Predictors of glycemic control in insulin-using adults with type 2 diabetes

OBJECTIVE - To determine the characteristics that influence glycemic control among insulin-using adults with type 2 diabetes. RESEARCH DESIGN AND METHODS - We studied all 1,333 eligible members of a large not-for-profit health maintenance organization who responded to a 1997 survey We tested associations among demographic, treatment, and psychometric variables with mean 1997 HbA(1c) values. The Problem Areas in Diabetes (PAID) instrument was used to assess the emotional effect of living with diabetes, and the Short Form 12 Physical Function Scale was used to assess the effect of physical limitations on daily activities. Based on differences between and within treatment groups, we built models to predict glycemic control for subgroups of subjects who were using insulin alone and those who were using insulin in combination with an oral hypoglycemic agent. RESULTS - Younger age, lower BMI, and increased emotional distress about diabetes (according to the PAID scale) were all significant predictors (P < 0.05) of worse glycemic control. However, except among individuals with an HbA(1c) level of >8.0 who were receiving combination therapy, only similar to 100% of the variance in glycemic control could be predicted by demographic, treatment, or psychometric characteristics. CONCLUSIONS - Personal characteristics explain little of the variation in glycemic control in insulin-using adults with type 2 diabetes. Possible explanations are that the reduced complexity of control in type 2 diabetes makes the disease less sensitive to personal factors than control in type 1 diabetes, that health-related behavior is less driven by personal and environmental characteristics among older individuals, or that, in populations exposed to aggressive glycemic control with oral hypoglycemic agents and nurse care managers, personal differences become largely irrelevant.