Health-Promoting Behaviors and their Associations with Frailty, Depression, and Social Support in Thai Community-Dwelling Older Adults: A Cross-Sectional Analysis
Article information
Abstract
Background
As the global population ages, including Thailand, health-promoting behaviors (HPBs) have emerged as critical challenges affecting the well-being of community-dwelling older adults. This situation calls for a deeper understanding of the modifiable protective factors involved. The current study aimed to examine the associations between frailty, depression, social support, health literacy (HL), and HPBs among community-dwelling older adults in Thailand.
Methods
A cross-sectional study with a convenience sample of 250 older adults was employed. We collected using validated questionnaires—including Tilburg Frailty Indicator, Thai Geriatric Depression Scale, Social Support Scale, Health Literacy Scale, and Health Promoting Behaviors Scale. We analyzed using descriptive statistics and stepwise multiple linear regression.
Results
This study involved a total of 250 community-dwelling older adults, with a mean age of 70.41 years. The participants had a mean frailty score of 6.78±1.95, indicating that most were classified as frail (score ≥5) based on the Tilburg Frailty Indicator. They also reported no significant depressive symptoms (3.80±2.62), moderate levels of social support (36.70±4.20), poor health literacy (33.15±2.71), and moderate health-promoting behaviors (101.19±7.67). Health literacy, social support, frailty, depression, and comorbidity conditions were significant predictors of health-promoting behaviors, collectively explaining 80.5% of the variance.
Conclusion
These findings emphasize that improved HL, greater social support, reduced frailty, and lower depression scores were associated with healthier behaviors in older adults. Multidisciplinary healthcare teams should consider these factors when designing their intervention strategies to gain a more comprehensive understanding and improve health outcomes.
INTRODUCTION
Global population aging population represents the most prominent demographic pattern of this 21st century. By 2050, the United Nations1) predicts that people older than 65 years will constitute 16% of global population numbers, which will represent twice the numbers from 2019. Consequently, the substantial changes in population demographics create significant barriers for all healthcare and social programs as well as public health programs.2) As life expectancy increases, ensuring the health, independence, and quality of life of older adults is a key public health concern.3)
Frailty is increasingly recognized in both research and clinical settings as a key concern in aging populations. It is a complex condition characterized by decreased physiological reserve and reduced ability to cope with stressors, increasing the risk of hospitalization, disability, falls, and mortality among older adults.4) Unlike discrete medical diagnoses, frailty reflects a state of cumulative vulnerability that progresses over time but may be mitigated with timely and appropriate interventions.5) Evidence suggests that frailty is a dynamic condition, influenced by personal, social, and environmental factors, and is not an inevitable consequence of aging.5)
Depression, another prevalent condition among older adults, has been shown to exacerbate frailty and accelerate health decline, affecting both mental and physical domains.6) Community-based studies estimate the prevalence of depression in older adults to range from 10% to 25%.7) Depression intensifies frailty while decreasing physical movement and motivation, and makes it harder for people to follow disease-prevention practices, thus deepening health complications.8) Despite growing interest in the intersection of frailty and depression, research examining their combined effects within community settings remains limited.
Furthermore, research has confirmed that social support functions as an essential protective factor that protects aging populations from negative health outcomes. According to the convoy model of social relationships, people who sustain supportive networks throughout their lives receive protection against stress while promoting their overall health.9,10) Empirical evidence indicates that strong social networks lead to decreased frailty combined with better psychological well-being and higher participation in health-promoting behaviors (HPBs) based on existing empirical studies.11,12) Recent studies further highlight how community-based social support influences cognitive functioning and psychological distress.13,14) However, these protective benefits are increasingly threatened by trends such as urbanization, migration, and social isolation.
