to self, others and the environment (Diener, 2013). Past studies have indicated that in America, where the population are less concern with the growing older population (Pew Research Center Survey, 2014), that despite experiencing late-life disability, adequate health care and psychosocial support promotes successful ageing among the older adults (Romo, Wallhagen, Yourman, Yeung, Eng, Micco, Perez-Stable & Smith, 2012). Other studies by Trigg, Watts, Jones, Tod and Elliman (2012) revealed lower subjective well-being among the older adults with dementia while a study by Steptoe, Deaton and Stone (2015) showed that older adults experience varying levels of well-being. In Asia studies by Lu, Kao and Hsieh (2010) and Sargent-Cox, Anstey and Luszcz (2012) indicated higher levels of subjective well-being while others revealed a medium level of subjective (Suh, Choi, Lee, Cha, & Jo, 2012). Studies on well-being and ageing process in African context are scanty. In Sub-Saharan Africa, a study by Aboderin, (2010) observe that the increasing population poses a major concern due to inadequate policy approaches towards understanding the ageing process in the region. The observations of Aboderin are echoed by Mwanyangala, Mayombana, Urassa, Charles, Mahutanga, Abdullah and Nathan (2010) who revealed a low subjective well-being and poor health status among the older population in Tanzanian. Mwanyangala et al adds that having good health status was significantly associated with being male, married and not being among the oldest old, a high level of education and higher level of socioeconomic status of the household. Using multivariate analysis two studies by Calys-Tagoe, Hewlett, Dako-Gyeke, Yawson, Baddoo, Seneadza, Mensah, Minicuci, Naidoo, Chatterji, Kowal and Biritwum, (2014) in Ghana and Phaswana-Mafuya, Peltzer, Chirinda, Kose, Hoosain, Ramlagan, Tabane and Davids (2013) in South Africa posits that being of a younger old age, being of male sex, having a high educational level and high income were associated with high levels of subjective well-being. The findings are similar with the studies conducted in Europe and Asia that well-being decreased with increasing age and that women reported poor subjective well-being.
In Kenya, people aged above 65 years accounted for 2.8% in 2015 and is projected to increase to 6.1% by 2050 (Population Pyramid of world, 2015b). As it is elsewhere across the globe, Kenyan older people form an increasingly important sub-group that requires adequate health care, psychosocial and economical support in regard to aging process. A review by Walaba (2014) found out that ageing Kenyans with positive experiences healthily lived beyond 100 years and still were useful to the society. In contrast a study conducted in a rural population in Western Kenya by Kabole, Kioli and Onkware, (2013) revealed that 63% of the older adults do not experience goodness in old age. Another study in Nairobi slums by Kyobutungi, Egondi and Ezeh (2010) found out that different groups of persons have different level of subjective well-being. In addition, a survey conducted by Kenya National Commission on Human Right (2009) andPew Research Center Survey (2014) show that many Kenyans have fears and anxieties about growing old. The reviewed literature provides evidence that link good governance with the development of strategies that enhance subjective well-being of older adults. Therefore, this study sought to gain an understanding into the influence of demographic characteristics on subjective well-being of older adults participating in Older Persons Cash Transfer Program in Kenya.
Both Intrapersonal and interpersonal experiences have been found to have a predictive influence on subjective well-being of older adults. This is demonstrated by an integrative model by Galinha and Pais-Ribeiro (2011) that the predictors of subjective well-being were exclusively intrapersonal; satisfaction with life, negative state affect and positive state affect. As is the case with the current study, this result underpins the importance of using the selfreporting method on ageing process. Another study by Siedlecki, Salthouse, Oishi and Jeswani (2014) examined the relationship among types of social support and facets of subjective well-being found out that there were no significant differences in predictors of subjective well-being across age. The results of this studies emphasize that predictors of subjective well-being are dynamic; its depended on factors such as demographic characteristics of the sample, variables being assessed, time frame and prevailing environmental situations. Similarly, a cross-sectional and correlation study by Suh, Choi, Lee, Cha and Jo (2012) among older Korean adults showed that older age and lower economic status reduced life satisfaction and that being female, having a monthly income, living with a spouse, having knowledge about aging were associated with enhanced wellbeing. They concluded that living with a spouse was among the most powerful predictors of well-being, followed by perceived health status.
