“Better health, a higher degree of well being, more marital satisfaction, less addiction, less suicide, better likelihood of healing.”  You can have it all if you are really religious, say Theodore Chamberlain and Christopher Hall in their new book Realized Religion: Research on the Relationship between Religion and Health.  Dr. Theodore J. Chamberlain is an associate professor of counseling psychology at Eastern College, St David’s, Pennsylvania.  Dr. Christopher A. Hall is an associate professor of biblical and theological studies in the same institution.

Their efforts, and the researchers they quote, are part of a wider movement to ‘scientise’ religion – a move to treat the process, psychology and phenomenology of religion like any other physical or psychological activity at least in terms of its empirical approachability.  It is a contentious approach and has been both approved and criticised from within and without the science-religion discussion arena.

The authors are full blooded in this book in their ambition to regularize, to normalize religion in the scientific community.  Both avowedly are committed Christians, but suggest that their personal orientations have not affected their evaluation of the more than 300 research articles quoted in the book.  Their contribution to this burgeoning field is to scan a large compendium of research and order it into broad health categories with some added commentary.

They lay out their claims specifically in the introduction.  “Realized religion” is defined as “the essential elements of religion that are made operational by being brought into concrete existence.”

Health is defined in broad physical, psychological and social terms.  These terms are so broad, and to some extent value laden, that some might have a little difficulty with them.  For example, is premarital sex a health issue or a religious view about behavior?

Chapter One is a consideration of the role of prayer in health.  The issue of the experimental demonstration of the supernatural efficacy of prayer was considered in a review of Russell Stannard’s “The God Experiment” published in October (2001.10.03) and November (2001.11.05) editions of Metanexus.  The same views and considerations by the writer apply to the current authors’ treatment of the topic.

Further chapters consider the issues of religion in faith healing, mental health, suicide, alcohol use and abuse, and life and marital satisfaction.  The final section of the book is devoted to a discussion of religious values, empirical research, and Christianity.

Broadly, their claim is that some 75% of the studies quoted show the health benefits correlated with the practice of religion.  They are also clear about how they would interpret this correlation.

Firstly, they suggest that the relationship is causal – religious practice produces better health.

Secondly, they claim it is the supernatural component of religion that is the causal agent.  “The transcendent interacting of the human with the divine account for the positive effect of religion on health and not simply the placebo phenomenon.”

The authors list voluminous correlational studies in support of their position then round off each chapter with the summary view that the case has been made for a particular health benefit being consequent on the serious practice of religion.  Typical of the research quotations is, “Hadaway and Roof (1978) found that 50% of people who belong to a church or synagogue rated their lives as extremely worthwhile compared to 39.4% of unaffiliated or non members.”

Their final conclusion is that “realized religion” is good for any or all aspects of health.

Experimental design

The authors’ thesis, research evaluation, and discussion invite consideration on a number of levels.  With such ambitious claims one might have expected a clear explication of how the field should be construed and how the research maps on the central issues.

Realized Religion is a very substantial introduction to the literature on the topic.  In terms of mapping the fundamentals of the relationship between health and religion very much further analysis and research are required.  The book might have benefited from such analysis even if it meant reducing the volume of quoted research to make room for it.

Before considering the literature quoted, it would have been very useful to know the criteria for choosing and rejecting studies for the series.  Did the authors just choose say the first 300 or so studies they came across?  Did they evaluate each on its technical merits independent of the conclusions?  It is very easy to see how considerable unconscious bias could influence the process particularly in such a soft research area.  Of significant concern is the asymmetry in the evaluation of the literature by the authors.  When considering the studies where religion is associated with individual or group psychopathology, the stance assumed by the authors is that the religion is fine and it is the individual’s or group’s native pathology which is causative.  If the association is positive it is religion which is credited not the native psychological state.  No attempt is made to justify this, and perhaps the authors are even unaware of it.  If the studies are sorted on this basis the conclusion was biased from the beginning.

A nested series of experimental issues need to be addressed:

* The basic question is, “How good are the raw correlational studies?”  Some informed criticism by the authors of the study designs, operational definitions of religiosity and health outcomes would have been reassuring.

