Date of Award

Spring 2019

Document Type


Degree Name

Doctor of Philosophy (PhD)


Health Services Research

Committee Director

Qi Harry Zhang

Committee Member

Deanne Shuman

Committee Member

Margarita de Veciana


Problem Statement: Gestational diabetes (GDM) has been reported to affect as many as 18% of all pregnancies in the U.S. This diagnosis is costly and presents health risks to both baby and mother. The main risk to the mother with a history of GDM is her increased risk for diabetes which has been estimated at 35% to 60% in the following 10 to 20 years; more recent studies report a 7 to 8 times occurrence in the following 3 to 6 years. It is recommended that lifestyle efforts of eating healthfully and exercise can reduce this risk

Methods: Subjects (n=153) from a diabetes in pregnancy clinic with a recent history of GDM completed a mailed 115-question survey (10 completed by phone). Questions assessed diet and exercise behavior, health beliefs, self-efficacy, environmental support, diabetes-related variables, and socio-demographics. Five multivariate logistic regression models were used to test the utility of the Health Belief Model with added constructs in predicting diet and exercise behavior. The models consisted of varying combinations of health beliefs, self-efficacy and environmental factors.

Results: Healthy eating was analyzed in four models, but none were significant. High calorie food/beverage intake was analyzed in four models, all were significant (p < .01). Exercising ≥ 30 min three days or more weekly was analyzed in five models, all were significant (p < .01). Exercising to a sweat three days or more weekly was analyzed in five models and all were significant (p < .01); the two models assessing health beliefs, self-efficacy, and environmental support showed the most strength of prediction of all the models studied. Benefits exceed barriers and self-efficacy showed the highest prediction across all the models studied.

Conclusions: By utilizing an Expanded Health Belief Model with the added constructs, self-efficacy and diabetes-related (family history of diabetes, diagnosed with diabetes) and specific ecological/environmental (social/community support) cues to action, we were able to improve prediction of compliance with healthy lifestyle recommendations in women with a recent history of GDM.