What accounts for quality of life among older adults? Examining chronic physical health, mental health, and social support among ethnically diverse populations in Ghana

ABSTRACT

Older adults have reduced income and chronic health problems that cause adverse quality of life outcomes. Social support also affects quality of life but there is not enough research that explains how interactions among these predictors affect older adults’ quality of life. To address this gap, we tested whether social support moderated the effect of mental health and chronic physical health on a multidimensional quality of life measure, comparing groups selected from three ethnically distinct communities in Ghana. We selected older adults who were at least 60 years old from three geographic/ethnic areas and administered the WHO multidimensional quality of life scale, geriatric mental status and social support (N = 353). The research findings showed that as individuals grow older their scores on all dimensions of the quality of life decreased. We also found that there were geographic differences in quality of life; the least urbanized and the most urbanized communities had poorer outcomes on all dimensions of quality of life. We also found that higher levels of depression were associated with lower quality of life. Finally, the research findings implied that lower levels of social support moderated the effect of mental health on past, present, and future activities and social participation.

KEYWORDS:

• Ageing
• depression
• Ghana
• quality of life
• psychological well-being

1. Introduction

The population of older adults has been increasing steadily over the years mainly due to improved health of the population in low- and middle-income countries (LMICs) and improved quality of life. This also means that more people are living into advanced old age (Gyasi & Phillips, 2020). In Ghana, an average growth rate of 0.51% has seen the population of adults who are 65 years and above increased from 2.5% in 1972 to 3.2% in 2021. The demographic shift in life expectancy has caused changes in the disease burden profiles of LMICs, with chronic noncommunicable diseases (NCDs) becoming a more common and growing public health challenge (de Graft Aikins & Agyeman, 2017). In addition to deteriorating physical health, ageing is also associated with a decline in changes in mental health and quality of life.

Quality of life is defined as the subjective perception of an ‘… individual's position in life in the context of their culture and value systems in which they live and in relation to their goals, expectations, standards and concerns’ (WHOQOL Group, 1993; pp. 153). Elements that predict quality of life are multifaceted, involving the physical, psychological, social, and economic factors. These facets can have either positive or negative outcomes on quality of life (Bilsky et al., 2016; Perry & Felce, 1995).

Quality of life of older adults is influenced by factors such as age, gender, health, social support, geopolitical factors and government policies (Kisvetrová et al., 2021; Patrício et al., 2014). Individuals who are healthy, have access to good medical care, have minimum stress, or are economically engaged usually have a better quality of life than those who do not (Low & Molzahn, 2007). In the same way, chronic ill-health, poverty, living in impoverished environments or conditions tend to result in poorer quality of life. Older adults may suffer chronic health problems, usually have financial challenges, may live in conditions that are less than optimal and therefore more likely to have poorer quality of life when compared to younger adults.

In this paper, we examined the factors that predict the quality of life of individuals older than 60 years in an LMIC, Ghana. Ghana, a multicultural country, is an example of contexts that are often under-represented in psychological research, particularly among older adults. This, therefore, presents the opportunity to address questions about ageing and quality of life comparing different cultural and economic groups. It also presents the opportunity to examine if these predictors affect different facets of quality of life inconsistently.

1.1. Predictors of quality of life among older adults

Increase in life expectancy and its associated higher prevalence of chronic physical health problems and poorer mental health among the elderly is a global phenomenon. The key indicators of quality of life include income/wealth, education, employability, physical health, mental health, and social domains such as marital status and social support (e.g. Cho et al., 2019). We focus on mental and chronic physical health problems and type of community (degree of urbanization) and examine the moderating effect of social support. We examine these effects on a multidimensional quality of life measure.

1.1.1. Mental health and chronic health condition

One of the most consistent findings in the psychological literature on ageing is that there is a moderate-to-high prevalence of depression among older adults. For example, in a rural population in China, it is reported that the prevalence of depression was about 5.9% with a higher prevalence among women than among men (Zhou et al., 2014). This rate is similar to prevalence in Ghana (6.7%) but lower in South Africa (2.7%) (Thapa et al., 2014). Depression causes substantial functional impairment which leads to deterioration of quality of life (Hofmann et al., 2017).

The prevalence of chronic physical illness and mental health is higher among older adults than among other demographic groups (de Graft Aikins & Agyeman, 2017). Older adults do not only become frail but are susceptible to different chronic diseases, most of which require treatment or hospitalization. Individuals who have multiple chronic health problems are associated with social and psychological problems that limit normal healthy functioning that adversely affect the quality of life of older adults (Megari, 2013). This association is even more stark when comparisons are made between communities distinguished by different socioeconomic characteristics, as may be the case between highly urbanized and non-urbanized communities.

