Background: Childhood obesity has an uneven socioeconomic distribution. In health interventions, equity aspects are crucial. In 2007–2008, the IDEFICS/Sweden health survey (Identification and prevention of dietary- and lifestyle-induced health effects in children and infants) was performed in children aged 2–9 years, followed by a health promoting community intervention.

Aims: To assess if families with disadvantageous socioeconomic and -demographic backgrounds and children with obesity were underrepresented.

Methods:  IDEFICS participants (N = 1,825) were compared with a referent child matched for age, gender and municipality (N = 1,825) by using registers from Statistics Sweden. Longitudinal growth data from birth to 8 years was collected from local health services (n = 3,650) to compare children's BMI standard deviation score (SDS) at age of inclusion in the IDEFICS study (BMI SDS-index) and the children's BMI-categories during the age-span between the groups.

Results: The referent population had significantly lower education and income, more often foreign backgrounds, single parenthood and mother-smokers than the IDEFICS population. BMI SDS-index at inclusion did not differ between groups but the obesity prevalence differed significantly at age 8 years (referents: 4.5% vs. IDEFICS: 2.9%). In the multivariate adjusted model, parental Swedish background showed the strongest association to participation (odds ratio = 1.91, 95% confidence interval (1.48–2.47).

Conclusion:  Children with disadvantageous socioeconomic and -demographic backgrounds were underrepresented in the Swedish IDEFICS study. BMI at inclusion had no selection effect, but developing obesity was significantly greater among referents. Selection biases are important to consider when interpreting results from health surveys and in planning intervention programmes.