Associations between maternal depression and child behaviour problems may be biased due to depression-related distortions in the maternal reports. In this study, we compared the agreement between maternal, teacher and child ratings of hyperactivity/inattention (H/I), conduct (CD) and emotional (EM) problems measured with the Strengths and Difficulties Questionnaire in a population with depressed and non-depressed mothers. The sample of 12,961 11-year-old children was drawn from the Danish National Birth Cohort. We also examined the association between maternal depression before and during the child's life and child H/I, CD and EM problems when reported by the different informants. Agreement between the informants was assessed with the Bland-Altman limits of agreement (LOA) and associations were analyzed using multiple linear regression models. We found large discrepancies between maternal and teacher ratings of child behavioural problems measured with the SDQ and less discrepancy between maternal and child self-ratings. Better agreement between informants was found in the population of depressed mothers compared to non-depressed mothers for H/I and CD problems, and not for EM symptoms. The LOA for all three subscales were wide in both populations, with the largest intervals ranging from - 6 (Lower LOA) to 3 (Upper LOA) for H/I. Statistically significant associations were found between maternal depression, and child H/I and CD problems when reported by the mother but not the teacher and only by the child, when maternal depression had been present during the child's life. Our findings emphasize the importance of considering each informants' ratings when obtaining ratings from multiple sources.
Keywords: Behaviour ratings; Informant discrepancies; Maternal depression; Rater agreement.