Background

Koenig and colleagues present two connected review articles in BJPsych Advances on religion and psychiatry. The first is a selective summary of research (Koenig Reference Koenig, Al-Zaben and VanderWeele2020a). The second makes recommendations for clinical practice (Koenig Reference Koenig, Peteet and VanderWeele2020b). Professor Koenig provoked controversy in 2008 with a paper on similar themes in the Psychiatric Bulletin (Koenig Reference Koenig2008), advocating integration of so-called ‘religion/spirituality’ into clinical practice. In my opinion, these new articles are best seen in the context of the continuing debate within UK psychiatry (Poole Reference Poole, Cook and Higgo2019).

There are some uncontroversial points here: patients’ faith should be treated with respect as an important part of their lives; religion can be both helpful and unhelpful to mental health; understanding the relationship between symptoms and religious beliefs is sometimes important; and it is legitimate for psychiatrists to work alongside chaplains and other religious leaders. I teach these points to medical students, just as I was taught them in the 1970s. Other aspects of the article are more contentious.

As is commonly the case from advocates of the integration of religion into clinical practice, the articles start with assertions that psychiatry has generally regarded religion as intrinsically pathological. I see little evidence that this is correct, even taking into account the citations here. However, what has worried sceptical clinicians about the growing literature on psychiatry and religion has been lack of clarity about relevant professional boundaries (Poole Reference Poole and Higgo2011).

Overview of evidence

There are gaps in the overview of research on religion and mental health, the most obvious being the omission of some major public concerns. Among these is the effect on mental health, short- and long-term, of child sex abuse perpetrated by priests and other religious figures in the context of their religious role (Lueger-Schuster Reference Lueger-Schuster, Kantor and Weindl2014), often facilitated or covered up by religious authorities (Death Reference Death2015). This has proven to be so serious and widespread that it has led to major change in the status of religious authority all over the world, especially the Roman Catholic church. Similarly, there is nothing here about unequivocally harmful religious counselling, such as sexual orientation conversion therapy. This has been condemned as ineffective and unethical by the Royal College of Psychiatrists (2014), but is still advocated and practised by some Christians in the UK (Mosely Reference Moseley2018). These omissions do not invalidate Koenig and colleagues’ articles, but they do mean that they lack balance between the benefits and harms of religion.

Recommendations

Like most recommendations for clinical practice, Koenig and colleagues’ clinical guidance is based more on opinion than evidence. The statement in the first article that ‘such studies […] have the potential to contribute evidence for causation’ (Koenig Reference Koenig, Al-Zaben and VanderWeele2020a) is appropriately cautious, as the totality of the evidence does not remotely meet the Bradford Hill criteria (Bradford Hill Reference Bradford Hill1965). Their clinical guidance, however, is implicitly based on an assumption of causality. This is not an obscure technical point. Much of the literature concerns Christians in the USA, who are the majority in the most religiously active of high-income countries. Religion is a social as well as a spiritual phenomenon, closely linked to ethnicity and class. These are potential confounding factors in most studies. Furthermore, in culturally diverse societies such as the UK, religion has many implications relating to power imbalances and to misuses of power. For example, patients belonging to Muslim minorities might reasonably feel nervous about discussion of their faith with non-Muslim psychiatrists, given prevalent levels of Islamophobia in the UK and USA. These concerns are highly salient to clinical practice, and it is a shame that there is no guidance here about how they might be tackled in practice. In my opinion, the clinical recommendations are not of universal applicability.

Assumptions

In the article on clinical applications (Koenig Reference Koenig, Peteet and VanderWeele2020b), the authors correctly state that there is little controversy over holistic care that takes into account the patient's religion, where this is the patient's wish. However, I am less comfortable about the related assertion that implies that mental health professionals should assume that religion is benign and beneficial. Both the full literature and experience show that, even among the devout, religion can be more ambiguous than this. The authors do make reference to problems caused by religion, but this is mainly described as ‘how patients can use religion neurotically’. It seems possible that sometimes the problem can be with religion and not the patient. It is right that psychiatrists should not seek to undermine their patients’ faith, but, in my opinion, they should also be careful not to make assumptions about it.

Moral injury

The clinical applications article makes reference to moral injury, a problem that has become topical with regard to decisions that healthcare professionals have to make in coping with the current coronavirus pandemic. They make reference to spiritually integrated cognitive processing therapy, a novel intervention for moral injury in the context of post-traumatic stress disorder (PTSD) that is under evaluation (Pearce Reference Pearce, Haynes and Rivera2018). This has been developed on the basis of the observation that moral injury and ‘spiritual struggles’ make it difficult to recover from PTSD and that US military personnel and veterans report high levels of religion/spirituality. I do not criticise the development of explicitly religious forms of therapy, but they are culture bound and therefore not necessarily generally applicable. Many people in the UK do not derive their morality from organised religion and, in any case, it is not clear that moral injury is best understood as a psychiatric disorder per se. Most importantly, a request for a religion- or spirituality-informed therapy does not absolve the therapist of responsibility to be vigilant for inadvertent violations of boundaries.

Conclusion

Few areas of psychiatry are immune from a mismatch between rhetoric and evidence. In the case of religion, I would be less concerned if it were not for persistent recommendations to change the boundaries of professional practice in a way that I and others believe to be inappropriate. This long debate appears to be far from over and readers will make up their own minds about the salience and validity of these contributions.