Weight status and pathways to eating disorder referral from primary care in England and Northern Ireland: a cross-sectional analysis of routine care data.

BACKGROUND: There is a need to understand how weight status influences access to eating disorder services. We aimed to examine characteristics of adults with eating disorder diagnoses in primary care settings in England and Northern Ireland and to assess the association between weight status and referral to specialist eating disorder services. METHODS: We conducted a cross-sectional analysis of electronic health records from Clinical Practice Research Datalink (CPRD) Aurum, which includes general practices in England and Northern Ireland, with an observation period of Jan 1, 2010, to Dec 31, 2023. We included people aged 18-80 years with at least one eating disorder diagnosis with or without an eating disorder referral, and excluded people with only an eating disorder referral without an eating disorder diagnosis. All data were extracted from CPRD Aurum except Index of Multiple Deprivation, which was extracted from the UK Office of National Statistics. To examine the association between the primary exposure of weight status (ie, with underweight vs without underweight) and the main outcome of eating disorder referrals from primary care, we conducted six sets of generalised estimating equation models accounting for practice-level clustering. All primary models were done with a complete-case analysis, with participants excluded if data were missing for any included variables. To explore potential multiplicative effects of the two exposure variables, weight status and eating disorder diagnostic category, on our outcome variable of eating disorder referral, we included an interaction term between weight and eating disorder diagnosis. FINDINGS: 35 404 people with at least one eating disorder diagnosis aged 18-80 years were included in our analysis, of whom 26 822 (75·8%) did not receive an eating disorder referral. Most people identified as White (29 545 [83·5%] of 35 404), followed by Asian (2154 [6·1%]), unknown (1325 [3·7%]), Black (1028 [2·9%]), mixed/multiple (682 [1·9%]), and other (110 [0·3%]). There were 28 944 (81·8%) female individuals and 6460 (18·2%) male individuals. The interaction term between anorexia nervosa diagnosis and underweight was positively associated with receiving an eating disorder referral (odds ratio 4·05 [95% CI 3·54-4·63]; p<0·0001). The interaction term between having an avoidant restrictive food intake disorder diagnosis and underweight was not statistically significant (0·84 [0·53-1·32]; p=0·54). Similarly, the interaction term between having a binge eating disorder diagnosis and underweight was not statistically significant (0·62 [0·10-3·88]; p=0·61). The interaction terms between having an eating disorder diagnosis and underweight for bulimia nervosa (0·76 [0·64-0·90]; p=0·0018), other specified feeding or eating disorder (0·76 [0·64-0·90]; p=0·0018), and generic eating disorder (0·72 [0·65-0·80]; p<0·0001) were negatively associated with receiving an eating disorder referral. INTERPRETATION: Findings from this analysis indicate that people in England and Northern Ireland might encounter barriers to accessing eating disorder treatment within the National Health Service, where specialist care often requires General Practitioner referral. Disparities in eating disorder referral by diagnosis indicate a need to improve recognition of eating disorders beyond anorexia nervosa. FUNDING: Medical Research Council, Arts and Humanities Research Council, Economic and Social Research Council, and the UK National Institute of Health Research.