Eating disorders hit 1 in 7 men by age 40, model estimates

Eating disorders are common psychiatric conditions carrying high mortality, yet good data on how they start, recur and get treated over a lifetime are thin. Surveys catch people at a single moment, which makes it hard to say how many will ever be affected or when risk peaks. That matters for deciding where prevention effort should go and at what age. It also matters for men, who are often left out of the mental picture of who develops an eating disorder. This study tried to fill the gap by modelling the whole course of these conditions from birth to age 40.

The researchers built an individual-level state transition model and calibrated it in April 2019 using a Bayesian approach that pulled together the available clinical and epidemiological data. The model was tuned to nationally representative US survey data from 2007 and 2011. A virtual cohort of 100 000 individuals, half of them male, was followed from birth to age 40 for four diagnoses: anorexia nervosa, bulimia nervosa, binge eating disorder, and other specified feeding and eating disorders. Outcomes were age-specific 12-month and lifetime prevalence, plus deaths per 100 000 people by age 40. They ran 1000 simulations to capture uncertainty and reported uncertainty intervals around every estimate.

Estimated annual prevalence peaked at roughly age 21 for both sexes, at 7.4% for male individuals and 10.3% for female individuals. Lifetime estimates rose to 14.3% for male individuals and 19.7% for female individuals by age 40, which is about 1 in 7 and 1 in 5 respectively. Ninety-five percent of first-time cases occurred by age 25, concentrating onset in adolescence and young adulthood. On mortality, current treatment coverage was estimated to avert 41.7 deaths per 100 000 people by age 40, while treating all cases could avert 70.5. Every one of these figures carried a wide uncertainty range.

The estimates sit well above the impression given by clinic caseloads, which is the point: modelling attempts to count people who never reach services. The male figures in particular cut against the assumption that these are overwhelmingly conditions of young women. The study also complicates a purely youth-focused prevention strategy, since it estimates substantial recurring illness at older ages even though onset clusters early. What it cannot do is verify itself: a model inherits the biases and gaps of the survey and clinical data feeding it, and the wide uncertainty intervals mean the true numbers could be considerably lower or higher.

The useful takeaway is one of recognition rather than action. If these estimates are anywhere near right, eating disorders are common enough that most people know someone affected, and a meaningful share of those people are male. The concentration of first cases in adolescence and early adulthood is a reason to take changes in eating, training or body preoccupation in that age band seriously rather than reading them as ambition. The mortality modelling suggests treatment access is not a marginal issue. None of this is a diagnostic tool, and a model cannot tell any individual anything about themselves.