1.1.2. Environmental/Community factors

There is ample research that shows that multiple environmental factors are associated with quality of life in older people (Gobbens & van Assen, 2018). Environments with better infrastructural facilities such as health and transportation and are safe tend to have populations with higher quality of life (Puts et al., 2007). These differences are likely to be more pronounced in low- and middle-income countries where limited resources do not allow for equitable distribution of resources such as provision of health-care facilities, transport, water, sanitation, and electricity (D’Ambrosio & Frick, 2007; Ravallion & Lokshin, 2010). Individuals who have access to good infrastructure are likely to have better living standards and high quality of life (Shucksmith et al., 2009).

1.1.3. Social support

The literature on the positive effect of social support on quality of life is extensive (Gallardo-Peralta et al., 2018; Helgeson, 2003; Kong et al., 2019; Şahin et al., 2019b). There is copious evidence that social support minimizes the effect of general life stressors, chronic health problems and mental health problems. For example, when social support is low, individuals 60 years and above in Ghana usually experience challenging circumstances (Braimah & Rosenberg, 2021). In another study, perceived social support accounted for more than one-fifth of the total variance in quality of life (Şahin et al., 2019a). Social support also mediated the relationship between factors such as poor physical health, disability or mental health and quality of life (Bélanger et al., 2016; Newsom & Schulz, 1996; Unalan et al., 2015). In yet another large-scale study in Chile, where researchers examined multiple levels of social support, they found that more than 25% of the variance in quality of life among older adults is explained by social support (Gallardo-Peralta et al., 2018).

1.1.4. Current study and hypotheses

There is a large body of literature that associate quality of life with financial independence, mental and physical health, physical environment and social support, and there is equally vast evidence that has shown that the role of social support as a moderator or mediator variable (Bélanger et al., 2016; Gallardo-Peralta et al., 2018). In this study, we are guided by two theoretical positions. First, that the effect of social support on quality of life is both direct and a moderator reducing adverse consequences of factors that impact quality of life. Second, we draw from the position proposed by Caron et al. (2019) (Caron et al., 2019) who argue that determinants of quality of life are multidimensional and therefore examining few variables fails to account for confounders in other important domains. According to Caron and colleagues, cross-sectional studies typically focus on populations with specific characteristics such as clinical groups or disadvantaged or ethnic sub-groups (Caron et al., 2019). These studies are unable to examine the inter-relatedness of determinants of quality of life. It is further argued that using a unidimensional measure of quality of life is too restrictive and has the potential to prevent the comparability of quality of life across different groups, even within the sample population (Kreitler & Kreitler, 2006).

The aim of this study is to determine the relationship between facets of quality of life, mental health and chronic physical health conditions, and the moderating effect of perceived social support in three ethnically distinct populations among people aged 60 years and older. To this end, we formulated the following hypotheses:

Hypothesis 1:

High levels of chronic health problems and higher levels of geriatric depression will lead to poorer outcomes in each of the five facets of quality of life (i.e. sensory abilities, autonomy, past, present and future activities, intimacy, and social participation).

Hypothesis 2:

Older adults in more urbanized communities will have better outcomes in each of the five facets of quality of life than those in less urban communities (i.e. sensory abilities, autonomy, past, present and future activities, intimacy, and social participation).

Hypothesis 3:

Perceived social support will moderate the relationship between the predictors (geriatric depression status and chronic health conditions) and facets of quality of life (i.e. sensory abilities, autonomy, past, present and future activities, intimacy, and social participation).

2. Materials and methods

2.1. Participants

Participants for this study were selected from three districts from three regions in Ghana: Ga West (Greater Accra), Sunyani (Ahafo) and Navrongo (Upper East). Our aim was to target the three broad geographic zones of the country—Southern zone, Middle zone, and the Northern zone, to provide a fair representation of country. The reason for this selection is because communities in the south of Ghana are more affluent, with levels of deprivation increasing as one moves away from the south and middle-belt regions to northern Ghana. Accra, located in the south, is the capital of Ghana and therefore there are facilities and infrastructure typically associated with most modern cities. The middle belt of Ghana (particularly in the Ashanti, Bono and Ahafo regions) has access to natural resources such as gold and therefore has relatively more wealth. The northern regions of Ghana are relatively less urbanized and under-developed and are associated with higher levels of poverty than the southern and middle regions of the country. The sampling technique first involved selecting three regions from each zone and then selecting one district from each region. The selection of the district followed a convenience method. We then selected three communities that had fair representation of urban, per-urban, and rural communities. At the final stage of selection, we used the purposive technique to select participants in selected households. We included all adults who are 60 years and older within the household. Every individual in a household aged 60 years and above was selected. The inclusion criterion was based on the qualifying age of 60 years and above. This is based on the categorization provided by Ghana Statistical Service that places individuals who are above 60 years in older adult groups (Awuviry-Newton et al., 2021). Previous studies have also defined older adults as individuals who are 60 years or above (Braimah & Rosenberg, 2021). In Ghana, the compulsory retirement age is 60 years and therefore individuals at this age assume different occupational and social roles, transitioning from one phase of life that involves active participation in social and economic activities to another where they are less engaged. The total number of participants who completed the survey was 353; 111 from Accra, 122 from Sunyani, and 120 from Navongro.

