Objectives To examine if there is a social gradient in early childhood head injuries among UK children.

Methods Cross-sectional study, using data from the UK Millennium Cohort Study (MCS). The second, third and fourth sweeps of the MCS were analysed separately, when children were 3, 5 and 7 years old. Logistic regression models were used to explore the associations between head injuries and family socioeconomic position (social class, household income, maternal education and area deprivation).

Results The unadjusted analyses showed different associations with socioeconomic indicators at different ages. At age 3 and 5 years, head injuries were associated with higher area deprivation, lower household income and parents not being in work or in the routine social class. At age 5 years head injuries were also associated with lower maternal education. At age 7 years only associations with area deprivation and maternal education were found. In adjusted analyses (mutually adjusted for all four socioeconomic indicators, maternal age, child age and child sex), the following associations were observed: at ages 3 and 5 years, higher levels of area deprivation were related to higher odds of head injuries. At age 3 years only, lower levels of maternal education were related to lower odds of head injuries. No social gradients were observed. At age 7 years, there were no significant associations between head injuries and any of the SEP measures.

Conclusion We observed no social gradients in early childhood head injuries. However, at ages 3 and 5 years, head injuries were more frequently reported for children living in more deprived areas.