The relationship between stages of change and glycemic control in patients with diabetes

This research involved development of measures for the Transtheoretical Model (TMC) variables for each of the primary self-management behaviors necessary for control of diabetes (oral medication taking, insulin use, diet, self-monitoring of blood glucose, and physical activity/exercise). Validity was assessed through examination of (a) relationships among TMC variables (stage of readiness, decisional balance, and self-efficacy), and (b) the ability of the stage measure to predict outcome as has been done in other behavior change initiatives. The diabetes outcome variable was glycemic control (HbAlc).; This study included two phases. Phase one involved the development of a questionnaire including TMC variables for each of the diabetes management behaviors. Thirty-five patients with diabetes were interviewed or included in focus groups to gain information regarding salient pros and cons (decisional balance items) and tempting situations (self-efficacy confidence items) for each of the diabetes management behaviors. The questionnaire was subjected to expert panels (TMC experts, diabetes experts, literacy level experts) and was pre-tested with diabetes patients.; Phase two involved a cross-sectional implementation of the questionnaire with a sample of 193 patients with Type 1 or Type 2 diabetes from four participating sites: two American Diabetes Association (ADA) Recognized Provider medical practices and two hospital-based indigent care clinics. Patients filled out the questionnaire at the ADA Recognized Provider sites and were interviewed at the indigent care sites due to literacy concerns. HbAlc was extracted from patients' charts.; The expected relationships between decisional balance and self efficacy and the stage measure were confirmed for most of the behaviors. Principal components analysis produced self-efficacy and decisional balance pro and con scales of more than four items with factor loadings over .60 (Cronbach's alphas ranged from .70 to .94). Multiple regression revealed significant variance contribution to HbAlc for stage of readiness for portions of all the diabetes management behaviors except blood glucose monitoring, with oral medication taking and insulin use contributing the most. Patient age, years since diabetes diagnosis, and diabetes-related emergency room episodes also contributed significant explanatory variance to HbAlc. A discussion of the implications of the findings is included, along with limitations and recommendations for future research.