Mobility limitations and emotional dysfunction in old age: The moderating effects of physical activity and social ties

This study aims to examine the association between mobility limitations and emotional dysfunction among older Ghanaians and evaluate the buffering effect of physical activity (PA) and social ties in this association.


| INTRODUCTION
Aging in sub-Saharan Africa (SSA) is much faster than in today's richer countries due to the rising life expectancy as a net of declining fertility and medical technology advancement. 1 This is highly likely to compromise functional health outcomes due, at least, in part, to poor environmental factors and the incidence of chronic and neurodegenerative conditions, including stroke, diabetes, Alzheimer's, and Parkinson's diseases. 2,3 Estimates show that 23%-47% of adults aged ≥50 years report some deficits in mobility and functional performance, including difficulty in walking, climbing, and changes in gait. 4,5 Research has linked mobility limitations (defined as the reduced capacity to move around or part of the body) to incident depression, 6 increased risk of falls, 7 psychological distress, 2,8 risk of death, 9 healthcare utilization, and long-term care expenditures. 7 Given that the World Health Organization's (WHO) International Classification of Functioning (ICF) model emphasizes a robust assessment of the epidemiology of mobility limitations, 10 understanding the deleterious effects of mobility limitations becomes essential for emotional wellbeing and healthy aging, particularly in SSA.
Systematic reviews and longitudinal studies have shown that mobility limitations and adverse mental and emotional health outcomes commonly coexist, 11,12 although the relationship could be bidirectional. 13 Using data from the English Longitudinal Study (n = 8780), Philip et al. 12 found a longitudinal association between mobility limitations and loneliness. Also, among 3772 older Malaysians, Ahmad et al. 14 found that those with IADL and ADL difficulties were two and three times, respectively, more likely to have depression. Using pooled data involving 8520 Norwegians, Backe et al. 15 found a significant link between mobility limitations and psychological distress, including anxiety and depression. The impaired mobility and emotional dysfunction link have been hypothesized to be mediated through pain catastrophizing, inflammatory responses, and neuroendocrine processes. These may reduce independence and promote affective and mood disorders. 7,16 Mobility limitations are frequently related to localized knee and back pains which can render older adults' emotional well-being at risk. 17,18 Reduced functional capacities may also induce stressrelated alienation and social deficits, including a poor sense of belonging, which are proximate risk factors for emotional dysfunction. 19 However, evidence on the association of mobility limitations with emotional dysfunction is limited in older adults in LMICs, although heterogeneity and sociocultural diversities exist in older global populations.
PA and social relationships are coping and protective resources for stressful life events and relate favorably with physical function and emotional well-being in later life. 20 Several systematic reviews and meta-analyses have shown that exercise intervention positively impacts emotional well-being in older adulthood. 21,22 Others have demonstrated that having a stronger constellation of social ties is significantly associated with improved mental and emotional health. 23,26 For example, in systematic reviews, pooled-and metaanalyses of 87 studies, Marquez et al. 24 found that PA improves the adult's quality of life and well-being. Sufredini et al.'s mixedmethods systematic review of 51 studies found that insufficient support from social networks was associated with greater depressive and anxiety symptoms. 25 Multidimensional social ties were related to repairing or healing physical and emotional dysfunction. 27 However, there is a paucity of research on the buffering effects of PA and social ties in the association between mobility limitations with emotional dysfunction among older adults in SSA. The novelty of the current study lies in the use of older samples from Ghana to explore how mobility limitations interact with PA and social ties to influence emotional dysfunction in later life, This is important dynamic in aging epidemiology with salient public health and healthy aging implication but largely neglected in LMICs.
This study, therefore, examines the association of mobility limitations with emotional dysfunction among older adults in Ghana and estimates the potential effect modification of the association by PA and social ties. We expected that 1) mobility limitations would positively influence emotional dysfunction, 2) the effect of mobility limitations on emotional dysfunction would significantly be lower among active older adults than their inactive counterparts, and 3) the effect of mobility limitations on emotional dysfunction would be lower among older persons with stronger than those with weaker social ties.

