Field of Science

Religion and health: the big meta-analysis shows it's about attending, not believing

A couple of epidemiologists from University College London have published a meta-analysis* of nearly 60 studies into the relationship between religion and health.

Their aim was to give a definitive statement not only whether religious people have better health, but to get some precision on what types of religiousness are important.

The figure shows the results from prospective studies (i.e. ones where they measure religiosity, and then some time down the road they measured health) in originally healthy people. The spot indicates their best estimate, and the lines the 'confidence interval'. Basically, if the horizontal line crosses the vertical one, then that particular factor probably has no effect on health.

There's a lot of stuff in there about study design and other factors, but I've picked out the important ones - three factors that do not connect religion and health.
  1. If you are a man, then it seems like being religious isn't connected to being more healthy - or if it is, the effect is quite small.
  2. Intrinsic aspects of religion (belief in a god concept, religious/spiritual well-being, religious/spiritual experience, and religious motivation/orientation) have no effect on health.
  3. Although organisational activities (such as going to Church) seem to have a big effect, non-organisational activities (prayer, meditation, or sacred book study) do not.
In other words, there seems to be a real health effect for women, at least, from participating in group religious activities. According to their analysis, this effect holds even after you control for happiness, strength of social networks, and unhealthy behaviour.

However, in people who are already sick, there is no benefit from any kind of religion. You die just as quick!

The results give some hard, objective numbers to the conclusions that have been made before (see earlier post). These really do seem like consistent, strong findings. Now the only question that needs answering is what's driving them!


*A meta-analysis is a tool that uses statistics to get the average results of a large number of studies.

ResearchBlogging.orgChida, Y., Steptoe, A., & Powell, L. (2009). Religiosity/Spirituality and Mortality Psychotherapy and Psychosomatics, 78 (2), 81-90 DOI: 10.1159/000190791


  1. I think it shows that believing without attending gives little measurable benefit, which is somewhat different :-)

  2. It is different, but it's also not the whole truth.

    What these studies show is that people who don't have any particularly strong beliefs get just as much out of church attendance.

    It may sound odd, but there are a lot of people who go to church but who aren't believers in any meaningful way - particularly in places like the US, where most of these studies are done. And these people get just as much benefit from churchgoing.

    So it really is about attendance, not belief.

  3. This comment has been removed by a blog administrator.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS