This preregistered multilevel meta-analysis (k=90 studies, 80 unique samples) examined the cross-sectional and longitudinal associations between **observed** (independently coded) parent–child interaction behaviors and childhood/adolescent depression (age <19 years), as well as observed child behaviors during those interactions. Five parenting behaviors and six child behaviors were analyzed from studies identified up to September 2024.
Low parental warmth/support and high harsh control/criticism each co-occurred with and prospectively predicted childhood depression (small effects: r=−.104 and r=.118, respectively). Childhood depression did not generally precede changes in observed parenting. Children with depression showed more negative affect (r=.130), reduced autonomy (r=−.132), and lower task engagement (r=−.106) — all small effects. Parental autonomy granting, guidance/structure, and depressed affect showed no significant associations with child depression.
- Most studies included mothers only, limiting generalizability to fathers and two-parent dynamics. - Few longitudinal studies controlled for baseline depression levels, so prospective effect sizes may be over- or underestimated. - The trim-and-fill procedure for child depressed affect suggested possible publication bias, rendering that specific result inconclusive.
Target **parental warmth and criticism** in prevention and intervention programs for childhood depression — these behaviors show small but consistent observed (not just perceived) associations with depression onset. Family interventions should also validate and address children's increased negativity, reduced autonomy, and low engagement, which are observable signals of depression that can strain parent–child relationships.
Explore related topics
