Association between social support and health-related outcomes among older adults with type 2 diabetes mellitus in a primary care setting

INTRODUCTION In Singapore, the increasing prevalence of diabetes mellitus is disproportionally greater among older adults.[1] With an ageing population, older adults form a growing segment of our diabetic population. These older adults also have an increased risk of diminished social support from various disruptive life events. Engel's biopsychosocial model emphasises the importance of psychosocial issues and their interactions with biomedical health, hence the need to address different dimensions of illness to ensure holistic care.[2] Social support may impact self-care behaviours, mood and pain perception, and consists of functional and structural support. Functional support includes instrumental support, referring to practical aid received, and emotional support. Structural support refers to the size and density of social networks.[3] In this study, we examined functional social support as we postulated that this may impact an older adult's ability to cope with diabetes mellitus. Most studies evaluating social support in diabetes mellitus were conducted among the general adult population, whereby social support was associated with improved self-care and self-rated health.[4,5,6] Our study aimed to explore the association between functional social support and self-care behaviours such as diet, exercise, medication adherence and health-related quality of life (HRQOL) among older adults with diabetes mellitus. We hypothesised that functional support has a positive association with diabetic self-care behaviours and HRQOL. METHODS A cross-sectional study was conducted in a National Healthcare Group (NHG) polyclinic between July 2021 and August 2022. Adults aged ≥ 60 years, who were diagnosed with type 2 diabetes mellitus for at least 12 months and able to speak English, Mandarin or Malay, were recruited via convenience sampling. The study excluded adults with severe hearing impairment, cognitive impairment or active psychosis. Questionnaires were self-administered with assistance provided as required. Ethics approval was obtained from NHG domain specific review board (reference: 2021/00200). A target sample size of 250 was derived based on an effect size of 0.15, alpha at 0.05, power at 0.95 and factoring for a 10% non-completion rate. Demographic questions included those on marital status, education, income and living arrangements. Functional social support was assessed with an eight-item self-administered survey covering emotional and instrumental support. The scores are transformed to a 0–100 scale, with higher scores denoting better support.[3] Modified medical outcomes study social support survey (mMOS-SSS) has been evaluated as valid and reliable, especially in geriatric assessments (see Supplemental digital Appendix at https://links.lww.com/SGMJ/A117).[7] Summary of diabetes self-care activities assesses various aspects of diabetes self-care via behaviour recall over the last 7 days, with each component calculated through averaging the scores of the respective items. As self-management is multidimensional, different components are not highly correlated, and should be assessed separately.[8] In this study, we assessed the domains of diet, exercise and medication adherence. EuroQol five dimension scale (EQ-5D) measures HRQOL based on five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression, on a 5-point Likert scale.[9] The E-5D utility score is then calculated, with values anchored at 1 (full health) and 0 (as bad as dead). Electronic medical records were extracted and they included (a) demographic information (age, gender, race); (b) glycaemic control (latest glycated haemoglobin [HbA1c] level in the last 12 months); (c) number of chronic medications, and (d) number of chronic medical conditions. Descriptive statistics were used for demographic characteristics. Two linear regression models were developed to assess the association between social support and each health-related outcome. Model 1 investigated the association between social support and health-related outcomes. In model 2, adjustment was carried out for demographic variables, number of chronic diseases and chronic medications. Outcomes of the analyses included B (model 1) and adjusted B (model 2) coefficients and their respective confidence intervals. Statistical analysis was done using IBM SPSS Statistics version 24 (IBM Corp, Armonk, NY, USA). A P value <0.05 was considered significant. RESULTS A total of 250 participants were recruited with a 59.8% response rate. The mean age was 66.4 ± 6.2 years. Among the participants, there were 59.2% male, 57.6% Chinese, most were married (76.4%) and lived with others (88.4%); 70% had completed secondary education and beyond. Of those who reported their monthly income, 44.6% reported it at SGD 2000 and above [Table 1].Table 1: Demographics and variables of interest.Mean HbA1c was 7.5% ± 1.1%, and the mean mMOS-SSS