2.2. Measures

In this study, three variables, geriatric depression, chronic physical illness, and social support were used to predict World Health Organization Quality of life measure for older adults (WHOQOL-OLD). Social support was used as a moderator. We also asked questions about socio-demographic and economic characteristics such as age, marital status, living arrangements (if they live on their own or depend on somebody) and income status.

Geriatric Depression Scale—short form is a scale developed as a screening tool for depression among older adults (Yesavage et al., 1983). There are two versions of the Geriatric Depression Scale, the original version made of 30 items and a short form made of 15 items (Sheikh & Yesavage, 1986). The short form was used for this study for ease of administration. It has been shown in previous studies to have good validity and high correlations with both the long form and other quality of life measures (Herrmann et al., 1996).

Perceived Social Support is a 12-item scale developed to assess perceptions of social support from three sources—family, friends, and significant others (Zimet et al., 1988). While the scale yields three dimensions, scores on the items can be aggregated for a composite score. In this study, we use the sum of scores on all 12 items.

To measure chronic health status, we asked participants to indicate the presence or absence of any chronic disease they have either being diagnosed with or receiving treatment for such as diabetes or hypertension. Score for chronic health status was created by simply counting the number of diseases endorsed by the participant. More symptoms indicate high levels (or worse) of chronic health status.

The World Health Organization Quality of Life-Old (WHOQOL_OLD) is a 24-item, 6-facet measure of quality of life for older adults. This is a brief version of the original WHO quality of life scale purposely designed to measure quality of life among older adults (Power et al., 2005). We used five dimensions of the quality of life scale and these are as follows: sensory abilities, autonomy, past, present, and future, social participation, and intimacy.

2.3. Data collection

The questionnaire, made up of the respective measures of the constructs, was translated using the standard translation and back translation methods (Üstun et al., 2005). Firstly, the questionnaire was translated into the three local languages commonly spoken within the selected communities—Ga in Accra, Akan Twi in Sunyani and Kasem in Navrongo. Secondly, the three new versions were then back translated into English by translators who were not familiar with the original version of the measures. The third stage involved a comparison between the back translated and the original English versions focusing on contextual and linguistic equivalence between the translated and original versions of the measures. The quality of life scale has been validated and published in a separate paper. It is not referenced yet to keep anonymity. The interviews were conducted by research assistants who had university degrees in the social sciences or related courses and were proficient in the local languages. The research assistants had a 5-day training during which they were taken through the English and then the translated versions. The research assistants pretested the questionnaires and were involved in the final corrections of the survey instrument. The interview for the data collection was conducted in the three local languages and for participants who were proficient in English, the interview was conducted in English.

We received ethical approval from an ethical committee in a local University (Protocol Number: ECH 105/17–18) and the research was done in compliance with ethical requirements in Ghana. All participants signed or thumb-printed an informed consent form. Participants received either phone credits or cakes of soap up to the equivalent of $1.00. Individuals who volunteered to participate were interviewed in their houses.

2.4. Data screening and analyses

We used the SPSS version 23 software to perform the statistical analyses. We first conducted descriptive analyses for the relevant variables, calculating frequencies and means and standard deviations. For the standardized measures, we calculated Cronbach’s Alpha and calculated normality using Skewness and Kurtosis tests. Analyses of the missing data showed that except for gender, missing data fell within 1% to 2%. To test the hypotheses, we used multiple linear regressions to analyze the effects of depression and the number of chronic diseases on the different measures of quality of life controlling for age, gender, marital status, and living arrangement. We also used regression to test for the moderation effect of social support.

