Early predictors of childhood restrictive eating: A population-based study

Objective: Childhood eating problems, in particular restrictive eating, are common. Knowledge and understanding of risk mechanisms is still scarce. We aimed to investigate prospective early risk factors for restrictive eating across child, maternal, obstetric, and sociodemographic domains in a population-based sample of Danish 5 to 7 year olds. Method: Data on restrictive eating patterns (picky eating, slow/poor eating, and emotional undereating) collected on 1327 children from the Copenhagen Child Cohort 2000 were linked with registered and routinely collected health nurse data (during the first year of life). Prospective risk factors were investigated in univariable and multivariable regression models. Results: Feeding problems in infancy were prospectively associated with childhood picky eating (odds ratio [OR] 2.02, 95% confidence interval [CI], 1.20-3.40) and emotional undereating (OR 1.49, 95% CI, 1.05-2.11). A high thriving index in infancy was inversely associated with both picky and slow/poor eating. Having 2 non-Danish-born parents predicted slow/poor eating (OR 5.29, 95% CI, 1.16-24.09) in multivariable analyses, as did maternal diagnosis of a psychiatric disorder before child age 5 years in univariable analyses (OR 6.08, 95% CI, 1.70-21.72). Conclusions: Feeding problems and poor growth in the first year of life show high continuity into childhood restrictive eating. Maternal psychopathology is an important and modifiable risk factor. These findings confirm that early signs of poor eating and growth are persistent and might be useful in predicting eating problems in mid-childhood.