Re-examining premature mortality in anorexia nervosa: a meta-analysis redux.
Level 3 - non-randomized controlled study
Systematic review and meta-analysis of observational cohort studies
PubMed 25214371 · doi:10.1016/j.comppsych.2014.07.017
What was done
Meta-analysis of 41 cohorts across 40 peer-reviewed studies (published 1966–2010) evaluating premature mortality in anorexia nervosa (AN). The authors used double data extraction, log-linear regression with an over-dispersed Poisson model, and calculated all-cause and suicide-specific standardized mortality ratios (SMRs) with 95% Poisson confidence intervals relative to general population females aged 15–34 years.
What was found
Participants with AN had an all-cause premature mortality SMR of 5.2 (95% CI: 3.7–7.5) and a suicide-specific SMR of 18.1 (95% CI: 11.5–28.7). These values were 10% and 49% lower, respectively, than earlier published SMR estimates. Premature mortality risk was highest in studies with older participants, though confounding by treatment was present; risk was also influenced by gender, ascertainment methods, and diagnostic criteria.
Why it matters
This study provides methodologically corrected, less inflated estimates of mortality in anorexia nervosa while confirming that premature death and suicide risks remain substantially elevated compared to the general population.
Limits
The underlying studies span over 40 years with shifting diagnostic criteria and ascertainment strategies. Total participant count was not reported in the abstract, and treatment confounding was noted among older cohorts.