In addition, health literacy (HL) has also emerged as a key determinant of health, influencing individuals’ capacity to manage their health across the lifespan. HL is defined as a set of cognitive and social skills that determine individuals’ motivation and ability to access, understand, and use information to make informed health decisions. This capability substantially influences their ability to navigate the healthcare system, adhere to prescribed treatments, and accept preventive health measures.15) People who demonstrate higher HL skills achieve better chronic disease self-management, which leads to more active health behavior practice and enhanced quality of life.16-18) Conversely, HL that is insufficient leads patients to experience worse medical outcomes together with higher admission rates and wider health inequalities.19)
Several studies suggest that HL may mediate the relationship between frailty and depression among older adults. Uemura et al.20) and Shin et al.21) reported that limited HL is associated with higher frailty, while Liu et al.14) found that HL mediates the association between social support, depression, and frailty in adults with chronic diseases. Additionally, mental HL—a subset of HL—has been shown to promote protective health behaviors among older adults.22)
HPBs such as physical exercise, nutritious eating, disease prevention practices, medication adherence, and stress control steps to sustain their personal independence and life quality.23) However, participation in these behaviors often declines due to health limitations, financial constraints, emotional challenges, and a lack of health information.24,25) Notably, studies show that proper HL combined with social support provides substantial benefits to older adults’ participation in HPBs, including among people who have frailty or chronic diseases.17,26)
Few studies have explored these variables collectively, especially in Southeast Asian contexts, where cultural and healthcare system differences may influence aging experiences and behaviors. For instance, a prior study17) identified perceived self-efficacy, HL, access to COVID-19 preventive resources, and social networks as critical determinants impacting HPBs among older adults in Thailand. Despite the growing acknowledgment of the interconnectedness of these factors, much current literature tends to examine them either independently or in narrowly defined groupings. In addressing this gap, the present study aims to synthesize frailty, depression, social support, and HL into a comprehensive model to predict better HPBs among older adults residing in Thailand. The insights gained from this culturally relevant research can inform the development of targeted interventions designed to promote healthy aging. Ultimately, this study aims to improve the well-being of community-dwelling older adults by gaining a deeper understanding of the dynamics at play among these critical health determinants.
MATERIALS AND METHODS
Research Design and Sampling
This cross-sectional study was carried out among older adults living in the community of Mahasarakham Province, situated in northeastern Thailand. Mahasarakham is a northeastern province characterized by predominantly rural communities, limited access to specialized healthcare, and strong communal and familial support structures, which may influence health behaviors and social dynamics. Therefore, findings may not be fully generalizable to urban or culturally distinct regions of Thailand. Participants were selected through convenience sampling, following these inclusion criteria: (1) aged 60 years or older of both genders; (2) residing within the community; (3) having no history of cognitive dysfunction, Alzheimer’s disease, or psychiatric conditions; and (4) capable of communicating in and understanding the Thai language. However, individuals with other visual, auditory, or sensory impairments, those diagnosed with a psychiatric disorder, and those unwilling to voluntarily take part were not included in the study. The convenience sampling approach was utilized, and the sample size was determined using the G*Power program.27) The effect size from a previous study was 0.15,28) with an alpha level set at 0.05 and a power of 0.8. As a result, we raised the sample size by around 30% to accommodate possible incomplete answers to the questionnaires. Thus, the total sample size for this study reached 250. While the convenience sampling approach allowed for a diverse sample within the local community, it also raises concerns regarding the generalizability of the findings, as they may not be representative of the broader older adult population within Thailand.
Research Instruments
In this study, a comprehensive suite of six research instruments was utilized to facilitate the investigation. Prior to their application, permission was secured from the respective owners for the use of all instruments. The Thai translations of the Tilburg Frailty Indicator (TFI), Thai Geriatric Depression Scale (TGDS), Social Support Scale (SSS), Health Literacy Scale (HLS), and Health-Promoting Behaviors Scale (HPBS) were employed, while the TFI had previously undergone translation from English to Thai, with validation conducted in the Thai context. To ensure the content validity of the questionnaires, they were rigorously reviewed by three subject-matter experts. Additionally, a pilot study was carried out involving 30 older adults who shared characteristics with the target population, enabling an assessment of the reliability of the questionnaires.
The Tilburg Frailty Indicator
TFI was used to assess frailty, distinguishing between various domains: physical, psychological, and social. Developed by a previous research team,29) this instrument consisted of 15 items related to frailty. The scoring system for the different domains is as follows: a score of 4 or higher on the eight items related to physical frailty indicates that an individual is physically frail; a score of 2 or higher on the four psychological frailty items signifies psychological frailty; and a score of 2 or higher on the three social frailty items denotes social frailty. To assess overall frailty, the total TFI score is calculated, with a score of 5 or higher classified as frail. The Cronbach’s alpha coefficient of the current study was 0.82.
