Applying the Theory of Planned Behaviour to persons with prediabetes and diabetes: An examination of intention and behaviour for healthy eating and physical activity

Healthy eating (HE) and physical activity (PA) are essential for effective management of type 2 diabetes and can prevent/delay the onset of diabetes in persons at elevated risk for diabetes (i.e., prediabetes). However, adherence to dietary and exercise recommendations tends to be quite low in persons with and at-risk for developing diabetes. There exists a critical need to better understand predictors of HE and PA to support these behaviours in this chronic disease population. The present dissertation employed a theoretical framework, the Theory of Planned Behaviour (TPB), to test the utility of this social-cognitive model for prospectively predicting determinants of HE and PA in two diabetes sub-populations. A belief elicitation study (N = 30) was conducted to elicit specific behavioural beliefs relevant to the sample populations, and findings were used as part of the main study questionnaire. For the main study, prediabetes (N = 117) and diabetes (N = 122) patients recruited from diabetes education classes completed a questionnaire (time 1) and follow-up behavioural survey (time 2) after 1 month. Results generally support the the TPB's utility for diabetes samples but suggest the theory is insufficient for prediabetes samples. Path analysis and multiple groups invariance testing suggest that predictors of HE and PA differ by behaviour and for persons at different stages of the diabetes disease course. For prediabetes patients, intention was not a significant predictor of behaviour. Gender, age, and perceived behavioural control predicted HE. There were no significant predictors of PA. For diabetes patients, intention and subjective norms predicted HE. Intention was the only predictor of PA. The TPB accounted for more variance in HE behaviour for both groups. Factors external to the theory would likely add to the prediction of both behaviours, especially PA. Differences in specific beliefs correlated to behaviour were also observed between groups; however the number of significant beliefs identified for intervention was low. The findings support and extend preliminary TPB research with other diabetes samples and suggest that the focus of interventions to promote HE and PA should differ for prediabetes and diabetes patients. Limitations, future research directions, and clinical implications are discussed.