Most of the extant research examining the link between social support, social networks, and mental health service use has focused on clinical samples and is limited to either psychiatric inpatient or outpatient services. This population-based study highlighted the association between social network and social support and use of different types of mental health services across different mental health conditions. It supported our hypotheses that the association between service use and social network and social support varied across friends, relatives, and spouse, but the association differed in that increased social network and social support was not uniformly associated with decreased use of services. However, this was not surprising because of the variation in existing evidence. There were similar patterns of service use across different mental health conditions, which suggest that one’s social network or social support was largely unrelated to the diagnosis.
Specifically, these findings indicated that for general medical service use, social network measures (for example, frequency of meeting friends or relatives) were directly related to service use, such that less frequent contact was associated with reduced odds of accessing services. In contrast, increased social support from the spouse was associated with increased service use. These findings contradict those in a literature review (12
), which found that less social network or social support was generally associated with more service use. However, our findings support more recent research that indicated that higher scores on a composite social network index (especially for contacts with relatives and friends) were associated with use of inpatient psychiatric services in local hospitals (14
It is important to note that we examined the effect of number and frequency of contacts separately. It is possible that for general medical services one’s social networks and social support are more likely to suggest early consultation because it carries less stigma than formal psychiatric services (28
). Also, when symptoms first appear, the cause of illness is often unclear and one typically assumes physical rather than psychological factors as the causal agent. However, once the cause is identified as related to mental health, the use of even general medical services to discuss the problem is reduced if one has good social support from relatives. We hypothesize that this might be a reflection of the stigma associated with more formal mental health services and mental health in general (29
Consistent with previous research (9
), increased social network (that is, contact with more relatives and greater social support from friends and relatives) was associated with less use of specialty psychiatric services. Unadjusted results showed that increased levels of social support were associated with reduced use of other human services, a result similar to that found in use of other mental health services and probably reflective of stigma (because other human services were used for mental illness). On the other hand, having contact with more than one friend was associated with increased use of other human services. This was evident even in the adjusted analysis. Increased use of other human services when in contact with more than six friends is an indication that a large number of casual contacts constitutes a less dense network than that found with relatives or spouse. The network of a large number of friends is less dense because of less intimacy, emotional intensity, and lower social support; however, such a large number of contacts tends to provide more information about available services, thus encouraging help seeking (32
). A network of casual contacts has been defined as constituting weak ties (33
), a concept based on the strength and quality of interactions within one’s network, where weak ties result in sharing of general information about multiple services and strong ties result in sharing of detailed information about fewer specific services. Increased social support could be indicative of such strong ties, thus being associated with reduced use of other human services, as seen in the unadjusted results. Others also found evidence to support the concept of weak ties, especially for nonformal care (34
It is important to consider some limitations when interpreting these findings. There was a relatively large loss to follow-up between 1993–1996 and 2004–2005, the majority of whom had died. Weighted analyses accounted for attrition and reduced biases resulting from attrition, but residual biases are possible. However, the pattern of results was quite similar across all the sets of analyses, and this provides more evidence to the robustness of the findings. The social support measures were developed specifically for the ECA program and were not a standardized social support scale. Moreover, they were self-reports and could not be corroborated by another source. The findings may have varied if a different measurement strategy had been employed. Barriers to care and stigma were also not tested and could be explored in future research. It might be difficult to generalize these findings to other populations across different countries or cultures. Moreover, for the purpose of these analyses, some of the more severe mental disorders, such as schizophrenia, were not included because of low prevalence.