Repeat a hundred times: correlation is not causation. The odds ratios look good. But the rate of serious mental illness (Schizophrenia, Bipolar and schizoaffective disorder) was 50/100 000 in this cohort of male conscripts. Most people with low emotional stability (ie, the neurotic) and social immaturity did not develop psychosis.

Late adolescent social maturity, mental energy, and emotional stability were inversely associated with schizophrenia and other nonaffective psychotic illnesses in a dose-dependent fashion after accounting for important confounders. Bipolar disorder had a U-shaped association with social maturity and was associated with low emotional stability. Schizoaffective disorder was solely associated with low emotional stability. Associations remained after excluding men who became ill within 5 years after their personality assessment, suggesting that these results are not purely attributable to prodromal changes in personality.

I am tempted to write a letter. The odds ratios are not great, the rate of psychosis fortunately rare, and one should mistrust the significance of the differences with very large numbers. This paper fits too neatly into the fashionable idea that psychosis links with personality: clinically this is clearly not always the case. What is the case is that a large number of our early intervention services, when using criteria not the first psychotic symptoms to accept patients, end up treating a lot of personality difficulties. Bentall gets some things right in his comment, and some things very wrong.

As acknowledged by the authors, one limitation of the data is that the personality measures, although reliable, were specifically designed for the purposes of the Swedish military and did not correspond to recognized personality models, such as Eysenck’s triad6 of neuroticism (N), extroversion (E), and toughmindedness or the now widely accepted 5-factor model (or big five model), which adds openness to experience (O), conscientiousness (C), and agreeableness (A) to Eysenck’s N and E. The authors suggest that high social maturity implies high scores on C, E, and A (extroversion coupled with high levels of responsibility); high mental energy implies high scores on O and C (high levels of initiative and persistence); and emotional stability corresponds to high scores on A and especially low scores on N. The standout finding is that low ratings for emotional stability were associated with all 4 outcomes, and a plausible interpretation is therefore that neuroticism is an important risk factor for psychosis.
…
However, if interactions are discovered between personality and environmental risk factors, a targeted public health approach to preventing severe mental illness may become possible. A pioneering study in Mauritius attempted to use a nursery school–based environmental and nutritional enrichment program to prevent the development of schizotypal personality characteristics, with some promising results, although autonomic measures at 3 years (likely to be an indicator of early emotional instability) did not moderate the findings. Future studies might consider whether social policies that promote supportive family and educational environments have the potential to prevent severe mental illness in vulnerable young individuals, thereby reducing the burden of mental illness for individuals, families, and society.

Bentall forgets that the biggest risk factors known are maternal starvation and immigration. His intervention is akin to fortifying frankenfood. It is better to have a functioning society in the first place.

But even then, the risk of psychosis will remain.