Frequently Asked Questions

Yes. In this randomized controlled trial, peer-delivered group cognitive-behavioral therapy was associated with statistically and clinically significant improvement in postpartum depression symptoms over 9 weeks. There was a significant group-by-time interaction for Edinburgh Postnatal Depression Scale (EPDS) scores from baseline to 9 weeks (F1,44.22 = 13.74, P < .01), and mean EPDS scores in the experimental group decreased by 5.4 points after treatment.

Participants randomized to the intervention were also much more likely than waitlist controls to have a clinically meaningful EPDS improvement of at least 4 points (OR = 32.14; 95% CI, 3.51 to 294.22).

Depression improved by a mean of 5.4 EPDS points in the experimental group from baseline to 9 weeks. In addition, using the Matthey Reliable Change Index criteria, 40.9% of treated participants (9/22) were classified as recovered and 54.5% (12/22) as improved at 9 weeks.

By comparison, in the waitlist control group, 6.3% (1/16) were classified as recovered and 31.3% (5/16) as improved, while 25.0% (4/16) deteriorated versus 4.5% (1/22) in the treatment group; these proportions differed significantly between groups (χ23 = 14.8, P < .01).

Yes. Among participants who had current major depressive disorder at baseline, those randomized to peer-delivered group CBT had 9 times the odds of no longer meeting diagnostic criteria for current major depressive disorder at 9 weeks compared with waitlist controls (OR = 9.00; 95% CI, 1.14 to 71.04).

At baseline, current major depressive disorder was present in 70.4% of participants in the experimental group (19/27) and 65.4% in the control group (17/26).

Yes. Peer-delivered group cognitive-behavioral therapy was associated with greater improvement in anxiety symptoms than waitlist control. There was a statistically significant group-by-time interaction for GAD-7 scores from baseline to 9 weeks (F1,38.85 = 20.77, P < .01).

Within the experimental group, the main effect of time was also significant (F1,45 = 20.56, P < .01), and mean GAD-7 scores decreased by approximately 5.5 points at 9 weeks.

Yes, the improvements in depression and anxiety were stable up to 6 months in the experimental group. EPDS scores were not significantly different from 9 weeks to 6 months (mean [SD] 11.1 [4.4] vs 10.6 [6.1]; t15 = 0.48, P = .64), and stability was high (r = 0.81, P < .01).

Anxiety findings were similar: GAD-7 scores were not significantly different from 9 weeks to 6 months (7.56 [4.27] vs 8.13 [4.87]; t15 = 0.82, P = .42), with high stability over time (r = 0.83, P < .01).

Not immediately after treatment. The trial found no statistically significant group-by-time interaction for perceived social support from baseline to 9 weeks, and no statistically significant group-by-time interactions for any Postpartum Bonding Questionnaire subscales over that same period.

However, within the experimental group, some mother-infant relationship measures improved by 6 months. The impaired bonding subscale improved from a mean (SD) of 10.81 (3.49) at 9 weeks to 6.69 (4.06) at 6 months (t15 = 3.72, P ≤ .01), and the rejection and pathological anger subscale improved from 6.38 (2.66) to 3.69 (3.32) (t15 = 3.01, P ≤ .01).

The intervention was a 9-week group cognitive-behavioral therapy program delivered in weekly 2-hour sessions by 2 peer facilitators who had recovered from postpartum depression. The first hour focused on instruction and practice of core CBT skills, and the second hour was unstructured discussion on topics relevant to postpartum depression, such as sleep and supports.

Core cognitive skills, including cognitive restructuring, were introduced in week 1. Behavioral techniques such as behavioral activation, relaxation, and goal-setting were introduced in week 2 and continued throughout the program.

Peer facilitators were women who had recovered from postpartum depression and had depression and anxiety scores below clinical cutoffs at recruitment. Ten peers were selected after a written application and telephone interview.

Their training included 2 days of classroom instruction, a 9-week observership of the CBT intervention delivered by 2 experienced therapists in a hospital setting, and weekly 1-hour supervision from an experienced therapist who reviewed session recordings. Facilitators delivered the intervention in pairs.

Yes, the intervention appeared feasible and acceptable within this study. Eighty-four percent of participants (38/45) attended 5 or more of the 9 weekly sessions, and 10 groups were delivered with a mean of 5 participants assigned to each group.

Therapeutic relationship ratings were also favorable. On the STAR-P, participants reported high positive collaboration with peer facilitators (mean [SD] 21.47 [2.92] out of 24), high positive peer facilitator input (10.35 [2.03] out of 12), and low non-supportive peer facilitator input (11.24 [1.35] out of 12, where the authors interpreted results as indicating low non-supportive input).

Participants were women aged 18 years or older who had an infant younger than 12 months, were fluent in English, and had an EPDS score of 10 or higher. They had to be free of bipolar disorder, psychotic disorders, and current substance use disorders on the Mini-International Neuropsychiatric Interview.

Of 105 women screened, 73 met eligibility criteria and were randomized: 37 to the experimental group and 36 to the waitlist control group.

The main limitations were a relatively small sample, substantial loss to follow-up, limited ethnic diversity, and use of a waitlist rather than placebo control. Recruitment stopped before the target sample was reached because of COVID-19, and follow-up completion fell from 54 participants at baseline to 38 at 9 weeks and 17 experimental participants at 6 months.

The authors also noted that the study took place in Canada, where universal health care is available, which may limit applicability to other settings. Participants were allowed to receive usual care during the study, peer leader adherence to the CBT model was not measured, and the study could not assess whether outcomes differed by individual peer facilitators.