Obesity Abstracts

Background: Genotype has an effect on body weight that starts early in life and increases with time. Eating behaviour may be a mediator.

Methods: We used a 97 single-nucleotide polymorphisms (SNPs) polygenic score (PGS) to index genetic risk and the Adult Eating Behaviour Questionnaire (AEBQ), to measure 8 appetite traits among 2593 Avon Longitudinal Study of Parents and Children (ALSPAC), when they were on average 23.8 (standard deviation (S.D.): 0.5) years old. We used regression-based mediation analysis to estimate the extent each trait mediated the association between PGS and BMI at age 24.4 (S.D.:0.7). We used multiple imputation to impute missing BMI and covariates and adjusted for age, sex and 20 ancestry-informative principal components. We repeated our analyses, adjusting for parental socioeconomic position, BMI, depression, eating disorders, smoking and breastfeeding, as well as for offspring personality, depression, physical activity and smoking.

Results: In complete-case, minimally adjusted analyses, a 1 S.D. higher PGS was associated with a 0.934 (95%CI 0.7, 1.2) kg/m² higher BMI. A 1-unit higher score in emotional overeating, was associated with a 1.2 (95%CI 1.0–1.5) kg/m² higher BMI. Emotional overeating mediated 7.2 (95% CI 1–13.3) % of the overall association. Food responsiveness was associated with higher BMI and satiety responsiveness, slowness in eating and emotional under-eating were associated with lower BMI, but there was limited evidence of mediation (proportion mediated 1.9–2.5%, P 0.09–0.515). Results were similar for multiple imputation, most adjusted analyses.

Conclusions: Targeting emotional overeating may lower genetic inequalities and reduce BMI. The small proportion mediated by most traits suggests that genetic makeup acts through other pathways. Studies with repeated and more accurate appetite traits measures are needed to strengthen causal inference regarding their associations with BMI.

Disclosures: EM is supported by a PhD studentship from GW4 MRC Biomed DTP (MR/N0137941/1). AEBQ data collection was funded by LDH’s MRC Population Health Scientist Fellowship (G1002375) and LJ’s unrestricted research grant from Kellogg Europe. LJ has had funding from Danone Baby Nutrition, the Alpro foundation and various UK research councils. TMF has consulted for Sanofi, Servier, and Boehinger Ingelheim, and holds an MRC CASE studentship sponsored by GSK.