Abstract
OBJECTIVE: Prenatal exposures have been suggested to have a programming effect on neural and metabolic development, which may affect the risk of eating disorders. We investigated the association between prospectively measured maternal gestational high-sensitivity C-reactive protein (hs-CRP), an established inflammatory biomarker, and subsequent risk of AN in daughters. METHOD: This nested case-control study with sibling-comparison design used systematic sampling from a register-based cohort including all eating disorder patients in Finland born 1991-2000 and diagnosed in specialized health care. Final sample included 150 full triads of females with severe AN (ICD-10 code F50.0), age- and sex-matched population controls, and biological sister controls (total N = 450). RESULTS: Mean gestational hs-CRP values were 4.10 mg/L (SD 5.22), 4.83 mg/L (SD 4.88), and 5.53 mg/L (SD 10.36), for individuals with AN, population controls, and sister controls, respectively. Higher hs-CRP was associated with decreased risk for AN when compared to sister controls (adjusted OR 0.68, 95% Cl 0.48-0.97, p = 0.03). Analyzing hs-CRP in tertiles, maternal hs-CRP in the highest tertile (≥ 5.13 mg/L) versus lowest tertile (≤ 1.94 mg/L) was associated with decreased risk for AN compared to sisters (adjusted OR 0.35, 95% Cl 0.15-0.80, p = 0.01), and to all controls combined (adjusted OR 0.52, 95% Cl 0.29-0.93, p = 0.03). DISCUSSION: We found no evidence that higher gestational CRP would increase the later risk of AN. On the contrary, lower maternal low-grade inflammation early in pregnancy was associated with an increased risk of AN in daughters.