Translating research results into everyday healthcare practice : an empirical study examining the influence of social-cognitive and perceived environmental factors on physical activity and weight during early stages of a lifestyle intervention

Abstract

Interventions aiming at increasing physical activity (PA) carry great potential to counteract lifestyle-related diseases such as type 2 diabetes mellitus. However, many interventions are only conducted in controlled settings without or only insufficient implementation into real-world settings (‘research-to-practice gap’). In order to improve public health, such interventions need to be translated into practice and therefore examined in terms of their practical effectiveness. Knowledge about the underlying mechanisms of an intervention, i.e. how an intervention affects behavioural determinants and how these affect the desired behavioural outcomes, is essential to design or replicate effective real-world interventions. The aim of this study was to explore the translation of research into practice by evaluating the effectiveness of a complex theory- and evidence-based lifestyle intervention (PREMIT) implemented in the real world. Data of 1484 participants of a behaviour change intervention were analysed. The behaviour change intervention PREMIT (PREview behaviour Modification Intervention Toolbox) was designed to support participants to maintain their initial weight loss in their natural living environment. Whereas the intervention was designed to improve participants’ healthy eating and physical activity behaviours, this study focused on PA behaviour and its determinants. Physical activity behaviour was assessed by accelerometers. During clinical investigation days (CID), participants’ body weight was registered and the participants filled in a battery of questionnaires with regard to social-cognitive and perceived environmental determinants (self-efficacy, expected benefits and expected disadvantages of physical activity, social support, temptations to be active). Analysis of Variance was used to investigate behavioural determinants’ change between two clinical investigation days for participants rarely and participants regularly attending PREMIT sessions in order to gain insights into the effect of the intervention on behavioural determinants. Structural equation analysis was used to examine the influence of social-cognitive and perceived environmental determinants on physical activity, and the influence of PA behaviour on weight change. Furthermore, the interrelation between determinants was investigated. Results showed that between CID 2 and CID 3 weight decreased for high attenders and increased for low attenders, suggesting PREMIT to be effective. Examination of the intervention mechanisms showed that the intervention did not influence all behavioural determinants and that not all behavioural determinants influenced PA behaviour. High attendance was associated with a decrease in temptations to be inactive and expected disadvantages of physical activity whereas low attenders showed an increase in temptations and expected disadvantages. There were no differences between low and high attenders with regard to benefits of physical activity, self-efficacy, and social support. Results of the structural equation model showed that expected disadvantages as well as social support significantly influenced PA behaviour (fewer expected disadvantages and more social support led to increased PA behaviour). Expected benefits, self-efficacy, and temptations were not significantly associated with PA behaviour. Increased social support and increased self- efficacy were associated with fewer temptations. Outcome expectancies (benefits and disadvantages) influenced self-efficacy. Physical activity influenced weight change. The results demonstrate that the PREMIT intervention was effective in a real-world setting and may therefore contribute to preventing and combatting the rising burden of T2DM and its related diseases. The analysis of the underlying mechanisms showed expected outcomes of PA behaviour and the social environment to be relevant components to consider in developing effective behaviour change interventions. As the hypothesised mechanisms could only partly be confirmed, further research into the exact mechanisms of interventions is needed. Furthermore, the extent to which the underlying mechanisms of interventions may be valid and effective on a community level needs to be examined in order to design large-scale interventions for people with prediabetes to have an impact on population health.