Treatment of maternal depression in a medication clinical trial and its effect on children

Abstract

Objective: Observational studies show that when a depressed mother's symptoms remit, her children's psychiatric symptoms decrease. Using randomized treatment assignment, the authors sought to determine the differential effects of a depressed mother's treatment on her child.

Method: The study was a randomized double-blind 12-week trial of escitalopram, bupropion, or the combination of the two in depressed mothers (N=76), with independent assessment of their children (N=135; ages 7-17 years).

Results: There were no significant treatment differences in mothers' depressive symptoms or remission. Children's depressive symptoms and functioning improved significantly among those whose mothers were in the escitalopram group (compared with those whose mothers were in the bupropion and combination treatment groups). Only in the escitalopram group was significant improvement of mother's depression associated with improvement in the child's symptoms. Exploratory analyses suggested that this may be due to changes in parental functioning: Mothers in the escitalopram group reported significantly greater improvement, compared with the other groups, in their ability to listen and talk to their children, who as a group reported that their mothers were more caring over the 12 weeks. Maternal baseline negative affectivity appeared to moderate the effect of maternal treatment on children, although the effect was not statistically significant. Children of mothers with low negative affectivity improved in all treatment groups. Children of mothers with high negative affectivity improved significantly only for those whose mothers were in the escitalopram group.

Conclusions: The effects of the depressed mother's improvement on her children may depend on her type of treatment. Depressed mothers with high anxious distress and irritability may require medications that reduce these symptoms in order to show the effect of her remission on her children.

Trial registration: ClinicalTrials.gov NCT00519428.

Figure 1

STAR*D-2 Flowchart

Figure 2. Estimated Trends in Mean Children’s Depression Inventory (CDI) Scores Over 12 Weeks by Maternal Treatment Assignment

Note. Figure shows means for 54 children of 27 combination-treated mothers, 35 children of 20 bupropion-treated mothers, and 46 children of 29 escitalopram-treated mothers. Lower scores indicate improvement. All analyses were adjusted for child age (centered) and gender, site, and within-family correlation. Time trends by treatment imply that children of escitalopram-treated mothers improved the most (Beta=−0.39, SE=0.06, t=−6.27, p<0.0001), followed by children of bupropion- (Beta=−0.20, SE=0.07, t=−2.82, p=0.006) and combination-treated mothers (Beta=−0.06, SE=0.06, t=−0.93, p=0.36). The overall week × treatment interaction was statistically significant (F=7.28, df= 2,227, p<0.001), implying that the change in child depression score over time differed significantly between treatment groups. Pairwise comparisons showed significant differences between children of mothers on combination vs. escitalopram treatment (Beta=0.33, SE=0.09, t=3.81, p=0.0002) and bupropion vs. escitalopram treatment (Beta=0.19, SE=0.09, t=2.04, p=0.04), and no significant differences between children of mothers on combination vs. bupropion treatment (Beta=−0.14, SE=0.09, t=1.51, p=0.13).

Figure 3. Estimated Trends in SAS-SR Parenting Item “Been able to talk to and listen to your child” Over 12 Weeks

Note. Figure shows means for 27 combination-treated mothers, 20 bupropion-treated mothers, and 29 escitalopram-treated mothers. Lower scores indicate improvement. All analyses were adjusted for site. Time trends by treatment imply that mothers on escitalopram improved the most (Beta=−0.10, SE=0.02, t=−5.83, p= <0.0001), followed by mothers on bupropion (Beta=−0.05, SE=0.02, t=−2.94, p=0.004) and combination treatment (Beta=−0.03, SE=0.01, t=−2.35, p=0.02). The overall week × treatment interaction was statistically significant (F=4.12, df=2,150, p=0.02), implying that change in scores on SAS-SR parent item #2 over time is significantly different between treatment groups. Pairwise comparisons showed significant differences between mothers on combination vs. escitalopram treatment (Beta=0.06, SE=0.02, t=2.82, p=0.005), marginally significant differences between mothers on bupropion vs. escitalopram (Beta=0.05, SE=0.02, t=1.92, p=0.057), and no significant differences between mothers on combination vs. bupropion treatment (Beta=0.02, SE=0.02, t=0.73, p=0.47).