Physical activity and lifestyle behaviours in obesity and the prevention of type 2 diabetes : the PREVIEW Project

Interventions are needed to stem the rise of obesity and type 2 diabetes. Physical activity (PA) is integral to lifestyle interventions, however, a paucity of research applying objective measures of PA exists in populations at risk of diabetes. Insight into changing PA and dietary behaviours is required to develop effective interventions. The aim of this thesis was to capture an insight into high-risk behaviours and behaviour change in pre-diabetic adults, and assess the feasibility of running an intervention in overweight and obese children. Study 1 investigated the associations between objectively measured PA and cardio-metabolic risk factors in pre-diabetic adults from 8 countries. Results indicated that moderate-to-vigorous physical activity (MVPA) was negatively associated with cardio-metabolic risk factors. However, associations between total PA and all risk factors were at least as strong as than MVPA. Study 2 examined the psychosocial correlates of objectively measured PA. Results showed that habit-strength and exercise intentions were negatively and positively associated with MVPA respectively. Two-way interactions between peer support and inactivity temptations and between age and social support, suggested that social support was of greater importance in older participants and in the presence of inactivity temptations. Associations between PA self-efficacyand goal adjustment were country specific. Study 3 investigated associations between body fat% and lifestyle behaviours in15,977 children aged 9-11yrs. Multilevel-models revealed body fat% wasnegatively associated with active transport to school, full fat milk and sweetenedbeverage consumption. Later bed time was positively associated with bodyfat%. No change was found in associations over a 10-year period. Study 4 used mixed methods case studies combining participant perceptions ofthe behaviour change process with objective outcome measures to assess thefeasibility of running an intervention in children. Three of the four cases showedimprovements in dietary and PA behaviour and reductions in BMI z-score,HOMA-IR and HbA1c. Semi-structured interviews indicated that behaviouralchanges occurred despite not always being detected by objective measures,possibly due to compensation effects or seasonal changes. Furthermore, goalsetting was considered useful but planning goals waned throughout the study. Compliance with self-monitoring techniques was low and largely reliant onparents.