We tested the reliability of the measures using Cronbach’s Alpha. We found that all the measures had moderate-to-high internal consistency (0.79 to 0.94) (Cortina, 1993). Measures of normality, using Skewness and Kurtosis showed normal distribution for all measures except the geriatric depression status. For the measures that were normally distributed, skewness values ranged from .096 to .701, and Kurtosis values ranged from −.618 to .375. These values, which fall between ±1, are acceptable indicators of normality (Nunnally & Bernstein, 1994). To test the hypotheses, we used One-way Analysis of Variance (ANOVA) and hierarchical multiple regression analyses. We also used Hayes Process Model for SPSS to test for the hypothesized moderator effect. The conditional effect of the predictors (chronic health status and mental health) on the quality of life measures was calculated at values of social support equal to the 16^th, 50^th, and 84^th percentiles of the distribution in the sample.

3. Results

3.1. Descriptive analyses

There were more females (67.2%) than males, and the average age was 71.65 years (SD = 9.22). When examining the age of participants in each community, the average age in Navrongo was slightly higher (73.70) than for Accra (70.93) and Sunyani (70.15). A little over a third (34.5%) were married or living with a regular partner. More than 70% were not employed or not on a regular income (Table 1). The data also showed that less than 50% of the participants received regular remittances from a relative, such as a child. Navrongo had the highest number of participants who are not employed (93.3%), followed by Accra (63.1%) and Sunyani (61.5%). More participants in Sunyani (37.7%) were on a regular income than observed for Accra (25.2%) and Navrongo (25.8%). The data also suggested almost 50% of the participants across the three locations received regular remittances from relatives—Accra − 50.5%, Navrongo − 47.9% and Sunyani − 48.4%. These statistics imply relative poverty among older adults.

Table 1. Descriptive information of demographic characteristics and key variables.

	District
	Accra	Navrongo	Sunyani
Gender	Frequency (%)
Female	76 (70.4%)	82 (70.7%)	71 (60.7%)
Male	32 (29.6%)	34 (29.3)	46 (39.3%)
Marital/Relationship status
Unmarried/live alone	31 (72.1%)	46 (61.7%)	48 (39.3%)
Married/cohabiting	80 (29.9%)	74 (38.3%)	74 (60.7%)
Unemployed status
Unemployed	70 (63.1%)	112 (93.3%)	75 (61.5%)
Employed	41 (36.9%)	8 (6.7%)	47 (38.5%)
Regular income
No	83 (74.8%)	89 (74.2%)	76 (62.3%)
Yes	28 (25.2%)	39 (25.8%)	46 (37.7%)
Regular remittances
No	55 (49.5%0	61 (52.1%)	63 (51.%)
Yes	56 (50.5%)	56 (47.1%)	59 (48.4%)
	Mean(SD)
Key variables
Age	70.98 (9.63)	73.75 (9.05)	70.25 (8.70)
Geriatric depression	6.41 (3.64)	5.98 (3.57)	4.87 (3.72)
Chronic health status	1.90 (1.00)	0.80 (0.90)	0.80 (0.77)
Social support	44.51 (15.96)	55.46 (13.65)	49.91 (16.95)
WHO quality of life facets
Sensory abilities	13.44 (3.68)	11.71 (4.50)	14.60 (4.33)
Autonomy	15.67 (3.76)	13.80 (3.79)	17.38 (2.72)
Past, present, & future	14.25 (3.80)	14.59 (3.54)	16.48 (3.10)
Social participation	13.45 (3.91)	10.18 (4.40)	15.30 (4.20)
Intimacy	15.07 (3.43)	17.74 (3.44)	17.34 (2.63)

An examination of the health status of participants showed that almost two-thirds of participants (n = 220; 62.32%) had at least one chronic health condition. The results showed that hypertension, sensory problems, diabetes, and arthritis were the most common health problems associated with older adults.

To examine the preliminary bivariate associations between age, geriatric depression status, social support and the quality of life dimensions, we calculated Pearson’s correlations. The direction of the correlations between age and quality of life dimensions suggested that older participants had poorer quality of life than younger participants. Geriatric depression was significantly associated with all the quality of life dimensions. The highest correlations were found with autonomy (r = −.51), social participation (r = −.53), past, present, and future activities (r = −.59). We also found significant negative associations between chronic health status and all quality of life dimensions, except intimacy. The highest correlation was with autonomy (r = −.27). See Table 2 for all correlations.

Table 2. Mean, SD, Cronbach alpha and Pearson correlations.

