INTRODUCTION
Falls are a primary geriatric syndrome, and the risk increases with age.1) The World Health Organization has suggested four categories of risk factors for falls: (1) biological causes (diseases of the musculoskeletal system, central nervous system, and sensory deprivation, among others), (2) behavioural (lack of exercise, alcohol consumption, inappropriate footwear), (3) environmental (insufficient lighting, slippery or uneven floors); and (4) socioeconomic (such as limited access to health and social services, low personal income, and poor community resources, including lack of social interaction).1) In a literature review on consequences of falls, only 5% of the total references corresponded to socioeconomic factors, and only 1% corresponded to a lack of social interaction.2) Thus, more evidence needs to be reported on social factors and interactions as risk factors for falls.
The social aspects potentially related to falls are social participation and social support. Investigators distinguish social participation that involves activities with other people (for example, participation in informal and organized social activities) from those activities undertaken for the benefit of others (for example, volunteering and civil commitment).3) Social participation has been described as one of the critical pillars of ‘active aging’.4) However, few studies have examined its relationship with falls. In one study, falls also caused a decrease in social participation, with physical frailty being a factor that exacerbated this effect.5) A recent study found an increased risk for falls among people over 55 with less social participation.6)
Several studies link social support with fewer falls. One study found that solid family networks were associated with decreased fall risk. Conversely, weaker friendship networks tend to decrease the risk of falls among older women.7) Other studies found that social isolation was a predictor of falls.8-11) A recent systematic review emphasized the importance of the relationship between falls and social isolation, loneliness, and living alone among older individuals.12)
Social frailty is insufficient social participation (or no participation) and the perception of a lack of contacts and support.13) Further, social frailty could be considered a lack of resources to satisfy basic social needs essential to function correctly and have physical and mental well-being.14) Worldwide, in community settings, the pooled prevalence of social frailty was reported as 18.8%, but it varies between 13.4% and 32.3% depending on the used social frailty measures.15) To our knowledge, few studies have linked social frailty to falls. Two recent studies found an association between social frailty and risk for falls.16,17)
In Colombia, about 30% of persons aged 60+ had a fall within the previous year.18) This study aimed to assess the association between social participation, children’s support, relationship with children, and social frailty with falls among older adults living in the community from Colombia. We will control for common fall risk factors identified in previous population-based studies including community-dwelling older adults. These include age, sex, vision problems, psychotropics consumption, multimorbidity, memory impairment, functional impairment, fear of falling, and environmental barriers.18-22) We hypothesize that older adults who fall more have less social participation, children’s support, and a poor relationship with children but greater social frailty after adjusting for other fall risk factors.
MATERIALS AND METHODS
Population and Design
The 2015 Survey on Health, Well-being, and Aging (SABE) in Colombia was a cross-sectional study that collected information on the health, well-being, aging, and quality of life of adults 60 or older living in communities across the country. The SABE Colombia is an extension of other SABE aging studies in Latin America and the Caribbean that defined older adults as persons aged 60 years and over, also based on the World Health Organization starting age for aging.23,24) Data was collected from April to September 2015 using in-person surveys. It employed a probabilistic, cluster, stratified, and multistage sampling design. Response rates were 62% to 77% (urban and rural, respectively); 23,694 older adults answered a structured questionnaire.25) In the original study, at the beginning of the interview, individuals with a total score <13 in the Folstein Mini-Mental State Examination26) were identified as having cognitive impairment. A proxy interview was conducted for these individuals in most of the survey. Participants whose proxies provided answers (n=4,690) were excluded from our analysis since they had cognitive impairment and were not allowed to answer specific questions related to the components of social frailty––one item taken from the Geriatric Depression Scale (GDS)27) and frequency of leaving home question. Therefore, the sample became 19,004. Furthermore, excluding those who reported not having children (n=1,317), and our sample for analysis was reduced to 17,687 (Fig. 1). The ethics committee approved the participants’ informed consent in the original study.25) Approval number 002-2021 was also obtained from the University of Valle Ethics Committee in 2021 for the secondary data analysis. This study complied the ethical guidelines for authorship and publishing in the Annals of Geriatric Medicine and Research.28)
Study Variables
Dependent variable
The dependent variable for this study was past-year fall, according to a yes answer to the following question: “In the past year, have you fallen to the ground?” Falls (code=1) vs. no-falls (code=0) categories were determined.
