Purpose: Poverty has been associated with high levels of behavior problems across childhood, yet patterns of associations over time remain understudied. This study aims: (a) to examine whether poverty predicts changes in behavior problems between 1.5 and 8 years of age; (b) to estimate potential selection bias for the observed associations.
Methods: We used the 1998-2006 waves of the Quebec Longitudinal Study of Child Development (N = 2120). Main outcomes were maternal ratings of hyperactivity, opposition and physical aggression from 1.5 to 8 years of age. Linear mixed-effects models were used to assess the longitudinal association between poverty and behavior problems. Models were re-estimated adjusting for wave nonresponse and using multiple imputation to account for attrition.
Results: Poverty predicted higher levels of behavior problems between 1.5 and 8 years of age. Poverty predicted hyperactivity and opposition in a time dependent manner. Hyperactivity [Bpoverty*age = 0.052; CI 95 % (0.002; 0.101)] and opposition [Bpoverty*age = 0.049; CI 95 % (0.018; 0.079)] increased at a faster rate up to age 5 years, and then decreased at a slower rate for poor than non-poor children. Physical aggression decreased at a steady rate over time for all children [Bpoverty*age = -0.030; p = 0.064). Estimates remained similar when accounting for attrition.
Conclusion: Poverty predicted higher levels of behavior problems between 1.5 and 8 years of age. The difference between poor and non-poor children was stable over time for physical aggression, but increased with age for hyperactivity and opposition. Attrition among poor children did not compromise the validity of results.
Keywords: Attrition; Hyperactivity; Opposition; Physical aggression; Poverty.