		2	3	4	5	6	7	8	9	10	11	12	13	14	15
1	Age (M = 71.65, SD = 9.32)	−.07	.17**	.04	.10	−.32**	−.04	.20**	.07	.01	−.20**	−.33**	−.32**	−.35**	−.22**
2	Gender (Male = 1)		.53**	−.05	.01	.03	.09	−.26**	−.15**	.16**	.15**	.09	.21**	.13*	.19**
3	Marital Status (Married = 1)			.01	−.10	.03	.10	−.16**	−.10	.13*	.10	.09	.09	.08	.15**
4	Live with someone (Yes = 1)				−.13*	.07	.02	.01	−.04	.09	.07	−.04	−.04	.00	.03
5	Regular income (Yes = 1)					−.08	−.02	−.04	.10	.22**	.05	−.07	−.04	−.06	.05
6	Employed (Yes = 1)						.02	−.24**	−.06	.03	.21**	.25**	.34**	.42**	.23**
7	Location (Accra = 0, Navrongo = 1, Sunyani = 2)							−.17**	−.17**	.13*	26**	−.59**	−.51**	−.52**	.44**
8	Depression (M = 9.26, SD = 3.69, α = .79)								.22**	−.31**	−.59**	−.42**	−.51**	−.52**	−.44**
9	Health (M = 0.92, SD = 0.91)									.12*	−.19**	−.27**	−.14**	−.21**	−.06
10	Social support (M = 50.10, SD = 16.20)										.42**	.16**	.27**	.19**	.43**
11	Sensory abilities (M = 13.25, SD = 4.36 α = .94)											.50**	.64**	.62**	.58**
12	Autonomy (M = 15.62, SD = 4.36 α = .89)												.62**	.60	.35**
13	Past, present & future activities (M = 15.14, SD = 3.61 α = .84)													.68**	.55**
14	Social participation (M = 12.99, SD = 4.69 α = .91)														.53**
15	Intimacy (M = 15.75, SD = 3.38 α = .93)

Note. ** Correlation is significant at the .01 level; * correlation is significant at the .05 level.

To examine location/community differences on the key variables, we used one-way ANOVA to test the differences among the locations on chronic health status, geriatric depression, social support and the quality of life dimensions. The summary of the results is reported in Table 3. We found that there were significant differences among the locations on chronic health status (F = 7.21), geriatric depression (F = 5.64) and social support (F = 14.25). We used Scheffe's posthoc test to examine group differences. Participants in Accra, the most urban location, had marginally higher mean on chronic health status than older adults in other locations. We also found a difference between participants in Accra and Navrongo on geriatric depression but no differences between Sunyani and other locations. On social support, posthoc analyses showed that all groups were different; participants in Navrongo and Accra had the highest and lowest means, respectively. There was a location effect on all dimensions of quality of life. Participants in Sunyani had significantly higher scores on all quality of life dimensions except for sensory abilities; the difference between Sunyani and Accra was not significant. There were significant differences between Accra and Navrongo on sensory abilities, autonomy and social participation in favor of Accra. The summary of the results is shown in Table 3.

Table 3. Summary of analysis of variance results for chronic health status, geriatric depression, social support and quality of life dimensions.

							WHO Quality of life Dimensions
	Chronic health status		Geriatric depression		Social support			Sensory abilities		Autonomy		Past, present & future activities		Social participation		Intimacy
	Mean	SD	Mean	SD	Mean	SD		Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Accra	1.19	1.00	6.41	3.64	44.51	15.96		13.44	3.68	15.67	3.76	14.25	3.80	13.45	3.91	15.07	3.43
Navrongo	0.80	0.90	5.98	3.57	55.46	13.65		11.71	4.50	13.80	3.79	14.59	3.54	10.18	4.40	14.74	3.44
Sunyani	0.80	0.77	4.87	3.72	49.90	16.95		14.60	4.33	17.38	2.72	16.45	3.10	15.30	4.20	17.34	2.63
F, df, p	7.21, 350, .001		5.64, 352, .004		14.25, 352, .001			14.49, 351, .001		32.63, 352, .001		14.15, 350, .001		45.94, 350, .001		23.86, 350, .001

We conducted multiple regression analyses to test for the effects of satisfaction with life, geriatric depression, and chronic health status on quality of life. We also used Hayes Process Model to test whether perceived social support moderated the relationship between satisfaction with life, geriatric depression, and chronic health status on quality of life. For all the multiple regression analyses, age, gender, marital status, living arrangement (whether the individual lived with someone or alone), availability of regular income and occupational status were entered as control variables.

3.2. Hierarchical regression analyses

We conducted hierarchical regression analyses to test the effects of chronic health status and geriatric depression on the five dimensions of quality of life. Social and demographic characteristics (age, gender, marital status, living arrangement, availability of regular income and occupational status) were entered at the first step (as control variables), because of the possibility that these variables affect quality of life. We dummy-coded location (Accra and Navrongo) and entered them at the second step. All binary variables were coded as 0 and 1. Geriatric depression and number of chronic health conditions were entered at the third step.

