This article could be summarized in two sentences, both of which should be unsurprising to most HN readers: 1. Publication bias means that the reported results

This article could be summarized in two sentences, both of which should be unsurprising to most HN readers: 1. Publication bias means that the reported results for efficacy of psychotherapy are likely to be stronger than the actual efficacy. 2. Observational studies showed a negative relationship between mental health outcomes and psychotherapy. The clickbait title aside, 1 is not really news as this bias affects every field, especially social science, and has been widely reported. 2 is basically a no-op. You cannot under a causal relationship based only on observational studies. Full stop. The idea that people who developed poor mental health sought professional help is if anything more plausible than the clickbaity alternative.