People who go to church live longer. Here’s why

    27 July 2016

    Over the last 20 years, research has gradually accumulated suggesting that religious service attendance is associated with better physical and mental health. For example, research articles have indicated that regular religious service attendance is associated with a 30 per cent reduction in depression, a five-fold reduction in the likelihood of suicide, and a 30 per cent reduction in mortality, over 16 years of follow-up.

    There have been a number of prior studies on religious service attendance and longevity. Many of these had been criticised for poor methodology, for instance allowing the possibility of reverse causation — ie, that only those who are healthy can attend services, so that attendance isn’t necessarily influencing health.

    Papers recently published out of Harvard University have tried to address this concern by using repeated measurements of service attendance and health over time to control for whether changes in health preceded changes in service attendance. The associations between religious service attendance and longevity, suicide and depression were all robust. Results indicated that compared with women who never attended religious services, women who attended more than once a week had a 33 per cent lower mortality risk during the study period. Those who attended weekly had a 26 per cent lower risk and those who attended less than once a week had a 13 per cent lower risk. (The data comes from women who worked as nurses in the US, most of whom identified as Catholic or Protestant, so most of the religious services would be at churches. However, the definition encompassed a range of different places of worship.)

    These results naturally raise the question as to why religious participation is associated with health. One explanation would be social support and, indeed, the results indicate that social support is important: by attending religious services, one develops social support which itself affects health, but this appears to account for only 20 per cent to 30 per cent of the effect.

    The research suggests there are likely to be several other mechanisms that are operative as well. In addition to social support, the social and behavioural norms at religious services (in the United States at least) seem to reduce the likelihood of smoking, which affects health. Another mechanism seems to concern one’s outlook on life: perhaps because of a message of hope or faith at religious services, there are higher levels of optimism among those attending and depression rates are lower; and these things have likewise been shown to affect physical health and longevity.

    Other mechanisms might also be operative. The development of self-discipline and a sense of meaning and purpose in life have been proposed in the literature as potential factors. The association between service attendance and health seems not to be explainable by just one mechanism alone. Rather, there appear to be many pathways from religion to health. Religious service attendance affects many aspects of a person’s life and the cumulative effect of all of these seems to have a substantial influence on health.

    One intriguing aspect is that it appears to be religious service attendance, rather than self-assessed religiosity or spirituality or private practices, that most powerfully predicts health. Something about the communal religious experience does seem to matter. Religious identity, spirituality, and private practices may of course still be important and meaningful within the context of religious life, but they do not appear to affect health as strongly.

    In an era in which people increasingly self-identify as ‘spiritual but not religious’ and in which the term ‘organised religion’ tends to carry negative connotations, this empirical research perhaps challenges our preconceptions and maybe suggests that personal spirituality, discarding all organised and communal aspects of religion, may not be an entirely satisfactory way forward.

    An interesting question is to what extent similar results would pertain to other forms of social participation. Again, the research indicates that social support explains only about a quarter of the effect on longevity; it is important, but not everything. There is, however, some literature suggesting that participation in other social groups likely has some effect on mortality, though the size of the effect tends to be somewhat smaller than it is for service attendance.

    My speculation, though we do not yet have data on this, would be that groups that not only have social gatherings, but also have a shared sense of meaning, healthy behavioural norms, and a common vision for life would have a larger effect on mortality than, say, merely showing up for a regular card game. Religious service attendance likely affects health not simply because of social support, but also because it potentially shapes so much of one’s outlook, behaviour, beliefs, and one’s sense of life’s meaning and purpose.

    The research has interesting, and sometimes challenging, implications for the practice of public health and medicine. Public health impact is often assessed based on how common an exposure is and how large its effects are. With religious service attendance, the exposure is relatively common: about 40 per cent of Americans report attending services weekly; the corresponding figure in the UK is about 10 or 15 per cent, which is considerably lower, but still relatively common compared to other types of social participation. Moreover, as discussed above, the research on service attendance and health appears to suggest a relatively large effect.

    All of this indicates that religious service attendance is an important, and probably under-appreciated, social determinant of health. It is something that should be taken into account in public health discussions. We would not think to neglect other powerful social determinants of health such as race, or gender, or social support in our discussions. The same should probably be true of religious service attendance as well.

    Finally, within the context of medicine, there have been debates as to whether and to what extent it is appropriate to discuss issues of religion and spirituality in a clinical setting. Whether religion should be discussed with patients during care will of course be context-specific. It may be thought to be more appropriate and important in end of life settings. Many patients — and in the United States it is the majority — say they think that doctors should consider patients’ spiritual needs. Many doctors feel uncomfortable doing so. Conversations will be easier if the clinician and the patient share the same, or a similar, faith, but these matters can be discussed in general terms as well. Training on how to do so can be helpful. If issues of religion and spirituality do come up in conversation, clinicians could inquire about service attendance as a potentially meaningful form of social participation.

    Decisions about religious practice and formation of religious beliefs are, of course, not generally made on the grounds of health. However, for those who already hold religious beliefs, but do not attend services, the research on religion and health does question whether they are perhaps missing something of the communal religious experience that is powerful, at least for health, and possibly for much else as well.

    Tyler J VanderWeele is professor of epidemiology at Harvard T H Chan School of Public Health