In Table 4, we have reported the unstandardized B weights, confidence interval and the R² for each regression analysis. The results showed that for each quality of life dimension, the overall regression result was significant (Table 4). Age and occupational status were significantly associated with all quality of life dimensions. Being younger and being employed were positively related to facets of quality of life. Gender was also significantly associated with autonomy (B = 1.54), social participation (1.08), and intimacy (B = 1.25): being male has a better outcome on quality of life.

Table 4. Relation between geriatric depression, chronic health status and facets of the WHOQOL-OLD (unstandardized regression coefficients).

	Sensory abilities		Autonomy		Past, present, & future activities		Social participation		Intimacy
Predictors	B	95% CI	B	95% CI	B	95% CI	B	95% CI	B	95% CI
Age	−0.13**	−0.18 – −.74	−0.92**	−0.13 – −.05	−0.60**	−0.10 – −0.02	−0.12**	0.22 - −0.12	−0.06*	−0.10 – −0.02
Gender (Male = 1)	0.44	−0.75–1.21	1.54**	0.56–2.51	0.79	0.23–1.81	1.08*	0.16–2.51	1.25*	0.32–2.18
Marital status (Married = 1)	0.10	−1.01–1.21	−0.44	−1.35–0.47	−0.12	−1.07–0.83	−0.29	−1.71–0.62	0.21	−0.66–1.07
Living arrangement (Living with someone = 1)	0.40	−1.71–1.91	−0.42	−1.49–0.66	0.730	−0.389–2.01	−0.12	−0.99–1.77	0.23	−0.79–1.26
Are you on a regular income? (Yes = 1)	0.25	−0.81–1.32	0.51	−0.37–1.39	0.80	−0.11–1.72	0.24	−0.83–1.31	0.09	−0.75–0.92
Occupational status (Yes, employed = 1)	1.52**	0.47–2.56	2.12**	1.226–2.99	1.05*	0.16–1.95	3.38**	2.33–4.43	1.20*	3.82–2.02
Location
Accra = 1	−0.82	−1.89–0.25	−1.42**	−2.27 – −0.56	−2.06**	−2.95 – −1.16	−1.61**	−2.63 – −0.60	−2.03**	−2.84 – −1.21
Navrongo = 1	−1.75*	−2.86 – −.64	−2.39**	−3.27 – −1.51	−1.04*	−1.97 – −0.10	−3.64**	−4.70 – −0.26	−1.79**	−2.63 – −0.94
Social support	0.06**	0.03–0.08	0.07**	0.06–0.10	0.09**	0.07–0.11	0.07**	0.05–0.10	0.09**	0.08–0.11
Geriatric Depression	−0.41**	−0.53 – −.29	−0.42**	−0.52 – −0.32	−.55**	−0.65 – −0.46	−0.55*	−0.67 – −0.44	−0.35**	−0.44 – −0.25
Chronic health status	−1.00**	−1.45 - −0.54	−0.14	−0.14–0.50	−.14	−0.49 – −0.23	−0.55**	−0.97 – −0.14	0.28	−0.78–0.64
R^2	0.32		0.46		0.46		0.52		0.42

Our first major goal was to test the assumption that higher levels of chronic health problems and higher levels of geriatric depression will lead to poorer outcomes on each of the five facets of quality of life. The results showed that there was a significant association between geriatric depression and all the dimensions of quality of life; higher scores on depression were associated with poorer outcomes on quality of life measures (Table 4). With respect to chronic health status, however, the results were significant for two dimensions: sensory abilities (B = −1.00) and social participation (B=−.55).

The second major goal in the study was to examine the assumption that participants in more urbanized communities will have higher scores on the dimensions of quality of life than those in less urbanized communities. For example, we expected that participants in Accra and Sunyani will have higher scores on quality of life measures than participants in Navrongo. We dummy coded the location variable and tested for the effects of Accra and Navrongo. The regression results showed a mixed pattern. We found that participants in Accra were more likely to have poorer quality of life on autonomy (B = −1.42), past, present, and future activities (B = −2.06), and intimacy (B = −2.03). Among participants from Navrongo, we found that they were more likely to have poorer outcomes on sensory abilities (B = −1.75) and social participation (−1.04).

The third major goal in this study was to examine the assumption that social support moderated the relationship between the predictors (geriatric depression and chronic health conditions) and the facets of quality of life. We used the Hayes Process Model to test for the moderator effect of social support. The results showed significant overall results but small and inconsistent moderator effects on the different dimensions of quality of life. There was a significant moderator effect for Autonomy (R² = 0.011, p < .022), and past, present, and future activities (R² = 0.014, p < .004). The results from the moderation showed that as the level of social support increases, the relationship between depression and the quality of life facets weakens (Table 5). The impact of depression on quality of life was highest when social support was lowest. With respect to chronic health status, there was a significant effect on past, present and future activities (R²=.252, p < .001).