Main independent variables
Social participation questions were related to the following (where higher scores indicate higher participation):
(1) Participating in groups, such as religious, sports, political, cultural, ecological, union, ethnic, or health. Total score 0–10.
(2) Supporting others such as helping at home, giving advice, helping with personal care (e.g., bathing, dressing, etc.), financial support (e.g., money or in-kind donations), emotional support (e.g., affective, company), or help with caring for grandchildren. Total score 0–6.
(3) Volunteering in groups such as social welfare services, senior centers, children’s home centers, schools or universities, health care providers, churches, or temples. Total score 0–6.
The questions to assess social frailty were adapted from the study of Makizako et al.,29) and measured as visiting family or friends frequently (no), leaving home infrequently (≤2 times a week) (yes), living alone (yes), talking to friends or family frequently (no), and feeling helpful to family (no) (Table 1). The last question is from the GDS.27) Total score 0–5 and categories were no-frail (0), prefrail (1), and socially frail (≥2).29)
Children’s support was measured by older individuals’ perceptions of the support received by their children––if they have children––as follows: “Regardless of whether or not they live with you, do they (he or she) provide you...?” The responses were categorized into helping with daily activities, providing food, clothing or transportation, cash, company, affection, or nothing, and the answers were yes/no.
In addition, the quality of the children’s relationship was assessed with the following question: “How is your relationship with them (he or she)?” The responses were categorized as very good/good, fair, bad/very bad, and do not know/does not respond, which were regrouped into two categories: unsatisfactory relationship (regular, bad, or very bad) and good relationship (very good or good).
Covariates
We chose other risk factors for falls among older adults in the community as covariates available in the database.1,18-20,22) Age (year) and sex (female or male self-reported) were included. Place of residence includes city, town, or rural village. Multimorbidity was assessed by self-report of being told by a health professional about having any of these medical conditions: hypertension, diabetes, heart disease, arthritis, cancer, chronic obstructive pulmonary disease, stroke, and osteoporosis (counted from 0 to 8). It was defined as having two or more medical conditions.18)
Functionality was measured by the Lawton scale that assesses independence for instrumental activities of daily living (IADLs), including using the telephone, taking medications, managing finances, preparing meals, shopping, and using transportation. Scores ranged from 0 to 6; lower scores meant decreased functional status. Low IADL scores were defined from 0 to 4.30) Poor vision was defined when a person reported having regular, bad, or very bad vision (vs. good or very good vision), either near or far vision, regardless of whether wearing glasses. Psychotropic medication intake was defined as participants taking sleeping medications, tranquilizers, or sedatives during the past 30 days. Poor memory was defined by self-rated memory reported as fair or poor (vs. excellent, very good, good).18) Environmental factors include (1) sidewalks with obstacles or irregular surfaces; (2) trash or debris in the sidewalks or gardens; and (3) low or no street illumination. Finally, fear of falling was used as an ordinal variable (none, little, regular, or a lot; score 1–4).
Statistical Analysis
Proportions (%) were estimated for categorical variables, and continuous variables were expressed as mean±standard deviation. Bivariate comparisons between falls and no-falls categories were tested by categories of social participation, social support, social frailty, and other covariates using the chi-square test. Multivariate logistic regression analyses determined the association between falls and independent variables. Thus, the odds ratios (ORs) with their respective 95% confidence intervals (CIs) were obtained. Due to collinearity, social frailty had a separate multivariate analysis model from social participation and social support. Models were checked using the Hosmer–Lemeshow procedure. Additional analyses were made to explore why social participation was associated with higher odds of falls. We used Pearson correlation coefficients (r) and ANOVA to test differences in social participation––participating in groups (0–10), supporting others (0–6) or volunteering (0–6)––across environmental barriers (0–3) and site of residence (city, town, or rural village), respectively. All analyses were performed in the statistical program SAS version 9.4 for Windows (SAS Institute Inc., Cary, NC, USA); the level of statistical significance selected was p≤0.05 for the two-tailed test.
RESULTS
Participants were 17,687 older individuals with a mean age of 69.3±7.2 years; 56.2% were women, and 29.6% reported having a fall in the past year. Social frailty was present in 36.6%. A total of 45.1% of the individuals participated in at least one group, which was most frequently a religious group (32.6%); 92.3% gave help or support to at least one other, most frequently by giving advice (78.6%); 16.2% participated in at least one volunteer group, most frequently a church or temple group (10.5%) (Table 2).
