The effectiveness of smoking cessation, alcohol reduction, diet and physical activity interventions in changing behaviours during pregnancy: A systematic review of systematic reviews

Background Pregnancy is a teachable moment for behaviour change. Multiple guidelines target pregnant women for behavioural intervention. This systematic review of systematic reviews reports the effectiveness of interventions delivered during pregnancy on changing women's behaviour across multiple behavioural domains. Methods Fourteen databases were searched for systematic reviews published from 2008, reporting interventions delivered during pregnancy targeting smoking, alcohol, diet or physical activity as outcomes. Data on behaviour change related to these behaviours are reported here. Quality was assessed using the JBI critical appraisal tool for umbrella reviews. Consistency in intervention effectiveness and gaps in the evidence-base are described. Results Searches identified 24,388 results; 109 were systematic reviews of behaviour change interventions delivered in pregnancy, and 36 reported behavioural outcomes. All smoking and alcohol reviews identified reported maternal behaviours as outcomes (n = 16 and 4 respectively), whereas only 16 out of 89 diet and/or physical activity reviews reported these behaviours. Most reviews were high quality (67%) and interventions were predominantly set in high-income countries. Overall, there was consistent evidence for improving healthy diet behaviours related to increasing fruit and vegetable consumption and decreasing carbohydrate intake, and fairly consistent evidence for increase in some measures of physical activity (METs and VO2 max) and for reductions in fat intake and smoking during pregnancy. There was a lack of consistent evidence across reviews reporting energy, protein, fibre, or micronutrient intakes; smoking cessation, abstinence or relapse; any alcohol behaviours. Conclusions The most consistent review evidence is for interventions improving dietary behaviours during pregnancy compared with other behaviours, although the majority of diet reviews prioritised reporting health-related outcomes over behavioural outcomes. Heterogeneity between reported behaviour outcomes limits ability to pool data in meta-analysis and more consistent reporting is needed. Limited data are available for alcohol interventions in pregnancy or interventions in low-or middle-income-countries, which are priority areas for future research.