Background About half of lone-parent families in the UK live in relative poverty (income below 60% of the national median) compared to a quarter of two-parent households, and family hardship is associated with poorer child mental health. There is little evidence on whether reducing income differences between lone- and two-parent households could reduce inequality in child mental health problems (CMHP). Using data from the UK Millennium Cohort Study (∼18000 children born 2000–2002), we investigated whether equalising income between lone- and two-parent households could reduce prevalence and inequality in CMHP. We also simulated achievement of a government target to reduce child poverty to less than 10% of the population (Child Poverty Act 2010), to assess whether tackling low income could reduce inequality in CMHP related to family structure.

Methods Exposure was family structure (lone- or two-parent household) when the child was 9m; mediator was equivalised household income per week at 3y; outcome was parent-report CMHP at 5y (Strengths and Difficulties Questionnaire; borderline-abnormal, yes/no). We accounted for attrition to MCS at 5y, and baseline and intermediate confounders. Analyses were carried out in an analytic sample of 11,193.

We estimated the controlled direct effect of family structure on CMHP in logistic marginal structural models, weighted for attrition, and adjusted for confounding and mediation by household income. Prevalence was assessed overall and according to family structure; inequalities were computed using risk ratios (RRs) and differences (RDs) [95% CIs]. We re-estimated the controlled direct effect after increasing income according to two simulations. In the first, we equalised income between lone- and two-parent households. In the second, we modelled achievement of the child poverty target by increasing income for some lone-and two-parent families, reducing the proportion of households living in relative poverty to less than 10%.

Results Prevalence of CMHP was 8.5% [7.6, 9.5], and children from lone-parent households were more likely to exhibit poorer mental health (RR 1.73; RD 5.70). Equalising income between lone- and two-parent households reduced prevalence (8.2% [7.3, 9.0]), and inequality (RR, 1.37; RD, 2.86) in CMHP. Reducing child poverty in both lone- and two-parent households also decreased prevalence in CMHP (7.9% [7.0, 8.9]), but resulted in smaller reductions in inequality due to family structure (RR 1.63; RD 4.60). Sensitivity analyses showed that associations between exposure, mediator and outcome were comparable across MCS sweeps.

Conclusion Inequalities in CMHP between lone- and two-parent families in the UK are large, but could be reduced by tackling income inequality.