Table 5. Moderator effects of perceived social support on quality of life dimensions—unstandardized coefficients.

	Geriatric depression				Chronic health status
	Autonomy		Past, present and future activities		Past, present and future activities
Social Support*	Effect	p	Effect	p	Effect	p
Low social support	−.556	.001	−.586	.001	−.837	.001
Intermediate social support	−.445	.001	−.465	.001	−.392	.011
High social support	−.346	.001	−.359	.001	−.018	.532

*low = 16^th, moderate = 50th, high = 84^th percentiles.

4. Discussion

The results of this study showed that quality of life is correlated with geriatric depression, chronic health status, and social support, although not consistently correlated across all facets of quality of life. We also found that there was a significant association between location and quality of life; participants in Accra had significantly poorer quality of life compared to participants in Navrongo and Sunyani.

4.1. Chronic physical and mental health leads to low quality of life

Older adults tend to have multiple chronic health problems, with almost half of older adults having two to three comorbid medical conditions. Multiple health and frequent ill-health among older adults adversely affect quality of life. The literature on geriatric depression is consistent with the finding that higher levels of depression are associated with lower quality of life (Jemal et al., 2021; Li & Conwell, 2007; Lin et al., 2014). This finding is also common among studies that focus on the general population (e.g. Cho et al., 2019). For example, Naumann and Byrne (2004) found that scores on WHO quality of life for older adults were strongly correlated with severity of depression, an indication that quality of life is closely tied to depression or mental health (Naumann & Byrne, 2004). Similarly, Cho and colleagues also found in South Korea that among older adults, mental health struggles among others were significantly associated with low quality of life (Cho et al., 2019). Scores on geriatric depression were significantly associated with all quality of life dimensions, and this significance was retained significantly after controlling for age, sex and other demographic variables. This suggests that mental health seems to have a unique association with quality of life, a finding that has often been cited in the literature (Cho et al., 2019; Naumann & Byrne, 2004).

4.2. Location/Community differences in quality of life

Research typically tends to show that living in or near large cities has advantages that include better access to infrastructure such as access to medical services and transportation, opportunities for employment, financial security, access to adequate health-care services, and good housing improves quality of life (Alkire et al., 2014; Shucksmith et al., 2009). Our findings partially support this with results that showed that being in Navrongo, the least urbanized community, was associated with low scores on all quality of life dimensions. Accra, the most urbanized community in the study, was also associated with low scores on three of the quality of life measures. While this was unexpected, there are possible reasons for the phenomenon. In large cities and more urban communities, the traditional familial safety nets provided by the extended family system have gradually become ineffective and unreliable because of migration and high rate of unemployment among the youth (Boon, 2007). With reduced income in old age, particularly for individuals who do not receive pension, financial strain poses a significant burden in a strictly monetized environment. The cost of amenities such as electricity and water are relatively expensive. These facilities, therefore, may be available but not accessible to vulnerable groups. With weak or inadequate support systems, older adults are more likely to experience poor quality of life. We may conclude that urbanization does not fully protect individuals from challenges associated with ageing. This finding, however, is not an isolated one. In Italy, for example, males aged 65 years and living in rural areas had better quality of life than their age counterparts living in urban areas (Carta et al., 2012).

4.3. Social support as a moderator

We found a significant moderator effect on two dimensions (autonomy and past, present and future activities); when social support was high, the impact of geriatric depression and chronic health status on quality of life was low. The implication is that, when older adults are provided with necessary care and attention, or they have people around to communicate with, the adverse impact of chronic physical and mental health is low (Wedgeworth et al., 2017). Illness has an adverse impact on quality of life. This gets worse as individuals grow older because older adults have multiple chronic health challenges and a declining ability to recover quickly and completely. What our research shows is that social support mitigates the negative impact that this naturally occurring process has on the quality of life of the ageing individual (Braimah & Rosenberg, 2021). Ghana is not a monolithic society but one of the most enduring characteristics of the Ghanaian culture is the support that is provided to the aged. This social buffer is waning rapidly. Migration, economic challenges, and search for improved living standards mean that younger individuals leave their communities for ‘greener pastures’. With dwindling protection from social support, older adults will begin to have worse outcomes in quality.

