Introduction

Aging relates to the gradual weakening in physical and psychological activities of the daily living of a person [1, 2]. Thus, disability in performing a basic or instrumental daily activity is the inability of older adults to accomplish a task without support [3]. Generally, disability encompasses impairments, activity limitations, and participation restrictions [4].

Functional disability is a common public health problem that affects the health and quality of life of older adults [5] and causes them to be dependent and often institutionalized [6]. It is also known to be associated with the increasing nursing home admissions, utilization of health care, and the need for family and friends caregivers [7].

Several measures have been used to assess the functional status of the older adults, specifically, the basic Activities of Daily Living (ADL), Instrumental Activities of Daily Living (IADL), and the World Health Organization Disability Assessment Schedule (WHODAS II) [8–10]. Activities of Daily Living are referred to as the activities necessary for an independent life. Examples of ADLs are bathing, dressing, toileting, eating, continence, and transferring. Instrumental Activities of Daily Living refers to activities that require enough ability to make decisions as well as a strong interaction with the environment. Examples of IADL are housekeeping, telephone, shopping, transportation, and managing one’s financial resources and its medication [9, 11]. Instrumental Activities of Daily Living deficiencies mostly precede ADL deficiencies [12]. The WHODAS II scale measures difficulties in activities and daily-life participation [13].

Factors associated with functional disability in older adults have been explored in previous studies using ADL, IADL, and WHODAS scales [8, 10, 14, 15]. Prevalence of functional disability has been identified not to only increase with age [16–18] but relate with gender [18], obesity [19], chronic diseases such as diabetes, stroke, heart diseases [20, 21], rural residence [22] and visual impairments [17].

Although functional disability has been widely studied in developed country settings, there is a paucity of studies examining age-related functional disability in sub-Saharan Africa, and in particular Ghana. Assessing the factors associated with functional disability in this context will help develop appropriate preventive and health care policies and add in the identification of groups of older people at increased risk for functional disability based on these characteristics. Therefore, the purpose of this study is to assess the factors associated with the difficulties in performing basic and instrumental activities of daily living among older adults in Ghana.

Data description

Measurements

Data analysis

Descriptive statistics of the data are presented using frequencies with their related percentages for categorical variables and summary statistics (mean, standard deviation, minimum, and maximum) for continuous variables. The SAGE Wave II data was observed to have a hierarchical structure of participants nested within survey clusters which could bias the standard errors. Considering the hierarchical structure of the data (older adults nested within households), this study employed a multilevel analysis to accommodate various levels of variability in the data. The multilevel model, which is a regression model, does not only include fixed effects associated with individuals and households, but also random effects that permit specific household-based changes. The model also permitted the inclusion of error terms at each level, which makes it possible to identify changes in variance at each level across models [29]. The random effects can be considered as terms that model other unobserved variables that can affect the functional disability of older adults. Thus, this study uses the random-effects model to estimate the magnitude and significance of clustering (to adjust for dependence) captured through the Intra Class Correlation (ICC). The ICC was computed as the ratio of the variance at the cluster level to the sum of the variances at the individual and cluster levels [30].

The models are specified as follows: Model 1 is a null model that has no predictors but the variance is partitioned into between-individuals and between-household components by adding household-level shifts in the individual-level random intercept value. Model 2 adds predictors at the individual level while maintaining household random intercepts. The predictors at the individual level are sex, age, marital status, highest educational level, number of chronic conditions, physical activity, smoking cigarettes, fruit intake, vegetable intake, and drink alcohol. The log-likelihood ratio test was -4433.8. The intraclass correlation coefficient (ICC) was 0.61, which accounted for 61% of the total variance in the WHODAS at the individual level. The AIC and BIC for model 2 are 8921.7 and 9067.0, respectively.

Model 3 includes only predictors at the household level while maintaining household random intercepts. The predictors at the household level are water, fuel, household size, residence, and wealth quintiles. The log-likelihood ratio test was -5072.0. The intraclass correlation coefficient (ICC) was 0.45, which accounted for 45% of the total variance in the WHODAS at the household level. The AIC and BIC for model 3 are 10166.0 and 10225.2 respectively.

Model 4 is a full model that accounts for individual and household effects while maintaining household random intercepts. The log-likelihood ratio test of -4389.5 and AIC and BIC of 8848.9 and 9037.4, respectively confirm improvement to Model 1 when individual and household predictors are added. The intraclass correlation coefficient (ICC) was 0.61, which accounted for 61% of the total variance in the WHODAS at the individual level. After considering the number of significant variables, the values of the different model selection criteria (AIC and BIC), and the fact that Model 4 has household level significant variables and lower variance of the random effects (which means lower unexplained differences among households), we selected the best fitting model.

All the analyses were performed using STATA version 14. 0 (Stata Corp, College Station, TX, USA) statistical software with model fitting performed using SVY command to account for the complex survey design [31].

Model parameters were obtained using maximum likelihood estimation. To compare different sub-models, we used the Akaike Information Criterion (AIC), Likelihood Ratio Test (LRT), and Bayesian Information Criterion (BIC). Models with lower values of AIC, LRT, or BIC are often preferred. Multicollinearity of the predicted variables was checked using the Variance Inflation Factor (VIF) from multiple linear model, and a VIF value below 10 was considered acceptable [32].