| The survey
This study used data from the Aging, Health, Psychological Wellbeing, and Health-seeking Behavior study, undertaken in Ghana between 2016 and 2018. The methodology of the survey has been described extensively in the literature. 28 Briefly, a nationally representative sample was obtained using a multistage clustered sampling design to provide data on older adults' social relationships, health, and healthcare. 28 Samples included individuals aged ≥50 years. The random selection approach via sampling strata considered sub-regional zones, districts, and rural-urban neighborhoods. Participants were selected based on a probability-proportional-to-size sampling approach. Face-to-face interviews were conducted by

| Emotional dysfunction
The outcome variable was a composite index computed by summing up seven indicators of the mental and psychological functioning of the respondents. The items were assessed with widely used self-rated and cross-culturally validated questions, including "Over the past 4 weeks, have you been happy, sad/depressed, nervous/uneasy, restless/ fidgety/bored, hopeless, lonely, and worthless/having no value". 29 The F I G U R E 1 Flow chart of the selection of study participants. GYASI ET AL. responses were recorded on four-point Likert-like scale options: 1 = none of the time to 4 = all of the time. "Happy" was reversely coded to ensure comparability. The overall score ranged from 7 to 28, with higher scores indicating higher emotional dysfunction (α = 0.89).

| Mobility limitations
We assessed the key exposure variable with seven functional impairments and mobility-related deficiency-based items from the Medical Outcomes Study Short Form-36 (MOS SF-36) Scale. The respondents were asked seven questions on the extent to which they undertake vigorous activities such as weeding, lifting heavy objects, etc., moderate activities, such as moving a table/chair, etc., climbing about several flights of stairs, etc. 29 Each item was scored on a scale from 1 = not limited at all to 4 = much limited. We generated a continuous count of mobility limitations (range 7-28). Higher scores represented greater mobility limitations, with each additional point reflecting additional mobility limitations with Cronbach's α = 0.83.

| Physical activity (PA)
We assessed PA with the International Physical Activity Questionnaire short form (IPAQ-SF), measuring the past week's dimensions of PA intensity (low, moderate, and vigorous). The items include: "During the last 7 days, on how many days 1) did you walk for at least 10 min at a time, including walking at work, home, and to travel from place to place? 2) did you do moderate PA like gardening, cleaning, bicycling at a regular pace, swimming, or other fitness activities? 3) did you do vigorous PA like heavy lifting, digging, gardening, construction work, chopping wood, aerobics, jogging/running, or fast bicycling?" The options were recorded on a continuous scale expressed as activity events per week, with higher scores indicating higher levels of PA.
The IPAQ has been validated in older Africans with good reliability and validity. 30

| Social ties
We assessed social ties with three-item screeners, which considered elements such as social contacts, social participation, and social bonding among older adults. We asked respondents: "What is the frequency of contact with your family/close friends in the past 30 days?", "How often in the last 30 days have you attended social activities (e.g., family meetings, religious services, social/club meetings, etc.)?", "Do you feel there is someone you can share your most private concerns and fears with?". 28 Responses for each item were recorded on a 5-point Likert scale. The options for the first two items were 1 = almost every day to 5 = Never, which were reverse coded.
The responses for the third item were scaled from 1 = Completely false to 5 = Completely true. We standardized and computed ST composite score, with higher scores reflecting higher levels of social ties (Cronbach's α = 0.88).

| Analytic strategy
The significance level was set at p < 0.05, and all statistical analyses were performed using SPSS 25.0 (IBM SPSS Inc., Chicago, USA). Data on demographic characteristics were presented as means for continuous and percentages for categorical variables. Four-level hierarchical OLS models comprising unadjusted and multivariable analyses were constructed to examine the association between mobility limitations and emotional dysfunction outcomes. We first fitted an unadjusted model to estimate the specific association of mobility limitations and covariates with emotional health. All variables that showed significant associations with emotional dysfunction were entered into the subsequent multivariable analysis. Regarding the multivariable analyses, step 1 included the potential confounders (Model 1). Model 2 added mobility limitations to assess its unique contribution to emotional dysfunction. Model three included the interaction terms (mobility limitations � PA; mobility limitations � social ties) to evaluate their potential effect modifications in the mobility limitations-emotional dysfunction association. We tested the conditional effects for a simple slope at one standard deviation above and below the mean when significant interaction effects were achieved.
In a sensitivity analysis, we performed separate unadjusted and adjusted OLS estimations for each specific emotional dysfunction component to investigate the relative impact of mobility limitations on each emotional dysfunction sub-type. Multicollinearity was assessed by calculating each model's variance inflation factor (VIF) and tolerance, ensuring this was within satisfactory ranges. Scale data were checked for normality with the Kolmogorov-Smirnov test, p > 0.05. Skewed and non-normally distributed variables were logtransformed for analytic purposes.

