Research on social networks has shown that social ties and social activities are positively related to physical health. Numerous studies show that, for one reason or the other, being socially more engaged and having stronger social ties helps us avoid alcoholism along with a host of other hazardous behaviors and diseases. Indeed the positive effects of social integration are strong enough to exhibit significant effects on your longevity.
Religiousness is one very important form of social engagement, and hence it has been a considered a well-established fact among sociologists and social psychologists that religiousness too increases longevity. Indeed, several studies (a few are cited below) estimate religion as reducing mortality by roughly 25%. I.e. for two otherwise identical participants in a longitudinal study, the estimated likelihood of either surviving until the time of the follow-up study differs by 25%.
Despite a large amount of data on this phenomenon, there remain a lot of questions regarding this observed association between religiousness and longevity. As the authors of one recent study put it, analyzing this data requires us to also ask questions such as:
"What causes people to be religious? Do those causes of religion also cause the religion-longevity association? What does religion cause? And are those effects of religion the mechanisms through which religion obtains its associations with longevity?"
For example, social psychologists have been able to show that religious people are on average more conscientious and more agreeable than their non-religious counterparts. They also possess greater self-control, and are therefore generally better at committing to all sorts of social institutions and community projects; all of which are known to cause improvements in health behaviors even for the non-religious.
Combined with the finding that religious people do indeed participate more in non-religious social institutions, we might hypothesize that it is not religious affiliation per se that exhibits the health effect, but rather that it is the behavior that stems from the underlying personality traits (which also lead people to become active in religion) that ultimately cause the effect.
Another interesting finding, regarding the health effects of religiousness is the somewhat dazzling effect that health effects from religious engagement seem to be stronger for women than they are for men. So far there has been no conclusive study on the reasons for this effect, but scientists - as I am told - do not think the effect likely to be the consequence of a divine preference for females (which may or may not have been your first guess).
The main hypotheses that scientists are now investigating, can be summed as follows:
Do we see that death comes later to religious people, because religious commitment causes people to commit to social institutions (e.g. family, marriage, community) which in turn improve health behavior? (In which case religion could be seen as a initial cause of increased health). Or do strong social ties generally encourage people to also affiliate themselves with religious institutions? (in which case religion would simply be a side effect of other factors improving health behavior). Or is it possibly even both. The hypotheses are yet to be fully dissected (a model for this type of research can be found in the McCoullough et al study), and the discussion should remind us of the difficulties involved in analyzing social science data for causality.
Main References:
McCullough ME, Friedman HS, Enders CK, & Martin LR. (2009) Does devoutness delay death? Psychological investment in religion and its association with longevity in the Terman sample. Journal of personality and social psychology, 97(5), 866-82. PMID: 19857007
Berkman, Lisa. (2000-9-26) Social Support, Social Networks, Social Cohesion and Health. Social Work in Health Care, 31(2), 3-14. DOI: 10.1300/J010v31n02_02
Uchino, Bert N. (2006-8-7) Social Support and Health: A Review of Physiological Processes Potentially Underlying Links to Disease Outcomes. Journal of Behavioral Medicine, 29(4), 377-387. DOI: 10.1007/s10865-006-9056-5
