OP140 Household associations between child and adult weight status in an ethnically diverse urban population: cross-sectional study using linked primary care and National Child Measurement Programme records

Background

Parental weight status is a recognised predictor of offspring weight, however little is known about the influence of weight status of adult household members who may not be biologically related. We hypothesised that children sharing a household with at least one adult with excess weight would be more likely to live with overweight or obesity than those sharing a household with adults of healthy weight.

Methods

We linked 66,106 (98.5%) 5-year-old National Child Measurement Programme participants (2013–18) in North East London to their primary care records using pseudonymised NHS numbers, and linked household members using pseudonymised Unique Property Reference Numbers. The primary outcome was child excess weight (UK1990 clinical standards: overweight≥91st; obese≥98th centiles). Main explanatory variable: household weight status (obese:≥1adult BMI≥30 kg/m2; overweight:≥1adult 30 kg/m2>BMI≥25 kg/m2; healthy: all adults BMI<25 kg/m2). We used multivariable ordinal regression to estimate the adjusted odds [aOR] and 95% confidence intervals [CI] of child excess weight when living in overweight or obese adult households, after adjusting for child sex and ethnic group, household composition, and area-level deprivation. We conducted sensitivity analyses stratifying households by number of adults with available weight measurement. Analyses were conducted in R.

Results

Respectively 4,279 (10.4%) and 3,476 (8.5%) of 41,092 children with available household data were living with overweight or obesity. 13,091 (31.9%) children lived in overweight and 21,080 (51.3%) in obese households. Children living in overweight or obese households were more likely to be in a higher weight category (aOR 1.57 [1.43–1.72] and 2.82 [2.58–3.08] respectively), as were those from South Asian ethnic groups or living in single adult households (1.62 [1.52–1.72], and 1.33; [1.23–1.44] respectively). Children from Black ethnic groups (0.38 [0.34–0.42]) and multigenerational households (0.87; [0.80–0.94]) were less likely to be in a higher weight category. Associations attenuated and increased respectively for the subsamples of households with only one or more than one adult with available weight measurement.

Conclusion

Young children live in varied non-nuclear households with shared health and dietary behaviours influenced by caregivers who are not always biologically related. In this ethnically diverse and disadvantaged population with high levels of child and adult obesity, we found strong associations with household-level adult weight status, and composition. Strengths include novel data linkages. Weights were not available for all adult household members. Further research is needed to understand mediators of these associations and the potential of adult-only interventions to support weight management in young children living with severe obesity.