Older adults perceived children’s support mainly in company and affection (76.7%), followed by other forms of support, such as financial aid and other aid, including food, clothing, or transportation, at much lower percentages (39% and 38.5%, respectively). Regarding the quality of their relationship with their children, most participants described it as good (very good or good) at 83.7% (Table 2).
In bivariate analyses, falls were significantly associated with being older, being a woman, having multimorbidity, a low Lawton score, poor vision, poor memory, fear of falling, more environmental barriers, taking psychotropics, or residing in a city. Questions used in the construction of social frailty that were associated with falls were visiting family or friends frequently (no), leaving home infrequently (yes), and feeling helpful to family (no). Social frailty as a category (score ≥2) was associated with more falls. Concerning social participation, older people who participated in more groups, helped others or volunteered had more falls. In contrast, those who received greater children’s support had fewer falls. Regarding the quality of perceived relationships with their children, a good relationship was also associated with fewer falls (Table 3).
In multivariate logistic regression analysis, social frailty was associated with higher odds of falls (OR=1.20; 95% CI 1.10–1.32) (Model 1 in Table 4). Participating in groups (OR=1.07; 95% CI 1.03–1.11), helping others (OR=1.04; 95% CI 1.02–1.06), or volunteering (OR=1.09; 95% CI 1.01–1.17) were also associated with higher odds of falls. In contrast, receiving help, affection, and company from children was associated with lower odds of falls (OR=0.95; 95% CI 0.93–0.97) compared to not receiving those. Also, having a good relationship with children was associated with lower odds of falls (OR=0.75; 95% CI 0.66–0.85) than an unsatisfactory relationship (Model 2 in Table 4). Other factors associated with more falls were age ≥75, being a woman, having multimorbidity, poor vision, a low Lawton score, poor memory, fear of falling, more environmental barriers, taking psychotropics, or residing in a city. In sensitivity analysis, a multivariate logistic regression model showed that social frailty re-categorized as frail (=1) vs. no frail or pre-frail (=0), resulted also associated with more falls (OR=1.16; 95% CI 1.08–1.25; p<0.0001).
In additional analyses exploring why social participation was unexpectedly associated with higher odds for falls, we found that social participation measures (participation in groups, supporting others, or volunteering) were significantly and positively correlated with a higher number of environmental barriers (Pearson r=0.10, p<0.0001; r=0.14, p<0.0001; and r=0.09, p<0.0001; respectively). We also found that city residents have higher frequencies of falls compared to those in the towns or rural villages (Table 3). A the same time, the means for social participation measures were higher in the cities (participation in groups at the cities 0.7±0.9, towns 0.6±0.9, and rural villages 0.5±0.8, F-value 46.81, p<0.0001; supporting others at the cities 3.2±1.6, towns 3.1±1.7, and rural villages 2.8±1.7, F-value 108.35, p<0.0001). In addition, given that religious groups have the greater percentage of participation in the groups (32.6%), we found that participating in religious groups (vs. not) was associated with being women, having multimorbidity, poor vision, fear of falling, and a greater number of environmental barriers (chi-square tests all p<0.001); however there was no association with functional status (IADL Lawton score). We also used only religious groups instead of participation in groups variable in a multivariate analysis, similar to Model 2 in Table 4, adjusting for all covariates, and found that religious group participants have higher odds for falls (OR=1.14; 95% CI 1.06–1.23; p=0.0004). Therefore, these findings may partly explain why social participation is associated with higher fall odds.
DISCUSSION
In this nationwide study in Colombia, we found that social frailty and participation in social groups or helping others were associated with higher odds of falls in older adults. By contrast, having a good relationship with children or receiving help, affection, and company from them were associated with lower odds of falls.
