Reducing risk behaviours after stroke: An overview of reviews interrogating primary study data using the Theoretical Domains Framework

Background: Lifestyle changes, in addition to preventive medications, optimise stroke secondary prevention. Evidence from systematic reviews support behaviour-change interventions post-stroke to address lifestyle-related risk. However, understanding of the theory-driven mediators that affect behaviour-change post-stroke is lacking. Methods: Electronic databases MEDLINE, Embase, Epistemonikos and Cochrane Library of Systematic Reviews were searched to March 2023 for systematic reviews addressing behaviour-change after stroke. Primary studies from identified systematic reviews were interrogated for evidence supporting theoretically-grounded interventions. Data were synthesized in new meta-analyses examining behaviour-change domains of the Theoretical Domains Framework (TDF) and secondary prevention outcomes. Results: From 71 identified SRs, 246 primary studies were screened. Only 19 trials (N = 2530 participants) were identified that employed theoretically-grounded interventions and measured associated mediators for behaviour-change. Identified mediators mapped to 5 of 14 possible TDF domains. Trial follow-up ranged between 1–12 months and no studies addressed primary outcomes of recurrent stroke or cardiovascular mortality and/or morbidity. Lifestyle interventions targeting mediators mapped to the TDF Knowledge domain may improve the likelihood of medication adherence (OR 6.08 [2.79, 13.26], I2 = 0%); physical activity participation (OR 2.97 [1.73, 5.12], I2 = 0%) and smoking cessation (OR 10.37 [3.22, 33.39], I2 = 20%) post-stroke, supported by low certainty evidence; Lifestyle interventions targeting mediators mapping to both TDF domains of Knowledge and Beliefs about Consequences may improve medication adherence post-stroke (SMD 0.36 [0.07, 0.64], I2 = 13%, very low certainty evidence); Lifestyle interventions targeting mediators mapped to Beliefs about Capabilities and Emotions domains may modulate low mood post-stroke (SMD -0.70 [-1.28, -0.12], I2 = 81%, low certainty evidence). Conclusion: Limited theory-based research and use of behaviour-change mediators exists within stroke secondary prevention trials. Knowledge, Beliefs about Consequences, and Emotions are the domains which positively influence risk-reducing behaviours post-stroke. Behaviour-change interventions should include these evidence-based constructs known to be effective. Future trials should address cardiovascular outcomes and ensure adequate follow-up time.