Background There is a paucity of literature investigating the association between maternal mental health and child injury risk, with many potential confounders difficult to measure and adjust for (e.g. supervision, safety practices). Using a self-controlled case series (SCCS), a within person design where individuals act as their own control, we aimed to investigate the temporal association between child poisoning rates and episodes of maternal depression and anxiety.

Methods Using a cohort of 209,418 mother-child pairs from England who had linked primary care and hospitalisation data from the Clinical Practice Research Datalink and Hospital Episode Statistics, we conducted a SCCS analysis of 2,646 children aged 0–4 who had one or more recorded poisoning between 1997 and 2014. Conditional Poisson regression was used to estimate incidence rate ratios (IRR) for child poisonings during medicated and unmedicated maternal depression and anxiety periods.

Results Child poisoning incidence was higher but IRRs were not significant in the 60 days before depression and/or anxiety episodes compared with well periods. There was no significant increase in poisoning risk during episodes of maternal depression (IRR 1.11, 95% confidence interval 0.73–1.69), depression with anxiety (1.19, 0.71–1.97), or anxiety (1.19, 0.61–2.31) when the mother was not prescribed medications for these conditions. Poisoning risk was however significantly elevated during periods of maternal depression treated with medication, with a 43% higher poisoning rate compared to well periods (1.43, 1.14–1.81).

Conclusions For children who experienced a poisoning, poisoning risk was increased during periods when the mother was treated with medications for depression compared to periods when the mother was well. This finding may reflect that periods of depression treated with medication may be more severe, but does support the need for preventative interventions to ensure safe medication storage and use.