score was 70.7 ± 25.9. Linear regression analysis (model 1) showed a significant association between social support and diet (B 2.71, 95% CI 0.55–4.87) as well as medication adherence (B 4.13, 95% CI 0.43–7.84). The associations with diet (B 2.16, 95% CI 0.09–4.23) and medication adherence (B 5.24, 95% CI 1.76–8.72) persisted with adjustment for demographic factors, number of medical conditions and medications (model 2). There was no significant association between social support and the other health-related outcome variables [Table 2].Table 2: Linear regressions of social support (mMOS-SS) and outcome variables.DISCUSSION Our study examined the associations between functional social support, diabetes mellitus self-care behaviours and HRQOL among 250 older adults with type 2 diabetes mellitus in a primary care setting. We found that functional social support was positively associated with diet and medication adherence. Engel espoused the need for a biopsychosocial approach in care provision, especially for conditions such as diabetes mellitus, due to the complex relationship between biomedical, psychological and social issues.[2] Our findings emphasise the need for clinicians providing diabetes mellitus care to consider and address issues related to social support. The positive association between social support and diet was concordant with findings of other studies,[6,10] highlighting the importance of social support on self-care behaviours. Social support may affect diet through grocery shopping, food preparation and shared meals,[11] particularly in the Asian context. Our finding of a positive association between social support and medication adherence was similar to the findings reported in a Chinese population.[12] Social support has been described to improve self-esteem and coping,[13] and may promote medication adherence through self-efficacy.[14] Persons with better social support may also have more help with medication reminders and monitoring. No association was found between functional support and exercise. We evaluated general functional support rather than support that was specific to exercise. Another study evaluating exercise-specific social support, such as having company while exercising, found a significant positive association.[6] This highlights the multidimensional nature of social support and the importance of addressing its various aspects. Various studies found a positive association between social support and HRQOL,[5,15] where social support may enhance psychological well-being and provide practical support to improve both mental and physical aspects of HRQOL. This was not observed in our study population. This may be due to the relatively high HRQOL, with 98.0% of participants having little or no anxiety/depression and 93.2% having little or no mobility issues, limiting our ability to appreciate the impact of social support on HRQOL. Further studies could purposively sample a higher proportion of participants with issues related to HRQOL to better understand the relationship between social support and HRQOL. Social support was not significantly associated with HbA1c in our study. Similarly, some studies found no significant association.[16,17,18] However, other studies have demonstrated social support being positively associated with glycaemic control,[19,20,21,22] suggesting a complex relationship between the two. Our findings suggest that functional support is positively associated with diet and medication adherence. In older diabetic adults with challenges in medication and dietary adherence, we should consider the adequacy of their functional social support, which can then be incorporated into an individualised treatment plan. The study had some limitations. Our study population had a lower prevalence of depressive/anxiety symptoms and higher educational levels.[23] This may not be reflective of the older Singapore population, which may demonstrate different interactions between social support and diabetic health outcomes. In addition, self-administered questionnaires are subject to social desirability and recall bias and may not reflect the actual behaviours. To our knowledge, this is the first study evaluating functional social support, self-care behaviours and HRQOL among older adults with diabetes in Singapore. Future research should consider exploring the mechanisms that may explain these pathways and whether social support mediates the relationship between self-care behaviours and glycaemic control. In conclusion, functional social support was positively associated with diet and medication adherence, but not exercise or HRQOL, in older adults with type 2 diabetes mellitus. Acknowledgement We would like to acknowledge the support of the National Healthcare Group Polyclinics (NHGP) Clinical Research Unit for this study. Financial support and sponsorship The study was funded by the National Medical Research Council under its Centre Grant Programme (NMRC/ CG/C019/2017). Conflicts of interest There are no conflicts of interest.