Our finding that social frailty was related to an elevated risk of falls agrees with previous studies that have linked social frailty with adverse outcomes in older individuals. These included a higher disability,29,31) mortality,32) reduction of cognitive and physical stimulation, accelerating the decline of cognitive and physical functioning.33,34) Social frailty is also associated with a reduction of physical and cognitive stimulation and a decline in both physical (grip strength and walking speed) and cognitive function (memory, attention, executive function, and processing speed).13,29) Our finding also agrees with a few recent studies that found a relationship between social frailty and falls.16,17) Thus, social frailty may represent the accumulation of stress among frail older adults,35) and frail individuals tend to fall more than those not frail.18)
Falls increase physical frailty in the person by decreasing physical resources; this reduces the ability to continue social activities, exacerbated by the increased likelihood of repeated falls and traumatic consequences.36) It also suggests that falls can impact relatives’ lives by increasing the need for their support. Some qualitative studies have suggested that falls can cause fear, anxiety, and feelings of helplessness among older individuals’ relatives, mainly spouses and children.37) Physical frailty and falls were very close constructions in that study. They shared similar risk factors, such as mobility disorders, and had similar consequences in terms of disability, mortality, and social participation.5) Other studies have also described an association between physical frailty components, such as grip strength and exhaustion, with falls.38)
In our study, older people participating in more groups, helping others, or volunteering experienced more falls. These results differ from the hypothesis regarding social participation, in which, according to previous studies, there is a relationship between more falls and social isolation or less social participation.5,6)
Additional findings may explain why greater social participation is associated with higher odds of falls in our study. Indeed, we found that social participation measures (participating in groups, supporting others, or volunteering) were significantly and positively correlated with a higher number of environmental barriers. We also found that participants residing in the cities have higher frequencies of falls and, at the same time, more social participation compared to those in the towns or rural villages. When we focused on the religious groups, who were the more prominent participants in the groups, we found that they had common risk factors for falls, such as being a woman, having multimorbidity, poor vision, fear of falling,18-22,39,40) and, especially, greater exposure to environmental barriers.41)
Thus, the odds of falls could increase in people with more opportunities to be exposed to risky scenarios outside the home, especially on the streets and sidewalks.41) Furthermore, well-known extrinsic factors predispose individuals to falls, including poor lighting, uneven surfaces, and slippery floors, causing 30% to 50% of falls in the older population.42) In Colombia, urban areas are not designed with the capabilities of older individuals in mind, so it is reasonable to think that older adults who are more exposed to environmental barriers could have a greater probability of falls.18,25)
This study serves as a basis to justify the importance of developing community-specific, person-centered prevention programs and age-friendly cities related to the objectives of the United Nations 2030 agenda to improve the safety and well-being of older adults.43) In addition, the initiative of friendly communities and cities of the World Health Organization is already beginning to bear some fruits, however in Colombia so far only three cities have started this project.44)
We propose raising awareness about the risk of falls in older adults when leaving home so people would be more attentive to the needs that an older adult may require when walking on the street, crossing a sidewalk or climbing a staircase. Moreover, people should have a proactive and friendly attitude towards older adults, which could prevent a fall.45) From the collective point of view, it is necessary to promote greater participation in policies concerning the well-being and safety of older adults and adaptation of the infrastructure in public spaces such as parks and road networks to reach universal accessibility with more confidence to walk and run without physical obstacles or increase safe and accessible mobility options for them. In addition, all schools should teach about the great value of the older person in society and the family to promote behaviors and attitudes of respect and inclusion towards older people. By contrast, ageism is associated with an increased risk of falling in older adults.45)
On the other hand, older individuals receiving greater support from their children and perceiving the relationship quality as good had fewer falls. Previous studies have suggested that perceived social support could protect against falls.46) Some possible explanations are that older adults with more children’s support may be more attentive to environmental dangers because their children encourage them to do so. They may also receive help from children to complete risky tasks.47) In this sense, our data favor that good children’s support is a protective factor against falls.
This study has some limitations. Due to its design, it was impossible to demonstrate the causality of the relationship between social participation, children’s support, and social frailty with falls. The possibility of confounding variables in our sample is high, so the data were adjusted to minimize these effects. Another limitation is the exclusion of participants who needed help from a person by proxy since they had a low cognitive status; however, the probability sampling methodology allows the rest of the sample to represent the older Colombian population living in the community.25)
Among the strengths is that this is the first study that jointly evaluates social support from children, social participation, and social frailty, concepts that are constantly growing and for which there is increasing interest worldwide. Social aspects are postulated as pillars of active aging.4) This study lays the foundation for continuing to investigate social aspects and their relationship with falls to encourage the inclusion of social aspects in public policies that reduce falls in older adults. There are already established guidelines to improve the well-being and safety of older adults. However there is still much to work on regarding public policies and education.
In conclusion, our findings show an association between several social aspects and falls. Older adults who are socially frail had more falls. Those participating in more groups, helping others, or volunteering had more falls. These findings are partially explained because more group participants reside in cities where they are more exposed to environmental barriers. In contrast, those with greater social support or good relationships with their children had fewer falls. Given our findings, both social environments, represented by social frailty status and family relationship or support, are important additional factors to consider in preventing or improving fall status in older adults.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
