Regional variation in lifestyle patterns and BMI in young children: the GECKO Drenthe cohort

Table 4 Prospective associations of the lifestyle patterns with standardized body mass index at 10–11 years

	Model 1^a		Model 2^b		Model 3^c		Model 4^d
	B	95% CI	B	95% CI	B	95% CI	B	95% CI
Lifestyle patterns
‘High activity’	0.005	(− 0.02; 0.03)	0.002	(− 0.03; 0.03)	0.003	(− 0.03; 0.03)	0.002	(− 0.03; 0.03)
‘Low screen time, high sleep and healthy diet’	− 0.076	(− 0.11; − 0.04)	− 0.072	(− 0.11; − 0.04)	− 0.071	(− 0.11; − 0.03)	− 0.071	(− 0.11; − 0.03)
‘High outdoor play’	0.036	(− 0.01; 0.09)	0.031	(− 0.02; 0.08)	0.030	(− 0.02; 0.08)	0.031	(− 0.02; 0.08)
Socio-economic status
Equivalized household income indicator	–	–	− 0.054	(− 0.09; − 0.02)	− 0.050	(− 0.09; − 0.02)	− 0.052	(− 0.09; − 0.02)
Spatial components
Rho (spatial term for zBMI 10–11 years)	–	–	–	–	0.004	(− 0.84; 0.84)	–	–
lag_SES	–	–	–	–	− 0.054	(− 0.18; 0.07)	–	–
lag_’high activity’	–	–	–	–	− 0.021	(− 0.11; 0.07)	–	–
lag_’low screen time, high sleep and healthy diet’	–	–	–	–	− 0.022	(− 0.14; 0.10)	–	–
lag_’high outdoor play’	–	–	–	–	− 0.034	(− 0.20; 0.13)	–	–
Lambda (spatial term for error)	–	–	–	–	0.052	(− 0.78; 0.89)	0.121	(0.02; 0.23)

Additional models were run including a dummy for moving, and moving to healthier/less healthy neighbourhoods (determined at age 5 using geographically weighted mean of zBMI 5 years, converted to tertiles). Neither variable was significantly associated with standardized body mass index at 10–11 years. Results available upon request
^aModel 1: analyses were adjusted for age, sex, weartime, energy intake, smoking during pregnancy and standardized body mass index (zBMI) at 5–6 years
^bModel 2: Model 1 + Equivalized Household Income Indicator
^cModel 3: Model 2 + spatial terms for the determinants, the outcome and the error term, and additional adjustment with a spatial term for zBMI at 5–6 years
^dModel 4: Model 2 + spatial term for the error

ISSN: 1476-072X