Social Support and Mortality: If You're Sick, Friends Can't Save You

To the Editor: A recent meta-analysis of 148 studies revealed that people with stronger social relationships had a 50% greater likelihood of survival than those with weaker social relationships,1 but there was a large amount of variation and heterogeneity in the studies examined on a number of variables such as age, sex, and physical condition. It is thus unclear whether the association between social support and mortality is the same in women and men and whether social support is still protective of mortality after controlling for physical condition. There were three main objectives of this analysis: to investigate the risk of all-cause mortality in older adults associated with social support, to identify sex differences in this association, and to determine whether this association is significant after taking baseline health status differences into consideration. Data were obtained from the Men, Women and Ageing project, which incorporates data from two population-based longitudinal studies that began in 1996: the 1921–1926 birth cohort of the Australian Longitudinal Study on Women's Health (ALSWH) and the Perth Health in Men Study (HIMS). Detailed methods for both studies have been described elsewhere.2, 3 The human research ethics committees of the University of Newcastle and the University of Queensland approved the research protocol for the ALSWH. The ethics committee of the University of Western Australia approved the HIMS research protocol. Participants in HIMS are aged 69 to 87 and live in an urban area (Perth, Western Australia), whereas ALSWH participants are a national sample including regional and rural areas and are aged 73 to 78. To eliminate potential confounding related to age and area of residence, only urban women (n = 4,252) and men aged 73 to 78 (n = 2,625) were included in the current analysis. Identification information was probabilistically matched to the Australian National Death Index to identify deaths. Survival time was calculated as the number of days between the return of the survey (in 1999 for ALSWH and between 2001 and 2004 for HIMS) and the date of death or survival at October 1, 2009. Social support was measured using an abbreviated version of the Duke Social Support Index (DSSI).4 The two subscales of the DSSI measure structural social support (network) and satisfaction with subjective support (satisfaction). Scores for the network subscale ranged from 0 to 21, with higher scores indicating more social contacts. Satisfaction scores ranged from 6 to 18, with higher scores indicative of greater satisfaction. The satisfaction scores were positively skewed, so the median score of 17 was used to dichotomize the variable so that 17 or greater was categorized as satisfied and less than 17 as not satisfied. Body mass index was calculated from height and weight (kg/m2) and categorized according to World Health Organization recommendations5 as underweight (<18.5 kg/m2), normal (18.5–24.9 kg/m2), overweight (25.0–29.9 kg/m2), and obese (≥30 kg/m2). Smoking categories were never, former, or current. Participants reported whether a doctor had ever diagnosed them with any of the following chronic medical conditions: heart disease, diabetes mellitus, hypertension, stroke, or chronic obstructive pulmonary disease (COPD). Participants were categorized as healthy if they reported no prior history of any chronic condition and unhealthy if they reported a history of at least one chronic condition. Education was determined according to the highest level attained (primary, secondary, and tertiary school). Marital status was categorized into three groups, partnered (married or living with partner), not partnered (single, separated, or never married), and widowed. For all-cause mortality, proportional hazard models were used to estimate the hazard ratios (HRs) for men and women for social network and satisfaction (reference = not satisfied). The final models included age, body mass index, smoking status, marital status, and highest level of education as covariates together with interaction terms with sex. To determine whether the presence of preexisting illness modified the relationship between social support and mortality, the survival analysis was repeated separately in healthy and unhealthy participants. Missing data were excluded from the analysis. The number of deaths between baseline and October 1, 2009, for men and women were 542 (20.7%) and 1,170 (27.5%), respectively. The mean follow-up time (to death or date of return of the last survey) was 5.9 years for men and 9.3 years for women. HRs of all-cause mortality according to social support are shown in Table 1. Larger network size (HR = 0.89, 95% confidence interval (CI) = 0.84–0.95) and greater satisfaction with support (HR = 0.85, 95% CI = 0.73–0.99) were associated with a lower risk of mortality. In healthy participants, larger network size was significantly associated with lower risk of mortality, but no significant association was found between satisfaction with social support and mortality. In unhealthy participants, neither satisfaction with social support nor network size predicted survival. The sex interaction terms for network size and satisfaction were not significant. Although these findings were consistent with those of others in showing that social support is protective of mortality, this relationship was not apparent in people who were categorized as unhealthy. These findings highlight the importance of good health in accessing and maintaining social support in late adulthood. Despite women in this cohort reporting higher social support than men,6 no sex differences were found in the effect of social support on mortality. These results were surprising given the amount of recent literature describing the protective attributes of social support on mortality, although they are consistent with earlier studies that reported that social support was not associated with survival for men or women once adjustment was made for health status.7, 8 A particular strength of the current study is its reliance on a measure of multiple aspects of social support: structural support and satisfaction with subjective support. The results suggest that health status may moderate any protective benefit of social support on mortality. We thank participants of the ALSWH and the HIMS for their valuable contribution to this project. Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the authors and has determined that the authors have no financial or any other kind of personal conflicts with this paper. The Men, Women and Ageing project is funded by National Health and Medical Research Council of Australia/Australian Research Council Ageing Well, Ageing Productively Strategic Award 409953. The ALSWH is funded by the Australian Government Department of Health and Ageing. The HIMS is supported by grants from the National Health and Medical Research Council of Australia (279408, 379600, 403963). Author Contributions: DM was responsible for the study concept. All authors were responsible for the study design, interpretation of data, drafting and critical revision of the manuscript. JL and AD: Statistical analysis and preparation of the data. All authors approved the final version of the article. DM is the guarantor. Sponsor's Role: The funding sources had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; or preparation, review, or approval of the manuscript.