Adverse outcomes in patients with a diagnosis of an eating disorder: primary care cohort study with linked secondary care and mortality records.

Objective: To examine the short and long term adverse physical and mental health outcomes, all cause mortality, and natural and unnatural deaths in individuals with a diagnosis of an eating disorder compared with a matched cohort without these disorders. Design: Primary care cohort study with linked secondary care and mortality records. Setting: English primary care electronic health records from the Clinical Practice Research Datalink, linked to Hospital Episode Statistics for hospital records and the Office for National Statistics for mortality data, 1 January 1998 to 30 November 2018. Participants: 24 709 patients, aged 10-44 years, with a diagnosis of an eating disorder, matched by age, sex, and general practice with up to 20 comparators without an eating disorder (n=493 001). Disorders examined were anorexia nervosa, bulimia nervosa, binge eating disorder, and all other types combined. Main outcome measures: Physical health conditions, mental health conditions, and mortality (all cause and cause specific). Hazard ratios and cumulative incidences were calculated to compare outcomes. Results: Individuals with eating disorders had substantially higher risks for coded adverse physical, mental health, and mortality outcomes. Within the first year after a diagnosis of an eating disorder, these individuals were six times more likely to develop renal failure (hazard ratio 6.0, 95% confidence interval (CI) 4.2 to 8.5; excess 15 events per 10 000 individuals, 95% CI 11 to 21) and nearly seven times more likely to develop liver disease (hazard ratio 6.7, 3.8 to 11.7; excess 6 per 10 000, 4 to 11). Risks were still increased after five years (hazard ratio for renal failure 2.6, 95% CI 2.0 to 3.4; hazard ratio for liver disease 3.7, 2.3 to 6.0), with cumulative excesses of 110 (95% CI 87 to 136) and 26 (17 to 39) per 10 000 individuals at 10 years for renal failure and liver disease, respectively. Mental health coded outcomes were markedly increased in the short term for depression (hazard ratio 7.3, 95% CI 6.6 to 8.1; excess 596 per 10 000 individuals, 95% CI 545 to 650) and self-harm (hazard ratio 9.4, 8.2 to 10.7; excess 309 per 10 000, 279 to 342) at one year. Mortality was also higher: within the first year, the risk of all cause mortality was more than fourfold (hazard ratio 4.6, 95% CI 3.1 to 7.0; excess 10 per 10 000 individuals, 95% CI 7 to 15) and was fivefold for unnatural deaths (hazard ratio 5.1, 2.2 to 11.9; excess 30 per 100 000, 13 to 61). Risks persisted beyond five years (hazard ratio 2.2, 95% CI 1.8 to 2.7 for all cause mortality; hazard ratio 3.2, 1.9 to 5.4 for unnatural deaths), corresponding to 43 (95% CI 33 to 54) excess all cause deaths per 10 000 individuals and 184 (125 to 262) excess unnatural deaths per 100 000 at five years. At 10 years, 95 (95% CI 75 to 118) excess all cause deaths per 10 000 individuals and 341 (236 to 479) unnatural deaths per 100 000 were found. Suicide risk was 13.7 times higher in the first year (95% CI 4.8 to 38.8; 24 excess deaths per 100 000 individuals, 95% CI 10 to 52) and remained increased at 10 years (hazard ratio 2.7, 1.3 to 5.8), accounting for 169 (103 to 266) excess deaths per 100 000 individuals. Conclusions: Individuals with a diagnosis of an eating disorder derived from coded primary care data were more likely to have recorded adverse physical and mental health outcomes, as well as higher mortality than individuals without an eating disorder matched by age, sex, and general practice. These findings highlight the need for integrated care strategies that deal with both the physical and mental health dimensions in the management of individuals with eating disorders.