Depressive symptoms in older adults: the role of physical activity and social support

Abstract Objective To evaluate the role of social support in the association between physical activity (PA) in its various domains and depressive symptoms in older adults. Methods This was a cross-sectional study involving 399 older adults seen at basic health units in the municipality of Recife, state of Pernambuco, northeastern Brazil. Linear regression was implemented using the forward method to assess the association between PA and its domains and depressive symptoms, as well as to test the possible moderating component of social support in this association. Results Of the total sample, 17.6% reported not practicing PA; 6.7% presented depressive symptoms, with a mean score of 1.59 points. A total of 18.6% of the older adults assessed reported not having social support. Older adults who did not practice total, transportation or domestic PA showed increased scores of depressive symptoms when compared to older adults who practiced PA in these domains. Older adults with social support also had fewer depressive symptoms. Practicing domestic, leisure and total PA, and/or having social support, decreased depressive symptoms in older adults. Having social support or practicing PA decreased depressive symptoms in older adults in a similar way. Conclusions Older adults with social support and who practiced PA, especially domestic and total PA, presented the lowest scores of depressive symptoms. The data showed the importance of providing elements to strengthen social support and PA practice in older adults living in poor communities as a strategy to prevent and reduce depressive symptoms.


Introduction
Population aging has been growing at an accelerated rate in recent decades, due to changes in demographic structures, low fertility, and increased life expectancy.
According to the World Health Organization (WHO), the world population over 60 years of age will double from 11 to 22% between 2000 and 2050. It is estimated that by 2020 there will be 200 million people over the age of 60 in Latin America, and that number should rise to 310 million in 2050. 1 By 2025, Brazil will be the sixth country with the highest number of older adults in the world, 2 and in the year 2060, the state of Pernambuco will have a population of 3,168,161 older adults, accounting for 31.86% of the northeastern population. 3 Even though aging is not synonymous with illness, there is an evident increase in the occurrence of chronic non-communicable diseases among older adults, even with the emergence of different public health policies that encourage healthy aging strategies. 4 The WHO estimates that 47 million of older adults have dementia or cognitive impairment, of which 60% live in developing countries such as Brazil. 5 Depressive disorders are mood disorders that cause a persistent feeling of sadness and loss of interest. 6 As important as seeking treatment for depression in older adults is to identify the factors that precede the chronic phase of the disease. 7 Depression tracking carried out through surveying depressive symptoms has been widely used in epidemiological studies. 8 The prevalence rates of this condition vary from 20-56% in outpatients, through 23-42% in hospitalized older adults, to .6% in community older adults. 9 Physical activity (PA) seems to be an effective tool in treating and preventing depression and its symptoms. 10 PA programs for older adults substantially reduce the severity of depressive symptoms, 11 tend to be well tolerated, 12 and increase physical fitness, in addition to being among the treatment options preferred by patients. 13  neurobiological, psychological and social. 15 In the neurobiological mechanism, PA seems to increase brainderived neurotrophic factor (BDNF), noradrenaline and serotonin (hormones responsible for well-being), leading to improvement in depression. 16 From the psychological point of view, PA is able to improve negative selfassessment, low self-esteem and low self-confidence among older adults. 17 Finally, the third mechanism by which PA acts on depression and related symptoms is the social mechanism. However, evidence supporting this mechanism is still not well-stablished. Some authors suggest that PA performed in groups promotes connection between participants, increasing the social support network and thereby improving depressive symptoms. 18,19 In contrast, other studies reported no differences in depression improvement when comparing exercises performed individually at home vs. exercises performed in groups in sports centers. 20 Understanding the relationship between PA and depressive symptoms in older adults, as well as the role of social support in this association, and taking into account cultural and regional differences can contribute to planning care policies aimed at healthy aging, therefore helping in the treatment and prevention of depression in older adults. Thus, the objective of the present study was to assess the role of social support in the association between PA and depressive symptoms in older adults seen at basic health units in the municipality of Recife, state of Pernambuco, northeastern Brazil.

