Avoidant/restrictive food intake disorder: Systematic review and meta-analysis demonstrating the impact of study quality on prevalence rates

Objectives: The prevalence of Avoidant/Restrictive Food Intake Disorder (ARFID) is unclear. This paper is the first to present meta-analysis based estimates of the prevalence of ARFID, and to assess the impact of the quality of the research on these estimates.

Design: A pre-registered (Prospero: CRD42023487621) systematic review and meta-analysis.

Methods: PubMed, PsychInfo, Web of Science and CINAHL were searched (final date of retrieval 30th July 2024) for peer reviewed papers published between 2013 and 2024. Random-effects and quality effects meta-analyses were used to compute and compare prevalence estimates and to evaluate the impact of study quality on prevalence rates. Subgroups were also considered (gender, age group, clinical status). Loney et al.'s (1998) Critical Appraisal of the Health Research Literature: Prevalence or Incidence of a Health Problem scale was used to assign each study a quality score across three categories - methodological validity (six points); interpretation of results (one point); and applicability of the results (one point).

Results: Twenty-six studies were identified (n = 122,861). Meta-analysis using random-effects indicated a prevalence of 11.14 % (95 % CI 8.16–14.5 %), whereas quality effects prevalence was 4.51 % (95 % CI 0.7–10.68 %). Similar contrasts were evident among subgroups.

Conclusions: Even taking the more conservative estimate of 4.51 %, this review demonstrates that ARFID is a common disorder, meriting further research and clinical and service developments. Future research needs to be more methodologically robust (larger samples; standardised diagnostic measures; clearer data presentation).