Results

Table 1 presents the descriptive statistics of individual and household level characteristics of the study participants. From Table 1, the mean of functional limitations of participants was 5.2 (± 5.9), with a minimum of 0 and a maximum of 36. Out of 1610 participants, 55.4% had no functional limitation. For individual-level factors, 35. 8% of the participants belong to the age group 60–69 years, 53.7% of the participants are females, 51.0% of the participants are married, 43.9% of the participants had no formal education, 65.0% of the participants have no chronic conditions, 51.0% of the participants have high physical activities, 92.7% of the participants do not smoke tobacco, 67.9% do not drink alcohol, 56.8% of the participants have normal BMI, 95.6% of the participants have inadequate fruit intake, and 84.7% of the participants have inadequate vegetable intake. The mean age of participants was 65.8 (± 9.7) years, with a minimum of 50 years and a maximum of 110 years. The mean BMI of participants was 23.5 (± 4.9) kg/m², with a minimum of 13.3 kg/m² and a maximum of 46.3 kg/m². For household-level factors, 62.7% of the participants have their source of water from pipe or trunk water, 88.6% of participants use no-electricity or gas for fuel, 58.8% of the participants have a household size of less than 5, 58.8% of the participants are in the rural residence, and 24.8% of the participants are poorer.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Descriptive statistics of individual and household-level characteristics of study participants.

https://doi.org/10.1371/journal.pone.0277125.t001

Table 2 summarizes the results for four different models considered for the multilevel analysis. The statistically significant fixed effect factors of Model 2 were age, education, smoking cigarette, drinking alcohol, BMI, and fruit intake. The statistically significant fixed effect factors of Model 3 were fuel, residence, and wealth quintile. From the AIC, BIC, and log-likelihood results, Model 4 provides the best fit to the data. The statistically significant fixed effect factors of Model 4 were sex, age, highest level of education, number of chronic conditions, BMI, smoking cigarette/tobacco, drinking alcohol, fruit intake, water, fuel, household size, residence, and wealth quintiles.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Regression results for two-level models of WHODAS II score.

https://doi.org/10.1371/journal.pone.0277125.t002

Female older adults have a higher functional disability compared to male older adults, which was statistically significant (Coefficient (β) = 0.482, SE = Standard Error (SE) = 0.222). Older adults who belonged to the age group 60–69 years (β = 0.645, SE = 0.232), 70–79 years (β = 1.680, SE = 0.264), and 80 years and above (β = 3.859, SE = 0.388) had higher functional disabilities compared to older adults who belonged to the age group 50–59 years. Older adults who have completed university had lower functional disability compared to older adults who had no formal education (β = −0.770, SE = 0.315).

Older adults who have three or more chronic conditions present had higher functional disability compared to older adults who had no chronic condition present (β = 1.067, SE = 0.517).

Older adults who are current smokers (β = 7.104, SE = 0.495) and former smokers (β = 8.201, SE = 0.566) of cigarette/tobacco had higher functional disability compared to older adults who are non-smokers of cigarette/tobacco. Older adults who are current alcoholic drinkers had higher functional disability compared to older adults who are non-alcoholic drinkers (β = 0.602, SE = 0.241).

Older adults who are underweight (β = 0.781, SE = 0.289), overweight (β = 4.062, SE = 0.343), and obese (β = −7.532, SE = 0.412) had higher functional disability compared to older adults who had normal weight.

Older adults who have adequate fruit intake had lower functional disability compared to older adults who had inadequate fruit intake (β = −1.738, SE = 0.482). Older adults who used piped/trunk water had lower functional disability compared to older adults who do not use piped/trunk water (β = −0.861, SE = 0.230). Older adults who used electricity/gas had lower functional disability compared to older adults who do not use electricity/gas (β = −1.072, SE = 0.331). Older adults whose household size is greater than or equal to 5 had functional disability compared to older adults whose household size is less than 5 (β = 0.534, SE = 0.202).

Older adults who lived in urban residence had lower functional disability compared to older adults who lived in rural residence (β = −0.460, SE = 0.223). Older adults who were classified as poorer (β = −0.778, SE = 0.395) and middle wealth quintile (β = −1.197, SE = 0.282) had lower functional disability compared to older adults who were classified as poorest.

Discussion

Functional disability is a major health problem among older adults worldwide because of their inability to do daily activities, significantly affecting their quality of life [33]. Older adults who had functional disability are at increased risk for institutionalization, hospitalization, and death [34]. There are various determinants of functional disability and in diverse settings [35].

The findings from this study suggest that individual and household level characteristics matter in explaining the functional disability of older adults in Ghana. For individual-level effects, this study found that older female adults are more commonly to have higher functional disability scores as compared to older male adults. These findings are consistent with other studies [25, 36–38]. The high functional disability in women maybe due to their higher life expectancies [39, 40]. Also, this study found that functional disability scores increases with age especially in the oldest old people and this is consistent with studies [14, 25, 36, 37, 41, 42]. Age is identified to be related to functional disability because health problems, comorbidities such as disabilities, hypertension, and disabilities increase with age [37, 43].

