Abstract

This study uses Australian data collected from 1992 to 2001 to estimate the impact of social networks on physical disability. Of the four social network sub-groups (children, relatives, friends, and confidants), only relatives demonstrated a consistent protective effect on transition to disability. It is possible that extended family relatives are able to offer advice about better health habits and access to health services that may be more easily accepted than from children and spouses, and even from friends. Background: Understanding factors associated with disability in the elderly is an important component of predicting health and social needs in an ageing population. While broad models seeking to explain disability as a function of individual health, environment, personal attributes, and psychosocial factors are relatively common, few models examine the specific effects of social networks on disability in the elderly in greater detail. Functional performance of activities of daily living (ADLs) such as bathing, toileting, and eating are commonly used as a proxy for disability, even though evidence suggests there may be earlier transitions to disability than can be identified using ADL measures. This study was designed to estimate the impact of various types of social networks on two different measures of physical disability.

Methods: Data were obtained from Australian Longitudinal Study of Ageing (ALSA), a large epidemiological study of the population carried out in six waves between 1992 and 2001. Disability was measured at each wave, while social networks and other factors included in the model were measured in the first wave only. Two measures of disability were considered: 1) mobility (defined as the ability to walk up and down a flight of stairs and to walk half a mile without help) and 2) Nagi disability (based on level of difficulty in performing five tasks including kneeling, lifting, and reaching). Respondents were considered disabled if they had difficulty with one or more of the component tasks. Social networks were broken down into four sub-groups: children, relatives, friends, and confidants. For children, relatives, and friends, social network scores were derived based on the number, proximity, and frequency of contact. For confidants, scores reflected the number, and whether the confidant was a spouse. Scores for each of the groups were summed to obtain a total score for social networks as a whole. The models controlled for the influence of variables such as age, gender, receipt of community or residential care, marital status, household income, education, self-rated health, number of chronic conditions, depressive symptoms, and cognitive function. Data were used to estimate the effect of social networks on the probability of transition from non-disabled to a disabled state during the study period.

Finding: Results were generally mixed. Among social network sub-groups, only relatives demonstrated a consistent protective effect on transition to disability. This protective effect was stronger for Nagi versus mobility disability. Social networks, as a whole had a significant protective effect on transition to Nagi disability, but not to mobility disability. When analysis was extended to include incomplete data (i.e., deceased or no follow-up data), only relatives had a consistently protective effect, similar to core analysis. However, individuals initially not disabled with high relative social network scores had a significantly higher probability of dying, a possibly contradictory finding that suggests caution in interpreting these results.

Conclusions: This study provides tentative evidence that social networks, in particular those of relatives outside the nuclear family, have a protective effect on the probability of elderly people becoming physically disabled over time. It is possible that extended family relatives are able to offer advice about better health habits and access to health services that may be more easily accepted than from children and spouses. Strength of the evidence is somewhat mitigated by the general lack of a significant effect for all other types of social networks, and by the finding in one analysis that social networks of relatives increased the likelihood of death. More research on the specific nature of the relationship between social support and physical disability in the elderly could help clarify these apparent inconsistencies.

Reference: Giles, LC, Metcalf, PA, Glonek, GF, Luszcz, MA, Andrews, GR "The Effects of Social Networks on Disability in Older Australians" Journal of Aging and Health August, 2004; 16(4), 517-538.