In this study, subjective well-being was considered alongside older adult's demographic characteristics of gender, age, living arrangements, income, sources of income, employment and self-perceived health. Gender has been found to influence the subjective well-being of older adults. Despite women having higher life expectancy than men, older men reported better health than their female counterparts in an Indonesian sample (Ng, Hakimi, Byass, Wilopo, & Wall, 2010), in Nairobi slums, Kenya (Kyobutungi et al., 2010), in rural South Africa community (Phaswana-Mafuya et al., 2013) and in a Brazilian sample (Cachioni et al., 2017). On the contrary, a study by Mehmet and Yardan (2012) found out that gender did not affect quality of life among Turkish older adults living in nursing homes. Age, marital status, education and socio-economic status also influence subjective well-being of older adults. An earlier study by Yang (2008) reports marital status significantly influenced subjective wellbeing of older adults. Studies indicate that participants in older age groups, those not in any marital relationship and low educational and low socioeconomic status indicate low subjective well-being (Ng, et al, 2010;Gomez-Olive, Thorogood, Clark, Kahn, & Tollman, 2010;Kyobutungi, et al, 2010). The researchers conclude that the declining health with increasing age is likely to increase demand for health care and other services as people grow older. Therefore, understanding the determinants of healthy ageing is essential in targeting health-promotion programmes. Other studies have given an indication that older adults are relatively satisfied with their ageing until relatively late in life when taking into account the factors associated with increased age such as poor health and widowhood, which explained for poorer subjective well-being in older cohorts (Cachioni, et al, 2017;Walaba 2014)). The reviewed literature gives dissimilar results regarding the influence of age on subjective wellbeing.
Income has been found to influence subjective well-being of older adults. An earlier study by Pinquart and Sörensen (2000) and Lee (2010) indicated that income was correlated more strongly with well-being. Another study by Ingrand, I. Paccalin, Liuu, Gil, and Ingrand, P. (2018) adds that perception of personal financial situation has a direct influence on quality of life. In tandem with this study, Dai, Zhang, and Li (2013) indicated that health, economic status and family relations had a direct influence on subjective well-being among the Chinese living in major cities. Paid employment is critical to the well-being of individuals; it provides a direct access to resources therefore, fostering satisfaction (Warr, 2003). Living arrangement is one of other factors that has an influence on subjective well-being. Studies in Vietnam reveal that having quality contact with adult children (Pinquart & Sörensen, 2000) and intergenerational co-residence (Yamada & Teerawichitchaian, 2015) significantly increased subjective well-being of older adults. Similar results by Reichstadt, Sengupta, Depp, Palinkas and Jeste (2010) indicated that older adults who maintained a social support system had a positive well-being in San Diego, California. Another study by Bryant, Bei, Gilson, Komiti, Jackson and Judd (2012) found out that relationship status was also significantly associated with satisfaction with life among community dwelling Austrian older adults. The above studies were conducted in community with well-known living arrangement. However, the living arrangement in the current study area is not well documented. Earlier studies have theorized that subjective well-being is shaped within the context of time and space (Nordbakke & Schwanen, 2013;Chung and Lee, 2011). The studies found out that different and multiple demographic characteristics influenced subjective well-being of older adults at different times in various parts of the world.
The reviewed studies were carried out in the developed world, with a few studies done in developing countries. If the current trend in ageing population continues, then more and more Kenyans are expected to live longer. Therefore, subjective well-being is an important concept that require an understanding within the Kenyan context.