| Sample characteristics
The mean age of the sample was approximately 66 years, the majority were women (63%), and more than one-half lived in urban areas, while some 43% were married (Table 1). Over 86% had up to primarylevel education, and the mean income was ¢308.

| Regression analysis
Unadjusted OLS regressions showed that all the control variables were significantly associated with poor emotional health ( Table 2). Table 3

| Additional analysis
In sensitivity analysis, mobility limitations significantly predicted increase in sadness, nervousness, restlessness, and hopelessness after full adjustment (Table 4).

| Main findings
This study of community-dwelling older adults in Ghana employed exploratory survey data to examine the association between mobility limitations and emotional dysfunction outcomes and evaluated the modifying roles of PA and social ties in this association. Our results T A B L E 1 Descriptive statistics of the sample.  indicated that higher levels of mobility limitations independently contributed to worse emotional dysfunction. This association remained significant when controlling for potential confounders.
However, PA and social ties significantly moderated this association.
Thus, older adults who were physically active and those who were socially connected were less likely to report emotional distress in the context of mobility limitations. These results disentangle the complex and under-explored interrelationships between these variables with important policy and public health implications, particularly in LMICs.

| Interpretations
To the best of our knowledge, this is the first study exploring the respective interactive effects of PA and social ties on the relationship between mobility limitations and emotional dysfunction among older adults, particularly in LMICs. Our findings supported the first hypothesis of the study that mobility limitations increased the risk of emotional dysfunction in old age. Although the current research is distinct, various studies have examined direct associations between mobility limitations and emotional dysfunction, particularly in many Western and high-income Asian countries. For example, using longitudinal data, Weinberger et al. 31 reported that mobility constraints increased depression among older Americans.
Similarly, in a prospective cohort and population-based study in Taiwan, Lee et al. 32 found that mobility limitations lead to depression and that independent predictors of improvement in depression over a 4-year follow-up period are fewer mobility and more social support.
In a study focused on 2400 older Swedish, Chiatti et al. 33 showed that walking at least 12 km daily and being socially engaged were asso-     and social events, promote psychological well-being while dampening worse emotional dysfunction. 23 Social networks are particularly beneficial to those with emotional dysfunction by mitigating the negative association with physical function. Low social support and strong demand for connection are related to depression in later life, 49 and social support may improve well-being. 32 Moreover, social ties offer psychological resources that may facilitate one's ability to cope with stress and encompass opportunities for venting feelings (emotional support). Significant interpersonal relations may reduce psychological dysfunction and, in turn, enhance endocrine and immune functioning. 8,32 These improvements were attributed to shared life events, social support, reciprocal appreciation, trust, and a sense of belonging through common social activities.

| Strengths and limitations
The strengths of the current study include the use of a large and

| CONCLUSIONS AND IMPLICATIONS
This representative study builds on the previously published literature suggesting that physical limitations are a risk factor for older persons' emotional dysfunction but PA and social ties may moderate this association. Interventions to address functional decline-induced emotional dysfunction in old age should consider promoting regular PA and strong social ties. These interventions should consider sociocultural mechanisms, such as the creation of parks with easily accessible exercise grounds in the community and the promotion of home-based exercise routines. 50

CONFLICT OF INTEREST STATEMENT
None.

DATA AVAILABILITY STATEMENT
The data underlying this article are available from the corresponding author on reasonable request.