Table 1. Definitions of explanatory variables.

Variable	Definition	Timing of data collection(age of child)a
Individual-level
Male sex	Yes/no	5 months
Overeating phenotype	‘Often’ eats too much and/or ‘sometimes’ or ‘often’ eats too fast	4 years
Breast-fed exclusively ≥ 3 months	Yes/no	5 months
Mother smoked during pregnancy	Yes/no	5 months
Mother’s obesity status (WHO)	BMI ≥30 based on self-reported height and weight	1.5 years
High birth weight	>4 kg vs ≤ 4 kg	5 months
Rapid weight gain 0–5 months	Highest two quintiles of average monthly weight gain from 0 to 5 months	5 months
Family/household
Low socioeconomic status (SES)b c	Lowest third of SES indicator versus middle and upper tertiles	Time-dependent
Single parent family	Yes/no	Time-dependent
Mother is an immigrant	Yes/no	5 months
Local environment
Living locationb	Categorical (Urban: Census Metropolitan Area with ≥ 100,000 inhabitants; Semi-urban: Census Agglomerations with 10,000 to <100,000 inhabitants; and Rural: Rural or Small Towns with <10,000 inhabitants)	Time-dependent
High social cohesiond	Scale score in the bottom 50%	Time-dependent
High disorderd	Scale score <3	Time-dependent
Materially deprived	Two highest factorial score quintiles (4 and 5)	5 months
Socially deprived	Two highest factorial score quintiles (4 and 5)	5 months

^aTime-dependent indicates that these variables were available at 4, 6, 7, 8, and 10 years of age and thus were treated as time-dependent explanatory variables in the analysis. If a variable was only measured once and occurred at or before baseline (4 y) it was treated as a ‘risk factor’ (time-stable).

^bMissing at 4 y of age for all children, value at age 3.5 y was carried forward to age 4.

^cFor more information on how this variable was calculated and interpreted, please see reference 32.

^dMeasured every other data collection cycle for all children (value at age 6 was carried forward for age 7).