Methods
This was a cross-sectional study, part of a broader  The sample size was established considering the interest in estimating the prevalence of several health variables; therefore, the expected prevalence was established at 50%. In addition, the following parameters were defined: error of 4 percentage points, 95% confidence intervals (95%CI), and 80% power.
Considering these factors, sample size calculation resulted in a minimum number of 597 participants using a sample design effect (deff = 1.5). Then, it was decided that the minimum sample should be inflated by an additional 30% to ensure sufficient statistical power for the analyses and to deal with possible refusals.
Respecting these parameters, the sample size to be Data were collected between the months of October and November 2017. A total of 399 older adults were monitored due to unforeseen events which occurred during the collection. Therefore, a posteriori sample calculation was carried out for this study, and it was possible to detect a significant risk ratio greater than 1.5, with 95%CI and statistical power greater than 80%. criteria were used to calculate a score: answers "no" were assigned a value of 0; a value of 1 was assigned to answers "yes" to the first two questions and to answers "sometimes" in the third question; finally, a value of 2 was assigned answers "yes" to the third question.
The score could vary from 0 to 4 points. Older adults without social support were considered to be those who obtained a score > 2 points.  Figure 2).

Results
A total of 397 older adults (77.3% women) living in the Family Health Strategy coverage areas in the city of Recife participated in the present study, with a mean age of 72.8±8.1 years. Most of the older adults had low education (90.4%) and received less than two minimum monthly salaries (87.4%). Also, the majority of the participants (59.1%) did not live with a partner, and 74.4% had a poor self-perception of their health (Table 1).   rather than go anywhere else. When evaluating older adults who answered "yes" to 4 or more of the previous questions, it was found that 6.7% (95%CI 4.5-9.7) of the older adults had depressive symptoms (Figure 1).
Older adults who had a poor self-perceived health presented a higher mean of depressive symptoms than older adults who had a good perception of their health.
Likewise, older adults without social support had a higher mean of depressive symptoms than older adults with social support. Older adults who did not perform domestic or total PA also had more depressive symptoms than older adults who practiced PA (Table 2).
Older adults who did not practice transportation PA    Values presented as β (95% confidence intervals). Model I: transportation PA + social support + adjustments (gender, age, self-perceived health); Model II: domestic PA + social support + adjustments (gender, age, self-perceived health); Model III: leisure PA + social support + adjustments (gender, age, self-perceived health); Model IV: total PA + social support + adjustments (gender, age, self-perceived health). R 2 of the models ranged between 0.12 and 0.16. Bold type indicates significant association (p ≤ 0.05). practice domestic PA compared to those who did. When assessing total PA, it was observed that older adults who did not practice any PA showed an increase of 0.54 (95%CI 0.24-0.83) points in the depressive symptom score when compared to older adults who practiced some PA (Table 3).
We also tested an interaction between social support and the different PA domains. There was no interaction between transportation PA and social support (p = 0.945); conversely, interactions between social support and total PA as well as between social support and domestic and leisure PA were observed.

Discussion
The present study showed that older adults who did not practice PA had higher depressive symptom scores when compared to those who practiced PA. In general, older adults without social support and who did not practice PA presented the highest depressive symptom scores, while practicing PA and/or having social support were associated with the lowest depressive symptom scores in this older adult group. Thus, practicing PA and/or having social support were associated with lower depressive symptom scores in older adults.
The association between PA and its domains and depressive symptoms in older adults here reported supports the findings of a previous meta-analysis 27 that suggested an inverse relationship between PA levels and morbidity and mortality from all causes, with stronger associations for domestic PA and leisure PA. In a study conducted in southern Brazil, physically active older adults were 68% less likely to experience depressive symptoms when compared to those who were insufficiently active. 28  Conversely, low social support was associated with higher depressive symptom scores. Older adults who are unable to adapt to different stress factors such as unfavorable environments, lack of social and family support and scarce social resources tend to have more depressive symptoms. 30 In contrast, participation in  Some limitations of this study must be mentioned.
The first one is the cross-sectional nature of the study, which does not enable to determine the causality between PA, depressive symptoms and social support.
Therefore, longitudinal and intervention studies are needed to confirm the direction of these associations.
Also, data on depressive symptoms, social support and mainly PA were obtained by self-reported measures, which may lead to memory bias. However, prior and standardized training was offered to recruited interviewers to minimize this potential source of error, and most of them were used to conducting academicscientific activities and carrying out fieldwork.

Conclusion
The present data suggest that practicing PA and/ or having social support are associated with lower depressive symptom scores, evidencing the importance of providing elements to strengthen social support and PA practice in older adults who belong to vulnerable communities.

Disclosure
No conflicts of interest declared concerning the publication of this article.