Objectives

To report on long-term trends in hospital admission rates for anorexia nervosa using two English datasets.

Design

We used data on hospital day-case and inpatient care across five decades in the Oxford Record Linkage Study (ORLS), and similar data for all England from 1990. We analysed rates of admission for anorexia nervosa in people aged 10–44 years, using hospital episodes (counting every admission) and first-recorded admissions (counting only the first record for each person).

Setting

Former Oxford NHS Region; and England.

Participants

None; anonymous statistical records were used.

Results

In the longstanding ORLS, the age-standardised first-recorded admission rate for women was 2.7 (95% confidence interval 1.6–3.8) per 100,000 female population aged 10–44 years in 1968–1971; 2.7 (2.1–3.3) in 1992–1996; and 6.3 (5.5–7.2) in 2007–2011. Male rates were zero in the 1960s; 0.07 (0.0–0.1) per 100,000 men in 1992–1996; and 0.4 (0.2–0.6) in 2007–2011. In England, female rates increased from 4.2 (4.0–4.4) in 1998–2001 to 6.9 (6.7–7.1) in 2007–2011; and the corresponding male rates were 0.2 (0.1–0.3) and 0.5 (0.4–0.6). Episode-based admission rates rose more than person-based rates. The highest rates by far were in girls and women aged 15–19 years.

Conclusions

In recent years, anorexia nervosa has become a greater burden on secondary care: not only have admission rates increased but so too have multiple admissions per person with anorexia nervosa. The increase in admission rates might reflect an increase in prevalence rates of anorexia nervosa in the general population, but other explanations, including lower clinical thresholds for admission, are possible and are discussed.