Background Social capital may be a social good in health terms, but it is not necessarily a universal good. Several studies have shown that while there is a positive association between ecological social capital and health in people with high individual-level social capital, this relationship is weaker or even reversed in those with low individual-level social capital. Such studies, however, have used relatively coarse levels of geography for quantifying ecological social capital. The present study looks at this relationship at a more fine-grained spatial scale.

Methods Data from the National Survey for Wales (n=27 828, weighted mean age=48.4) were linked to previously published small-area estimates (n=410) of ecological social capital for Wales. Mixed effects models were then used to assess whether the relationship between mental well-being and self-reported health on one hand, and ecological social capital (sense of belonging) on the other, was moderated by individual-level social capital.

Results The models found the same moderation of the relationship that has been demonstrated previously: Although ecological social capital is positively associated with health in respondents with high individual-level social capital, the relationship is negative in those with low individual-level social capital.

Conclusion This study replicates this association at a spatial scale orders of magnitude more fine-grained than had been shown previously. Ecological social capital is not an unambiguously positive factor for public health, and may be a risk factor for marginalised people.