The purpose of this research was to investigate the impact of social support on adherence to a therapeutic regimen among obese Type II diabetic clinic patients. The aims of the study were to compare the effect of small reference group management with an educational advice program on the therapeutic outcomes of adherence and to test a theoretical model proposed to explain adherence to a diabetic regimen. The concepts of the model were derived from symbolic interactionist and role theory. Forty-one patients attending an outpatient diabetic clinic were randomly assigned to the two treatment programs and 32 patients completed the 6 month study. Subjects met for a total of 10 one-hour sessions which were held weekly the first month, biweekly the second month, and monthly the final 4 months. Demographic information, knowledge of diabetes, and measures of the model variables (social support, health locus of control, health perceptions, diabetic belief, regimen adherence), were collected at the start (t₁) and the termination of the study (t₁₀). Analysis of audiotapes recorded at t₂ and t₉ indicated that patient-initiated interactions were significantly longer in duration for the small reference group. The experimental group also demonstrated significantly lower blood glucose values at t₁₀, however, there was not a corresponding difference between groups in glycohemoglobin, urine glucose, or percent overweight. The total sample experienced a significant mean reduction in glycohemoglobin (-11.1%; p < .05) and an increase in social support and knowledge. Serial blood glucose and glycohemoglobin values indicated improvement in diabetic control was more marked when subjects met weekly or biweekly and stabilized or worsened when meetings were reduced. The staged recursive theoretical model was analyzed using multiple regression statistics. The empirical test of the model of t₁ indicated minimal support in terms of significant explained variance in 6 of the 8 dependent variables. In addition, some of the path coefficients indicated a possible interaction effect between resistance and locus of control. The t₁₀ model demonstrated cohesion among the first three stages of independent variables and greater explained variance, however, multicollinearity proved to be a significant problem when interpreting the path coefficients.

published_or_final_version / abstract / toc / Psychiatry / Master / Master of Philosophy