In May 2026, an Associated Press report revisited a familiar claim: people who attend worship regularly often report better mental health and lower levels of some serious risks. The article also included the other side of the evidence. Religious communities can provide belonging, structure, and hope, but they can also produce shame, exclusion, trauma, and barriers to treatment.
For someone who has left Christianity, the claim can land as an accusation disguised as health advice. If churchgoers are doing better, perhaps leaving was the mistake. Perhaps the cure for loneliness or anxiety is to return, even without believing. That conclusion goes beyond what the evidence can show.
The research does support a narrower statement: religious attendance is often associated with better outcomes in particular populations. The harder question is whether attendance causes the improvement. A study published on October 1, 2026, gives that distinction unusual importance.
A new study finds no clear treatment effect
Gabriele Prati’s 2026 study in Public Health used several waves of the UK Household Longitudinal Study, including 86,093 adults. The analysis was designed to address time-varying confounding, the problem that a person’s health, satisfaction, and attendance can all change and influence one another over time.
After accounting for baseline factors and changing measures of health and life satisfaction, frequent religious attendance did not significantly predict later mental-health scores. The confidence intervals still allowed for small benefits or harms, so the result is not proof that attendance never helps anyone. It does challenge the stronger claim that attendance is a generally effective mental-health intervention.
That result is consistent with Prati’s 2025 analysis in Psychological Science. Using 10 waves of British data covering 18 years and 29,298 people, the study examined whether changes within the same person predicted later changes in mental health and life satisfaction. Most associations were not statistically significant. The few significant results ran in both directions, suggesting possible benefits in some circumstances and possible costs in others.
Why earlier studies often look more positive
Many studies compare frequent attenders with people who rarely or never attend. Those are not interchangeable groups. Regular attenders may have stronger family networks, more stable routines, healthier behaviors, greater income, or an established sense of purpose. People experiencing depression, disability, alienation, or conflict may find attendance harder and gradually withdraw. Better health can make attendance easier just as attendance might affect health.
Researchers try to adjust for these differences, but an observational study cannot measure every relevant feature of a life. Even a study that follows people for years can confuse differences between people with changes caused by attendance. That is why within-person analyses and methods designed for changing confounders matter.
The positive evidence should not be dismissed. A 2020 study of three prospective US cohorts associated service attendance with lower levels of depression, anxiety, hopelessness, and loneliness, along with greater life satisfaction, social integration, and purpose. It found little association with most physical diseases. The authors framed attendance as potentially meaningful social integration for people who already hold religious beliefs, not as a prescription for nonbelievers.
“Going to church” contains many different experiences
Attendance bundles together activities that researchers cannot always separate cleanly. A person may sing with other people, follow a weekly rhythm, hear a coherent story about suffering, receive meals during an illness, volunteer, care for children, see friends, and occupy a recognized role in a community. Any combination of those experiences could matter.
The content of the congregation matters too. A welcoming church and a coercive church are both counted as religious attendance. A community that supports therapy is not equivalent to one that treats mental illness as weak faith. A congregation that affirms a person’s family is not equivalent to one that makes belonging conditional on hiding part of the self.
The possible harms are not merely theoretical. A 2025 systematic scoping review of clergy sexual abuse literature described spiritual harm as involving intertwined spiritual, emotional, and psychological effects. The review focused on a severe and specific form of abuse, so it should not be generalized to ordinary church participation. It does establish why studies of average benefits cannot be used to overrule the experience of someone harmed in a religious setting.
Can secular communities provide the same benefits?
Some functions of church are plainly available elsewhere. Regular gatherings, mutual aid, music, shared meals, volunteering, intergenerational friendships, and moral commitments do not require belief in God. That does not mean a book club or humanist group automatically reproduces the density of a long-established congregation. Churches often coordinate several forms of support in one place and repeat them over decades.
Research on nonreligious people is beginning to examine that complexity rather than treating everyone without a religion as one group. A 2025 national US survey studied identity, time spent nonreligious, involvement in nonreligious organizations, and whether nonreligion felt comforting or anxiety-producing. The strongest predictor of wellbeing among nonreligious respondents was that affective orientation, and participation in nonreligious organizations helped shape it over time.
That finding does not prove that secular groups duplicate every benefit of church. It does show that “nonreligious” is not a mental-health condition. Identity, social environment, time, and the emotional meaning of leaving all matter.
The strongest Christian response
A thoughtful Christian response is that churches are not clinics and worship is not supposed to be judged only by its effect on mood. Christians gather because they believe God exists, Christ was raised, worship is owed, and the church is a covenant community. Mental-health benefits may follow, but they are not the truth test of Christianity. A faithful church should also recognize that illness is not evidence of weak belief and that professional care can be necessary.
That answer avoids promising that worship will make everyone psychologically healthier. It also puts the historical and theological claims back where they belong. If Christianity is true, it is not true because attenders receive favorable survey scores. If it is false, social benefits do not make its supernatural claims true.
What the evidence means after leaving
Leaving church can remove more than doctrine. It can remove a calendar, a vocabulary for grief, familiar music, adults who watched you grow up, people who notice an absence, and a place where help is organized. Losing those things can hurt without showing that the beliefs were correct.
The practical question is not whether every former Christian should return. It is which functions disappeared and which are worth rebuilding. One person may need recurring social contact. Another may miss ritual or service. Someone else may need distance from religious settings before any community feels safe. Therapy, mutual-aid groups, secular organizations, sports, neighborhood work, choirs, classes, and friendships can each replace part of what a congregation once supplied, though rarely all at once.
The best current evidence does not support using church attendance as a universal mental-health prescription. It supports a more ordinary conclusion: stable relationships, meaningful participation, supportive communities, and coherent routines often help people. Churches sometimes provide them unusually well. They can also undermine them. The task after belief is to identify the difference.