Current Study
This causal-comparative study sought to determine the level of subjective well-being of older adults and how selected demographic data related to the population and within groups influenced subjective well-being of older adults. It also sought to determine whether the dimensions of subjective well-being and the selected demographic variable had a predictive power on subjective well-being.

Participants and Procedure
The participants were 140 older adults enrolled into the Older Persons Cash Transfer programme that provides about 18 USD to poor households aged 65 years and above (GOK, 2009;GOK, 2012). The programme seeks to improve subjective well-being of nonpensionable older Kenyans, who were not formally employed as envisioned under the Social Protection Policy (GOK, 2006). This study was carried in June, 2018 in Kajulu ward, Kisumu East sub-county, Kisumu County, Kenya. Procedures were approved by Maseno University School of Graduate Studies. The research permit number 15184 was granted by the Kenya National Commission of science, Technology and Innovation. The consent and permission was granted by the County Commissioner and the Director of Education, Kisumu County. Proportionate stratified random and simple random sampling methods were used

Measures
Subjective well-being was conceptualized both as a one-dimensional and multi-dimensional concept. As a one-dimensional concept the older adults were asked a single question 'How can you describe your overall health?' The respondents were given alternative responses ranging from very poor, poor, moderate, good and very good health status. As a multidimension model subjective well-being consists of three components; satisfaction with life (SWLS, Diener, Emmons, Larsen, & Griffin, 1985), relative presence of positive emotions and relative absence of negative emotions (PANAS, Thompson, 2007). The five-item on SWLS was completed by the older adults to measure cognitive-judgmental process of satisfaction with life experiences. The items were scored on 5-point Likert scale and interpreted in terms of relative life satisfaction ranging from 1 = 'Extremely dissatisfied' 2 = 'Dissatisfied' 3 = 'Neutral' and 4 = 'Satisfied' 5 = 'Extremely Satisfied'. Secondly, PANAS was used to assess the emotional experiences of the older adults. PANAS consists of 10 words that describe different feelings and emotions. Five of the items measure the presence of positive affects; being inspired, alert, attentive, active and determined. The other 5 words measure the relative absence of negative affects; being afraid, upset, nervous, ashamed and hostile. The items on PANAS were rated on a 5-point scale ranging from 1 = 'Very slightly or not at all' 2 = 'A little' 3 = 'Moderately' and 4 = 'Quite a bit' 5 = 'Extremely' to measure the extent to which their affect has been experienced during the past few weeks. Merz, Malcane, Roesch, Ko, Emerson, Roma, Roma and Sadler (2013) acknowledges that both positive and negative affect represent independent constructs ranging from low to high levels of emotional experience. The scores for both SWLS and PANAS were computed as group data for the sampled population respectively. If the older adult is highly satisfied with life and frequently experiences positive emotions and relatively low negative emotions, the older adult is said to have high subjective well-being (Eryılmaz, 2010).

Data Analysis
Descriptive analyses were conducted to explore on the older adults' demographic characteristics: gender, age, education level, employment status, marital status, income levels, and other sources of income, living arrangements and the self-rated overall perceived health. Correlation and regression analysis was used to determine the level and predictors of subjective well-being. The Post Hoc Tukey HSD analyses was used to test for the variance in between group of demographic characteristics on subjective well-being of older adults.

Subjective Well-Being of Older Adults
Subjective well-being of older adults was measured both as one-dimensional as well as a multi-dimensional concept. As a one-dimensional concept the results in Table 1 revealed that the majority of the older adults experienced a poor self-perceived health status which translates to low subjective well-being. The ANOVA results in Table 2 revealed that majority of older adults were dissatisfied with their life (M = 2.24, SD = .42), they indicated varied responses of positive affect (M = 2.70, SD = 1.08) and that the majority (79%, n = 114; M = 1.46, SD = .58) experienced very little of the negative affect. Unlike positive emotions that were spread across 5-point Likert scale, negative affect was skewed towards the lower level of the scale. The responses on the three dimensions of satisfaction with life, positive affect and negative affect were computed into a single score showed that a majority (74%, n = 103; M = 2.24, SD = .42) of older adults experienced a low subjective well-being. The two measures both as a unidimensional concept and a multi-dimensional concept gave similar results that older adults experience a lower subjective well-being  Table 3 revealed that 97.5% of any variance in subjective well-being of older adults could be explained by variations in satisfaction with life, presence of positive affect and relative absence of negative affect and two demographic characteristics of age bracket and highest academic level. This explained a good prediction power for future observations.  Table 4 revealed p value <.001 which is less than the set value of 0.05, indicating that the regression model was statistically significant in predicting subjective well-being. Further the results showed that the relative absence of negative affect (β = .326, p = < .001) was the strongest predictor of subjective well-being, followed by presence of positive affect (β = .321, p = < .001), lastly the dimension of satisfaction with life (β = .317, p = < .001). On the contrary, the demographic characteristics did not reveal themselves as predictive variables of subjective well-being.