Thai Geriatric Depression Scale
The TGDS was utilized to evaluate depressive symptoms among participants. This 15-item instrument, which was translated by a previous research team in 2013,30) has demonstrated strong psychometric properties in geriatric outpatient settings, establishing a cut-off score of greater than 5 for clinical significance. Each item is scored with one point for a positive response, with the exception of items 1, 5, 7, 11, and 15, which receive points when answered negatively. Total scores ranging from 0 to 5 are classified as indicative of normal psychological functioning, whereas scores between 6 and 15 suggest the presence of depressive symptoms. The Cronbach’s alpha coefficient of this tool was determined to be 0.80.
Social Support Scale
SSS assessed the supportive care older adults receive from their social networks, including family and friends, and their influence on health prioritization. Developed by Suksatan and Ounprasertsuk,31) “this 16-item scale uses a 4-point scale (1=strongly disagree, 4=strongly agree). Total scores range from 16 to 64, categorized as low support (14–30), moderate support (31–47), and high support (48–64),” with higher scores indicating greater social support. The Cronbach’s alpha coefficient in this study was 0.87.
Health Literacy Scale
HLS was used to evaluate HL in older adults, assessing functional, communicative, and critical HL through 14 items, which were developed by a previous research team.32) Responses were rated from 1 (never) to 4 (regular) across three categories: basic (five questions), interactive (five questions), and critical (four questions). Higher total scores (maximum 56) indicate better HL, except for basic HL, where a higher score reflects lower literacy. Total scores were categorized as follows: <33.60 (poor), 33.61–39.19 (fair), 39.20–44.79 (good), and >44.80 (very good). The study reported a Cronbach’s alpha coefficient of 0.91.
Health-Promoting Behaviors Scale
The HPBS was used to evaluate older adults’ HPBs as part of their lifestyle. We utilized the Thai version of the HPBS,33) consisting of 43 items across six domains: health responsibility (items 1–7), physical activity (items 8–14), nutrition (items 15–22), interpersonal relations (items 23–29), spiritual growth (items 30–37), and stress management (items 38–43). Responses were measured on a 4-point scale (4=strongly agree, 1=strongly disagree). The overall mean score was classified into three levels: low (<103), moderate (103–138), and high (>138). The Cronbach’s alpha coefficient was 0.89, indicating good reliability.
Sociodemographic Variables
The demographic questionnaire was composed of 12 items, including both open-ended and multiple-choice questions. These items addressed key variables such as age, body mass index (BMI), occupation, monthly income, religion, marital status, educational attainment, health insurance coverage, perceived health status, comorbid conditions, life events, and satisfaction with residence.
Data Collection
Prior to the commencement of the study, we secured approval from the director of the Human Research Ethics Committee. We then provided participants with a thorough explanation of the study's objectives and procedures, as well as the measures we would take to protect their rights throughout the process. After obtaining their written consent to participate, we guided them on how to complete the questionnaires and encouraged them to ask any questions they might have. Once the participants completed the questionnaires, we carefully reviewed them for completeness and accuracy before proceeding with the statistical analysis.
Data Analysis
All collected data were analyzed using SPSS version 26 (IBM, Armonk, NY, USA). We utilized descriptive statistics (e.g., frequency, percentage, mean, and standard deviation) to present the results of this study. Additionally, stepwise multiple regression analysis was conducted to identify predictive factors influencing HPBs among older adults. Stepwise multiple regression was selected for its utility in exploratory analysis to identify key predictors among multiple interrelated variables in the absence of a clearly established theoretical model.34) While we acknowledge its limitations, including risk of overfitting and instability, it provided a practical method for reducing model complexity in this cross-sectional design. Prior to conducting regression analysis, we assessed multicollinearity using variance inflation factors (VIFs) and tolerance values. All VIFs were <2.0, and all tolerance values were >0.5, indicating no serious multicollinearity among predictors. These assumptions (linearity, independence of errors, homoscedasticity, normality of residuals, and multicollinearity) were carefully assessed to confirm that the stepwise multiple regression analysis would provide reliable and interpretable results. A p-value <0.05 was considered statistically significant.
Ethical Statement
The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethical Committee Review Board of Mahasarakham University (No. 159-110/2021, dated May 20, 2021).
RESULTS
The study included 250 community-dwelling older adults, with a mean age of 70.41 years and a mean BMI of 22.50 kg/m2. Most participants had agriculture (40.80%) and had a monthly income below 6,000 Thai baht (58.4%). Most participants identified as Buddhist (98.4%), were married (60.8%), and had completed primary school or lower (64.0%). Regarding health coverage, 73.6% were enrolled in the Universal Health Coverage scheme. Most participants rated their health status as fair (36.8%), and 51.6% reported having one or more comorbid conditions. In terms of life events, most participants had not recently experienced major events such as the death of a loved one (66.0%), serious illness (86.4%), or crime (96.8%). Additionally, a large proportion (93.2%) reported satisfaction with their current place of residence (Table 1).
