Background Maternal employment in the UK has increased significantly in recent years, but evidence of its relationship with child socio-emotional behaviour is mixed. Given the policy importance of both family employment and children’s wellbeing, this relationship needs to be better understood. This study seeks to investigate whether cumulative exposure to maternal employment is associated with socio-emotional behaviour at age 7.

Methods The study comprises a longitudinal analysis of the United Kingdom Millennium Cohort Study (MCS), based on 10,723 singleton children who participated in all four sweeps of the study (at ages 9 months, 3, 5 and 7 years). Socio-emotional behaviour was measured using total Strengths and Difficulties Questionnaire (SDQ) scores, dichotomised using an established cut-off for borderline / abnormal behaviour. Risk ratios (RR) were estimated using Poisson regression for borderline / abnormal SDQ scores at age 7 according to cumulative maternal employment over four sweeps of the MCS (at 9 months, 3, 5 and 7 years), unadjusted (uRR) and adjusted (aRR) for confounders (age at first live birth, ethnicity, family size, lone parenthood, and maternal psychological distress, all recorded at 9 months). In order to disentangle whether there was a differential impact of earlier and current employment, additional analyses estimated the risk associated with employment up to age 5 years, adjusting for employment at age 7.

Results The risk of borderline/abnormal scores decreased linearly by number of MCS sweeps that mothers were in employment; after adjustment for confounders the gradient attenuated but remained significant. Children growing up in households where mothers were employed in all MCS sweeps were significantly less likely to display borderline / abnormal behaviour at age 7 (compared to never in paid employment): uRR=0.37 (0.32–0.44); aRR=0.68 (0.56–0.81). Analyses differentiating early employment (between ages 9 months and 5 years) and current employment (at age 7) suggest that the cumulative impact of prior non-employment may be ameliorated once a mother gains employment. After adjusting for current employment and confounders, there was no increased risk of borderline / abnormal socio-emotional behaviour associated with non-employment in all earlier sweeps up to age 5 compared to being in paid employment continuously (baseline): aRR=1.01 (0.82–1.23).

Conclusion This longitudinal analysis supports an association between maternal employment and child socio-emotional behaviour. Possible explanations and implications will be discussed.