RELAX digital therapy lowered anxiety in pregnancy in a phase 2b trial of 268 people

A team led by King's College London has reported results of RELAX (Reducing Levels of Anxiety), a phase 2b randomized clinical trial of a self-led digital intervention for pregnant people. The paper appeared in JAMA network open. The trial randomized 268 participants in the UK at 16 to 28 weeks' gestation who had high repetitive negative thinking (RNT) and up to moderate anxiety. Recruitment ran from June 2023 to July 2024, and the final assessment took place in April 2025 doi.orgJAMA network opendoi.orgOpen source ↗.

Anxiety disorders affect 1 in 5 pregnant women and birthing people. RNT is a risk factor for anxiety, and the intervention targets it.

Intervention and trial design

Participants were randomized to RELAX plus usual care or to usual care alone; 134 people were allocated to the RELAX group. The course consisted of twelve 17-minute sessions over 4 weeks. In each session, participants worked through scenarios of ambiguous everyday situations, some related to pregnancy, and were trained to generate more positive interpretations doi.orgJAMA network opendoi.orgOpen source ↗.

People with a current psychiatric diagnosis or suicide risk were not enrolled. Assessments took place at baseline and at 4, 8 and 36 weeks after randomization; the last point corresponds to 12-24 weeks after birth. The primary outcome was anxiety on the GAD-7 scale at 8 weeks. Analysis used a modified intention-to-treat approach. In the RELAX group, 81.3% completed at least 10 of the 12 sessions.

Results

At 8 weeks, anxiety was lower in the RELAX group: the mean GAD-7 difference was -1.5 (95% CI, -2.4 to -0.6; P = .001). Depression (PHQ-9, -0.8; EPDS, -1.1), RNT (RTQ-10, -2.7), worry (PSWQ, -3.6) and perinatal anxiety (PASS, -5.5) were also lower doi.orgJAMA network opendoi.orgOpen source ↗.

At 36 weeks, three effects persisted: general anxiety (GAD-7, -1.0), postnatal depression (EPDS, -1.3) and RNT (RTQ-10, -2.4). Functioning (WSAS) did not differ at either time point. At 36 weeks, perinatal anxiety, general depression and worry also did not differ between groups.

The intervention produced more positive interpretations than usual care (mean difference, 2.1; 95% CI, 1.8-2.3). However, this change did not mediate the effect on GAD-7 at 8 weeks (indirect effect, -0.2; 95% CI, -0.4 to 0.1). The authors conclude that integrating RELAX into maternity care for people at risk of increased anxiety may improve maternal mental health.