Although this paper takes a perspective mainly from Ghana, a sub-Saharan LMIC, the quality of life, and by extension, mental health of older people is a major global public health issue that has far-reaching implications for health of other sociodemographic groups. Understanding the mental health needs of the ageing populations provides opportunities to provide relevant interventions to respond appropriately. Our findings show that there is a multiplicity of factors associated with quality of life, and the relationship between these variables and quality of life dimensions may be indirect. This has implications for research and for policy. With the knowledge that predictors of quality of life do not have direct effects, researchers should be guided to explore bidirectional or nonlinear effects. These findings should also be reflected in policy about the aged. With the knowledge that living in relatively more urbanized communities does not provide protection for older adults, it means policymakers cannot provide ‘one-size fits all’ policies but rather provide needs-based policies that directly address challenges that individuals face within their communities.

The strength of this study is the inclusion of three ethnically distinct communities, selected from different geographic zones in Ghana, which therefore makes this sample representative of the population country. The study also used a translated version of the multidimensional WHOQOL_Old measure. This allows for cross-cultural comparisons both within Ghana and with other contexts outside Ghana or Africa. The mixed pattern of results on different facets of quality of life is evidence that supports Caron et al. (2019) view that when we examine multiple variables we can better examine the inter-relatedness of the variables.

5. Study limitations

The present study is not without limitations. First, as this study was based on a cross-sectional research design, it is difficult to establish causal relationships. We are therefore cautious in our interpretation of the results since causality implies that one variable occurs in time before another. We cannot prove that in this study. For example, while the results show that social support predicts quality of life, we are mindful of the fact that these associations, as is the case with correlational studies, are possibly influenced by variables other than the variables understudy.

Another potential limitation to this study is that the key measures are self-report measures. Self-report method is generally affected by response bias including social desirability and the ability to recall information accurately. Again, our measure of health status was computed from the number of chronic diseases suffered by the participant. Health status can be more complex than the presence or absence of disease conditions. For example, the ability to afford necessary care, which may include expensive medical processes and medications are likely to outweigh the effects of number ailments. Severity of illness and nature and extent of debilitation (more than multimorbidity) can cause poor quality of life outcomes. In this study, we did not measure these complexities and are therefore cautious in our interpretation of the effect of health on quality of life. The effect of socioeconomic status (SES) on wellbeing and quality of life is well established. We did not directly measure SES and wealth because of the limited validity of responses about income. We therefore used income availability as a proxy. While this may provide some indication of SES, we are mindful of the limitation this approach has in accurately estimating the effect of SES on quality of life. Finally, we acknowledge that studying older adults above 60 years living in three ethnically bound zones does not necessarily represent all older adults in Ghana. There are 261 districts in Ghana in 16 regions made up of multiple ethnicities, religious and educational groups. We selected participants from three regions, obtaining a sample of districts that represents less than 30% of the number of districts.

6. Conclusion

The major objectives in this study were to examine if having higher levels of chronic health problems and geriatric depression will be negatively associated with multidimensional quality of life among older adults. We also examined if living in more urbanized communities will have better quality of life than those in less urban communities and finally examined if perceived social support influenced the relationship between geriatric depression status and chronic health conditions' multidimensional quality of life.

We found that older age, geriatric depression, and lower level of social support were associated with low quality of life. The results of this study underscore the importance of social support and mental health care as significant factors in the lives of older adults. These relationships, as we found, varied slightly across geographic locations. Being in an urbanized community, while important, does not completely insulate older adults from difficulties that older adults face. This study contributes to the limited current knowledge about explaining the multifaceted predictors of quality of life among older adults, especially in sub-Saharan Africa. We have also provided data that show the impact that social support plays on individuals who are at risk of mental health and quality of life impairment. These should guide caregivers and policymakers in developing interventions for older adults.

Authors’ contributions

AA drafted the manuscript. AA, CSA and AdGA revised the manuscript. AA, CSA and AdGA approved the manuscript.

Availability of data and materials

The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Ethics approval and Consent to participate

Approval was received from the Ethics Committee for the Humanities, University of Ghana (Protocol Number: ECH 105/17–18). All participants signed or thumb-printed an informed consent form before participating in the study. The entire process was approved by the local IRB and the funding agency. The research was done in compliance with ethical requirements in Ghana, and all methods used were in compliance with the relevant guidelines and regulations.

List of abbreviations

WHO	=	World Health Organization
WHOQOL	=	World Health Organization Quality of life

Role of the funding source

The funder has no role in the collection, analysis and interpretation of data and in the writing of the report or in the decision to submit the article for publication.

Acknowledgements

We acknowledge the following for their role in the data collection: Accra: Pearl Lamptey, Naa Akler Odonkor, and Queen Angela Newman Navrongo: Felix Ananga, Nancy Bugase, and Mabel Koyiba Sunyani: Joseph Manukure and Rebecca K. Mintah,

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This project was made possible with financial support from the University of Ghana Research Fund. [Grant Number: UGRF/10/ILG-079/2016-2017].