Influence of Demographic Characteristics on the level of Subjective Well-Being of Older Adults
The one-way Analysis of Variance results in Table 5 revealed that demographic characteristics of marital status, other sources of income, academic level, and self-perceived health had a statistical influence on subjective well-being of older adults. Post hoc Tukey HSD test was conducted to establish where the mean differences lied between groups of the respective categories.

Post Hoc Tukey HSD Test Analyses
The analysis on marital status indicated that all categories of being married, separated and widowed (p <.001) had a significant influence on older adult's subjective well-being as indicated in Table 6. In spite of 60%; n = 84 of the respondents getting added income from their children and relatives, Table 7 below shows that subjective well-being was highly influenced by older adults who have savings followed by those who have invested in businesses and finally those who get financial assistance from their children and relatives. Further, the study results revealed that subjective well-being is higher in older adults who attained secondary level of education. This gave an indication that having a higher educational level is associated with a higher subjective well-being of older adults as shown in Table 8. Moderate self-perceived health and poor self-perceived health (r = .275 * , p = .013) had more influence on subjective well-being (Table 9). Further, a half of the respondents indicated to having a poor self-perceived health and none gave an indication of experiencing good or very good self-perceived health.

Discussion
This study investigated into the influence of demographic characteristics on subjective wellbeing of older adults. The results revealed that the older adults experienced lower level of positive affect, and low level of absence of negative affect and were dissatisfied with their lives and experienced a low level of subjective well-being. These findings are similar to previous studies that found low level of subjective well-being in low income countries across the world (Pew Research Center Survey, 2014;Aboderin, 2010, Mwanyangala, et al. 2010 among the rural Tanzanian older adults. This results are also consistent with that of Kabole et al. (2013) that revealed 63% of the older adults do not experience goodness in old age. The unidimensional measure of well-being gave an indication that the older adults experience a poor self-perceived health status. This agrees with the results of (Romo, et al, 2012) that found a low level of well-being among the older adults with dementia. However, this results are dissimilar with findings of Lu et al (2010) and Sargent-Cox et al. (2012) that revealed a higher levels of subjective well-being while other studies revealed a medium level of subjective (Suh et al. 2012). Other studies, for example, Steptoe et al., (2015) showed that that well-being is low at all ages in sub-Saharan Africa.
The present study also sought to determine predictors of subjective well-being. The results revealed predictors of subjective well-being were purely intrapersonal; relative absence of negative affect (β = .326, p = < .001), presence of positive affect (β = .321, p = < .001) and satisfaction with life (β = .317, p = < .001). The demographic characteristics were not revealed as predictive variables of subjective well-being. This results were similar to the findings by Galinha and Pais-Ribeiro (2011) that intrapersonal predictive variables from within the individual self. Further, the results by Suh et al (2012) among Korean older adults showed that interpersonal variables; living with a spouse and perceived health status were among the predictors of well-being. The results of the current study point to the finding of longitudinal studies (Chung & Lee (2011);Siedlecki et al 2014;Nordbakke & Schwanen, 2013) recognises that predictors of subjecting well-being are dynamic depending on sample characteristics, timeframe and environment. Therefore, the need for a longitudinal study within the Kenyan context.
Thirdly the current study also determined the influences of demographic characteristic on subjective well-being. Marital status, other sources of income, academic level and overall self-perceived health had a significant influence on subjective well-being of older adults. Marital status was found to have the highest significant influence on subjective well-being of older adults. The categories of marital status; living with a spouse, being separated and widowed all had a significant influence on subjective well-being of older adults. The results of this study are consistent with the findings of Kenya National Commission on Human Right (2009) and Yang (2008) which identified marital status as significantly affecting subjective well-being of older adults. However, a study by Calys-Tagoe et al., (2014) revealed that marital status had insignificant effect on subjective well-being of the Ghanaians older population.
Sources of income in this case; own savings, business or investment significantly influenced the subjective well-being of older adults. The previous studies have indicated significant relationship between income (Calys-Tagoe, et al. 2014;Phaswana-Mafuya, et al. 2013, Lee, 2010 and levels of socio-economic status (Suh et al, 2012) with subjective well-being. This result relates with the findings of Ingrand et al (2018) that personal financial situation has a direct link with quality of life. Conversely, there was no literature to compare the relationship between subjective well-being and the sources of income. Notably for this study, it is not the amount of money received but the source of the funds that matters. This study showed that for the older adults who receive cash transfer from the government, having yet another source of income and particularly self-earned money; from savings, business or investment is an added attribute towards enhancing subjective well-being. This has an implication towards the design of the older person's cash transfer programme and interventions.
Earlier studies have found that higher education level was associated with psychological well-being (Cachioni et al., 2017;Ng et al., 2010;Gomez-Olive et al., 2010;Kyobutungi et al., 2010;Phaswana-Mafuya et al., 2013). The reviewed literature studies have compared the results of level of education with other demographic characteristics. For example, a study by Pinquart and Sörensen (2000) indicated that income and education were correlated to wellbeing. Contrary to Pinquart and Sörensen findings, the current study observes that education had a significant difference with subjective well-being while income insignificantly influenced subjective well-being. Findings by Mwanyangala, et al (2010) indicated that both a high academic level and a higher socio-economic status among the rural population in Tanzania were associated with a good health status.
Self-perceived health was found to have a significant mean difference with subjective wellbeing of older adults. This agree with studies such as that of Lee (2010) in Korean society and by Low et al, (2013) conducted in 20 other countries including Kenya. The results of current study disagree with the findings of a national study by Dai et al, (2013) that family relations had a stronger effect than health and economic status on subjective well-being. On the contrary, the current study reveal living arrangement had insignificant influence on subjective well-being of older adults.