Sociodemographic status of the participants
Frailty, Depression, Social Support, HL, and HPBs
Participants demonstrated a mean total frailty score of 6.78±1.95 based on the TFI, indicating that most were classified as frail (TFI score ≥5). The subscale scores were 4.71±1.95 for physical frailty, 1.16±1.08 for psychological frailty, and 0.91±0.64 for social frailty. Depression scores indicated a normal level of depressive symptoms, with a mean score of 3.80±2.62. Social support was moderate, with a mean score of 36.70±4.20. HL was classified as poor, with a mean total score of 33.15±4.45, including subscale means of 11.44±2.71 for functional HL, 12.11±2.50 for communicative HL, and 9.58±2.23 for critical HL. Participants' overall HPBs were rated as moderate, with an average score of 101.19±7.67. The mean scores for health responsibility, physical activity, nutrition, interpersonal relations, spiritual growth, and stress management were 16.20±2.56, 15.32±2.64, 19.62±2.50, 17.20±2.90, 18.58±2.87, and 14.43±2.37, respectively (Table 2).
Results of frailty, depression, social support, health literacy, and health-promoting behaviors scores
Predictors of HPBs
Stepwise multiple regression analysis identified five significant predictors of HPBs: HL, social support, frailty, depression, and comorbid conditions. Standardized beta coefficients indicated that HL (β=0.426) was the strongest predictor, followed by social support (β=0.335), frailty (β=0.310), depression (β=0.290), and comorbid conditions (β=0.079). These variables together explained 80.5% of the variance in HPBs (Adjusted R²=0.805, p<0.001). The detailed regression results are shown in Table 3.
Regression of individual health-promoting behaviors
DISCUSSION
This study examined the interrelations between frailty levels and depression, together with social support and HL, and their impact on HPBs among Thai older adults who live in the community. The findings showed that these multiple factors enabled the explanation of substantial changes in HPBs, which demonstrates their essential functions in shaping older people's health outcomes. This section interprets the results in light of existing research, discusses possible causal mechanisms, and suggests implications for intervention and future studies.
Among all variables assessed, HL was the strongest factor associated with HPBs within this study. This finding aligns with previous research demonstrating that HL forms the foundation for effective health self-management, disease prevention, and engagement in positive health behaviors.35) Individuals with high are more likely to understand medical information, adhere to treatment regimens, and make informed health decisions, including engaging in physical activity, improving their diet, and following recommended screenings.16,19) Moreover, a prior study, along with other previous research, also supports the role of HL as a mediator linking chronic disease management and quality of life.36) Our findings emphasize the critical role that HL plays in everyday health-related decisions among older adults, especially those seeking to maintain independence.
Social support was also a strong predictor of HPBs in our study. This result supports previous findings indicating that emotional and instrumental support significantly influence health behavior adoption and maintenance in older adults. A prior study found that social relationships directly impact health actions, mental states, and functional abilities in elderly populations.37) A study in Taiwan also found that social activity and emotional support improve physical performance and reduce frailty progression.38) Moreover, a study in South Korea highlights the role of social engagement in improving psychological well-being and physical function, ultimately enhancing resilience and delaying the onset of frailty.13) Our findings reaffirm the importance of strengthening social networks and community ties to facilitate health-promoting practices.
Frailty was a significant but complex predictor of HPBs. Although frailty is often associated with reduced physical capacity,39) it does not inherently prevent older adults from engaging in health-promoting activities. Our findings suggest that frailty may not be a complete barrier to healthy behavior if appropriate support systems and tailored interventions—such as gentle exercise programs and nutritional counseling—are in place.40) Moreover, early identification of frailty and implementation of preventive strategies in community settings are crucial for preserving function and promoting active aging.