* What is the evidence for causation in these studies?  Consider, for example, how an independent psychological factor could be the determinant of both low blood pressure and of the psychological predisposition to attend church regularly.  This would mean that attending church is not actually the cause of low blood pressure.

At least three major psychological variables confound the interpretation of many such studies.

1.  A vast research literature on psychosomatic health has already well documented many of the relationships between health and various positive psychological and social processes.  Married men have better health than unmarried men do (interestingly the female correlate doesn’t hold).  People with strong positive life attitudes and social networks live longer, survive illness better etc.  It’s even been shown that having a shed, a workshop, an office retreat is very good for health.  Being chronically critical is bad for your coronary arteries: thus it behooves reviewers to be gentle for their own health’s sake.

2.  Quite independent of these issues is the well-mapped arena of the placebo effect.  The authors have some unusual views of placebo phenomena.  Their take is that it is unreal, or imagined or insubstantial in some way.  This is not the case.  The placebo effect by definition produces exactly the same outcome as the main effect in a clinical trial.  The difference is that the outcome is mediated by the beliefs of the person about the treatment rather than by the treatment process itself.  Additionally they seem to confuse placebo phenomena with the positive psychosocial processes that generally correlate with better health though these are completely independent phenomena.

3.  Apart from any views about supernatural phenomena, religious and spiritual experiences may be potent psychological factors in a person’s life.  Purely on this basis alone you would predict health correlations as occur with other positive psychological experiences.

Supernatural causality

Most importantly, the whole arena of discussion swings on the axis of supernatural causation.  Chamberlain and Hall locate supernatural causality in the religion/health arena as the central issue (correctly I believe).  Yet the studies quoted make no attempt to establish this central fact.  For a study to do this it would have to control for the expected positive psychological input from religious activity in order to tease out the additional supernatural component.

There is a universe of difference between religion as one of the very many varieties of psychosocial boons to health and supernatural intervention in the causal schema (miraculous by definition).  It is of general interest that religion may be one of life’s many positive health effecto s but believers generally (and the authors in particular) understand it as something very different to say getting the health benefits of having an enjoyable hobby.  Supernatural causality was thus the defining thesis and the “swing issue” of the book.  Yet it is not clear whether a single study of the very many quoted is directed towards supporting that view.

If supernatural causality is intended as the explanation in the health/religion relationship there are many issues to address.  Is it supernatural activity on the part of the person themselves that is causative?  Is God or some other agency understood to be performing a miracle?  What are the implications for explanation in science at large of such a worldview?  There is an enormous inferential gap between correlational studies of religious activity and health and any kind of case for supernatural causality.  Multiplying the number of such studies does not help to close that gap.  The question of how committed religious groups are to the scientific and experimental route need to be addressed at the beginning.  If the only acceptable science to such groups is the science which confirms a preexisting belief then this is not science at all and it would be more honest to simply abandon the program.

The wider perspective on religion and health

The religion/health relationship varies enormously depending on whose religion and whose health we are considering.  It is perhaps a defining sign of our culture that our natural orientation to the issue is, “my religion and my health.”

My religion may be good for my health but my religion could be very bad for your health.  E.g., the religion of slave owners was very bad for the health of slaves.  This is an extreme example to make the general point.  Yet it is important because exactly the same question of perspective arises when the relationship is more subtle and covert as it is in the complex socioeconomic orders we live under.  All religion which exists in coalition with socioeconomic oppression is bad for the health of many people and good for the health of some.

The strongest predictor we have of an individual’s health is their degree of wealth or poverty.  One could assess the thesis of the book simply by asking whether the nations of the planet which are most religious have the lowest differentials of wealth and poverty.

What kind of religion?

What kind of religion is being considered in the book?  Chamberlain and Hall seem to have a kind of middle of the road conservatism in mind.  They are largely ignoring the charismatic religions of faith healing and extreme fundamentalism on the one hand and the liberal traditions on the other.  No objective criteria are claimed for distinguishing religions and the selection is apparently just their own personal preference.  Does this then mean that the claim being made is restricted to a particular variety of  Christianity?

Chamberlain and Hall have done a lot of helpful footwork in assembling the health/religion literature.  From here on a lot of headwork and tailored experimentation will be required to demonstrate the supernaturally mediated health efficacy of religion, realized or otherwise.