Behaviour change interventions for low-income groups: meta-analysis of behaviour change techniques, delivery and context

Abstract

Introduction Health inequalities could be reduced by effective interventions to change diet, physical activity and smoking for low-income groups. This meta-analysis explored the behaviour change techniques (BCTs) and aspects of intervention context and delivery associated with increased healthy eating, physical activity and smoking cessation for low-income adults. Methods This was a secondary analysis of a systematic review of interventions targeting change in diet, physical activity or smoking cessation (searching from 1995-2014). There were 35 RCTs containing 45 interventions with 17,000 healthy low-income adults. The association with behaviour change of 46 BCTs identified using taxonomy v1 and 13 delivery/context variables as specified by the TiDieR checklist was examined. Findings In dietary interventions, one BCT (self-monitoring) and two delivery/context features (personal contact, targeting multiple behaviours) were associated with increased effectiveness; three other BCTs (feedback on behaviour, prompts and cues, information about emotional consequences) had reduced effectiveness. In physical activity interventions, three BCTs (behavioural practice/rehearsal, instructions on how to be physically active and information about antecedents) and two delivery/context variables (home or community delivery; focus on physical activity only) were associated with effectiveness. No BCTs or delivery/context features were associated with smoking cessation. Conclusions Several BCTs and delivery/context variables were associated with greater effectiveness of diet and physical activity interventions in low-income groups. Some BCTs and delivery/context variables were associated with reduced effectiveness; many could not be evaluated. These findings add to the evidence base for interventions to improve health outcomes in low income groups.