Limitations, Implications and Future Directions
The limitations of this study relate to designs and instrumentation. The research examined behavioural, cognitive and emotional experiences connected to long recall of past events. The recall of these activities could have potential to under estimation or over estimation based on older adults' affective status at time of responding to questionnaire. Secondly, study population was confined to older adults who were participating in Older Persons Cash Transfer programme. These limitations prevented generalizations of study findings to other samples. Another limitation is in regard to small sample size in current study that limits analyses. In spite of these shortcomings, this study contributes to knowledge on subjective well-being of older adults.
The current study has the following implications and recommendations. First, poor selfperceived health emerged as a significant variable on subjective well-being of the older adults. The study therefore, recommends inclusion of psychological interventions in mental health care systems for ageing persons. Secondly, current study showed that secondary educational level had a significant influence on subjective well-being. In the pursuit to augment subjective well-being of older adults, this study recommends strengthening of educational program to encapsulate most citizens beyond secondary schooling. Lastly, this study revealed the importance of self-earned money towards subjective well-being. The study recommends Governments to enhance her citizen's earnings, savings and investment abilities during the productive working period so as to promote self-earned money in old age.
The current study suggests the following key directions for future research. Previous studies have given an indication that concept of subjective well-being is contoured within the context of time and space. Future research may purpose to carry out a longitudinal study with a larger sample size and a wider geographical coverage to allow for comparison and generalization of results. Secondly, current study involved low income and low academic level cohort. Therefore, this study suggests inclusion of high income and higher academic group to help increase an understanding into the concept of subjective well-being among the Kenyan older adults. Own source of income had a significant influence on subjective well-being, therefore further research would benefit in examining the influence of sources of income on subjective well-being of older adults.

Conclusion
The older adults experienced a lower level of subjective well-being, are dissatisfied with life, experience a little presence of positive affect and a little absence of negative affect. Higher academic level, being married, having own source of income and a poor self-perceived health status had a significant influence on subjective well-being of older adults. The study suggest that older adults require attention that is focused in priority areas of their individualized lives.