Furthermore, the study found that older adults who have completed tertiary education are more commonly to have less functional disability scores compared with older adults with no formal education, similar to other studies [14, 41]. According to Goswami et al. [14] and Sinalkar et al. [41], educated older adults may have higher health literacy, maintain a better IADL performance, and maintain their physical function with a positive attitude.

Moreover, the study identified that older adults with three or more chronic conditions are more commonly to have higher functional disability scores compared with older adults with no chronic condition which is similar to a study by Ralph et al. [44]. According to Goswami et al. [14], older adults are more prone to communicable and non-communicable diseases which result in functional disability score.

The study found that older adults who have adequate fruit intake are more commonly to have less functional disability scores compared with older adults who have inadequate fruit intake which is similar to a study by Alipanah et al. [27]. According to Nicklett and Kadell [26], fruit intake protects older adults against the development of functional limitations and chronic conditions.

This study found that older adults who smoke tobacco or drink alcohol are more commonly to have higher functional disability scores compared with older adults who do not smoke tobacco or drink alcohol which is similar to the findings of Artaud et al. [45]. Smoking tobacco and alcohol intake increases the functional disability of older adults which makes them have long years in ill health [46–48].

However, the study found that physical activities was not associated with functional disability of older adults which is contrary to a study by Ćwirlej-Sozańska et al. [9]. Regular physical activity such as low intensity walking to more vigorous sports aids in physical and mental functions and the decrease of some effects of chronic diseases such as muscular weakness [49, 50].

For household-level effects, this study found that older adults who have household size more than five are more commonly to have less functional disability scores compared with older adults who have a household size less than five which is similar to other studies [51]. According to Saito et al. [52], the functional disability increases among older adults living alone and those living only with a spouse. This is because, older adults with few household members are unable to get support to carry out their functional abilities such as walking, dressing. This results in psychological disorders such as depression in older adults [53, 54].

Also, this study found that older adults who reside in urban areas are more commonly to have less functional disability scores compared with older adults who reside in rural areas which is similar to other studies [55, 56]. In Ghana, most of the occupation in the rural areas are farming activities which involve more physical activities and also, there are fewer activities in the rural areas that involve the use of buses or cars as means of transportation. Laditka et al. [57] stated that older adults in rural areas experience a longer expected period of disability in performing basic activities of daily living than older adults in urban areas. The higher rate of disabilities among older adults residing in rural areas may be due to less education, less access to health care and preventive services, pooer living conditions, and poverty [55]. Furthermore, this study found that older adults who drink pipe or trunk water are more commonly to have less functional disability scores compared with older adults who do not drink pipe or trunk water. Thomas et al. [58] stated that older adults are susceptible to dehydration as a result of acute and chronic health problems such as functional disability, which worsen thirst, decrease the ability to drink adequately, and increase the loss of skin, urinary, and respiratory fluid. Hence, older adults developing the habit of drinking water from the pipe helps them not to be dehydrated.

Finally, this study found that functional disability scores decreased among older adults who belong to the second to fifth wealth quintile. According to Wahrendorf et al. [59], there is a high disability among older adults with low wealth compared to those with high wealth. Poor older adults are unable to afford better healthcare when they are experiencing functional disability. Elwan [60] stated that conditions related to poverty such as inadequate water and sanitation, lack of access to healthcare, and poor living conditions increase the risk of disabilities in low- and middle-income countries. Nevertheless, it is easier for wealthy families to get caregivers to assist older adults with functional disabilities [61].

This study adds to the body of knowledge on aging and functional disability in a developing country setting such as Ghana. This paper makes three scholarly contributions. First, in the context of a developing nation such as Ghana, this study assesses the influence of individual-level factors such as performing physical activities, vegetable and fruit intake, and household level factors such as water, household size, and wealth quantile on functional disability with special emphasis on the older adults. The multilevel individual and household effects make the findings fascinating and the implications to the policy very vital.

Second, this study finds factors like fruit intake that may reduce functional disability among older adults. Thus, our findings highlights the need for nutritional interventions such as a high intake of fruits to reduce the functional disability among older adults. This is crucial right now because diets high in sugar, fat, and animal-based protein and sedentary lifestyle changes in developing countries make older adults more vulnerable to obesity, which has an impact on the functional disability. Finally, functional disability among older adults in Ghana presents a number of health risks, making it a public health issue that must be addressed with the appropriate policies, awareness campaigns, and interventions.

There were some limitations. This study is cross-sectional, thus, causal inference is limited. Rigorous designs such as longitudinal or randomized controlled trial with a strong qualitative data to better understand the relationship between individual and household related factors and functional disability of older adults.

Conclusion

The influence of individual and household level factors on the functional disability of older adults in Ghana is real. Functional disability among older adults is frequent and not only lowers the quality of life of the older adults but strains society’s limited resources for assistance, care, and rehabilitation. With the increasing population of the older adults in Ghana, there is an urgent need for the development of national policy on physical activity and nutritional interventions for the older adults which is currently not emphasized in the Nation Health Policy.