Depression also negatively influenced HPBs, consistent with prior research showing that depressive symptoms can reduce motivation, energy levels, and cognitive functioning.41) Depression interacts with frailty through a mutual enhancement process that produces continuous deterioration of mental and physical conditions.5) Therefore, the treatment of depressive symptoms in older adults needs immediate attention because it leads to improved health behavior promotion and better health outcomes. Addressing depression in older adults through psychological counseling, cognitive-behavioral therapy, and social programs can improve both mental health and engagement in health-promoting activities.42)
Our findings suggest that participants with comorbid conditions are more likely to engage in HPBs. This phenomenon may be attributed to the concept of a “teachable moment,” whereby individuals become motivated to manage their health following a diagnosis.43) While these results are encouraging, it is imperative to acknowledge that effective management of multiple chronic conditions necessitates coordinated care and tailored educational interventions. The positive correlation between comorbidities and HPBs may reflect enhanced health awareness and behavioral modifications prompted by health-related experiences. Furthermore, sociocultural norms in Thailand, including a strong respect for health guidance from healthcare professionals and family caregivers, may further facilitate this adaptive response.
Strengths and Limitations
This study has several noteworthy strengths. First, it presents interrelated factors influencing HPBs among community-dwelling older adults, incorporating frailty, depression, social support, and HL into a unified predictive model. Unlike previous studies that examined these variables in isolation or limited combinations, this study provides an integrated framework that enhances understanding of multidimensional health determinants in aging populations. Second, the study utilized validated, reliable measurement tools, and applied rigorous statistical methods—specifically stepwise multiple linear regression—to identify key predictors of HPBs. The relatively large sample size (n=250) enhances the generalizability of findings within similar community settings in Thailand. However, several limitations must be acknowledged. This research was carried out in a single northeastern province of Thailand, utilizing convenience sampling methods. Although the sample size was deemed sufficient for achieving statistical power, the results may not be entirely representative of older adults in other regions or the broader Thai population. To improve external validity, future studies should consider employing stratified random sampling techniques that encompass a variety of geographic locations. The cross-sectional design limits the ability to establish causal relationships among the variables. Longitudinal research is needed to explore how changes in frailty, depression, social support, and HL affect HPBs over time.
Additionally, the reliance on self-reported data may introduce biases such as recall bias and social desirability bias, which could affect the accuracy of the responses. Future studies should consider incorporating objective health assessments and longitudinal tracking to minimize these limitations. Cultural specificity is another limitation, as the sample was drawn exclusively from older adults living in one region of Thailand. The findings may not be directly generalizable to older adults in different cultural, geographic, or socioeconomic contexts. Comparative studies across various countries or regions are recommended to validate and expand upon these results. Moreover, although the model accounted for several psychosocial factors (e.g., self-control and self-awareness) and health-related factors (e.g., self-efficacy), other important variables (e.g., cognitive function, socioeconomic status, access to healthcare, and environmental supports) were not measured in this study and should be included in future research. Future research should integrate these additional determinants using ecological or systems-based approaches to develop more comprehensive intervention strategies. Finally, this study did not employ an integrated analytical framework such as path analysis or structural equation modeling, which could provide a more nuanced understanding of mediation or moderation effects among psychosocial variables. Future longitudinal studies should investigate these pathways more thoroughly better to understand the causal relationships and mechanisms underlying HPBs.
Conclusion
This study highlights the multifaceted relationships among frailty, depression, social support, HL, and HPBs in community-dwelling older adults in Thailand. HL was most strongly associated with HPBs, followed by social support, frailty, depression, and comorbid conditions. These findings underscore the importance of developing comprehensive intervention strategies to improve the health outcomes of aging populations. By identifying the key psychosocial and health-related factors influencing HPBs, this study contributes to the growing body of literature advocating for integrated health promotion models. Older adults require targeted interventions that address not only physical health but also emotional, social, and informational needs. As the global population ages, investing in such holistic and culturally sensitive programs is critical for achieving healthy aging and reducing the burden on healthcare systems. Future interventions in health promotion should focus on community-based health literacy initiatives and peer-led social support networks specifically designed for older adults. Policymakers must prioritize integrated and culturally appropriate strategies to foster healthy aging within Thai communities.
Notes
This research project was financially supported by Mahasarakham University. We would like to thank the participants for providing valuable data and the reviewers for their insightful comments and suggestions that helped improve the quality of this study.
CONFLICT OF INTEREST
The researchers claim no conflicts of interest.
FUNDING
None.
AUTHOR CONTRIBUTIONS
Conceptualization, BC, WS, DL; Data curation, BC; Funding acquisition, BC, WS, DL; Investigation, BC, WS, DL; Methodology, BC, WS, DL; Project administration, BC; Supervision, HM; Writing-original draft, BC, WS, DL, HM; Writing-review & editing, BC, WS, DL, HM.