Interpretation
We found no difference in psychological well-being at 29–34 weeks of gestation between the intervention group and the control group. In contrast a systematic review from 2018 concluded that prenatal exercise reduces the odds and the severity of prenatal depressive symptoms (29). The authors included thirteen studies (n=1,076) in the analysis, seven of these had a population at risk of or currently diagnosed with depression (29). Our result might be explained by a comparable level of general physical activity at 29–34 weeks of gestation in both the intervention and the control group, median physical activity being 4 hours (min 0–max 24) in the intervention group and 4 hours (min 0–max 16) in the control group. This indicates that the intervention did not help women to increase their physical activity rather just changed the physical activity behavior, which may be explained by the womens high level of physical activity preconceptionel in this study population (30). The perprotocol analysis of women attending ≥75% of the exercise sessions showed a statistically significant higher WHO-5 mean in the intervention subgroup relative to the control group (29). While we, based on the literature (22), predefined a mean difference between the two groups of minimum 7.75 point as clinically significant (19). Further, this is a selected subgroup and therefore it is not possible to draw conclusions from our perprotocol analysis as it can be flawed (31).
At eight weeks postpartum, women in the intervention group had a statistically significant higher mean WHO-5 than women in the control group signifying a greater level of psychological well-being. In line with our result, a systematic review found that physical activity during pregnancy reduced the risk of postpartum depressive symptoms, however, in a population of women at average risk for depression (32). In contrast, Davenport (29) found that exercise during pregnancy did not have an effect on postpartum psychological well-being, however the exercise interventions were primarily homebased exercise. Our positive results postpartum may be explained by psychosocial factors such as increased perceived peer-support and reduced loneliness, as previously reported for women in the intervention group (33). This is in line with other studies which imply that peer support may reduce the risk of postpartum depression by enabling sharing of experiences, reassurance that other mothers experience similar feelings, and a sense of belonging (34,35).
Regarding maternal delivery outcomes we did not conform our hypothesis that the intervention group would have a significant shorter duration of labour (36). This may be due to a high level of physicial activity in the control group. An interesting finding was that only 15.8% of women in the intervention group had labor induced, opposed to 27.2% in the control group. Although this is in line with findings from a prospective cohort study (37) and a non-randomised intervention study (38), it may represent a spurious finding and we interpret this with caution.