Health behaviour change in cardiovascular disease prevention and management: meta-review of behaviour change techniques to affect self-regulation

Abstract

Self-regulation processes assume a major role in health behaviour theory and are postulated as important mechanisms of action in behavioural interventions to improve health prevention and management. The need to better understand mechanisms of behaviour change interventions for cardiovascular diseases (CVD) called for conducting a meta-review of meta-analyses for interventions targeting self-regulation processes. The protocol, preregistered on Open Science Framework (OSF), found 15 eligible meta-analyses, published between 2006 and August 2019, which quantitatively assessed the role of self-regulatory mechanisms and behaviour change techniques (BCTs). Quality of the meta-analyses varied widely according to AMSTAR-2 criteria. Several BCTs, assumed to engage self-regulatory mechanisms, were unevenly represented in CVD meta-analytic reviews. Self-monitoring, the most frequently studied self-regulatory BCT, seemed to improve health behaviour change and health outcomes but these results merit cautious interpretation. Findings for other self-regulatory BCTs were less promising. No studies in the CVD domain directly tested engagement of self-regulation processes. A general challenge for this area stems from reliance on post-hoc tests of the effects of BCTs in multiple-component interventions. Recent advances in BCT taxonomies and the experimental medicine approach to engaging self-regulation mechanisms, however, provide opportunities to improve CVD prevention and management behavioural interventions.

Keywords: Meta-review; behaviour change techniques; cardiovascular disease; intervention; prevention; self-regulation.

Figure 1.

PRISMA flow chart: Flow of reports and meta-analyses into the meta-review. Note. MH = Mental Health.

Figure 2.

Quality Assessment Results according to AMSTAR 2 ratings Represents satisfaction of 16 criteria across 15 meta-analyses (two of the original criteria were split to distinguish between reviews that included only RCTs and those including RCTs and NRSIs). The figure displays the proportion of Yes’, Partial Yes’, and No’s satisfying criteria for all meta-analyses. Coding of subsets of items that supported a “Yes/Partial Yes/No” distinction for each review is found in Supplement 2. COIs = Conflicts of interest. k= Number of reviews. NRSIs = Nonrandomised studies of healthcare interventions. PICO = Population, intervention, control group, outcome. RCTs = Randomised controlled trials. RoB = Risk of bias.

Figure 3.

3-D heatmap depicting the BCTs (X-axis) assessed in each of the fifteen meta-analyses (Y-axis) and the number of studies contributing to each BCT analysis (Z-axis) Color (i.e., red, orange, yellow, light green, green) indicates AMSTAR quality; red through green represent from lower- to higher-quality meta-analyses. For reviews including multiple health outcomes for the same intervention component, the analysis with the greatest number of studies was selected for this figure. For the analyses conducted on self-monitoring and feedback by Morrissey et al. (2017) and Goodwin et al. (2016), the eligible number of studies is depicted on the Z-axis because the number of studies included in testing specific BCTs analysis was unavailable. Sr= Self-Regulation.