Abstract
OBJECTIVE: Gestational diabetes mellitus (GDM) poses significant health risks during pregnancy, with dietary adherence being crucial for effective management. This study aims to identify factors influencing dietary compliance to enhance patient outcome. METHODS: This retrospective cohort study analyzed 189 GDM patients from Wuhan Children's Hospital between January 2021 and June 2023. The patients were categorized into good and poor dietary adherence groups using the Perceived Dietary Adherence Questionnaire. Variables such as demographic data, disease duration, educational attainment, income, employment status, obstetric history, and dietary sources, were collected. Knowledge levels were evaluated using the Gestational Diabetes Mellitus Knowledge Questionnaire (GDMKQ), and social support was assessed by the Medical Outcomes Study Social Support Survey. RESULTS: A multifactorial logistic regression model was developed to predict poor dietary compliance, and the risk factors included lower educational attainment (Coefficient: 1.249; Odds Ratio (OR): 3.487), lower income (Coefficient: 2.282; OR: 3.602), and takeout breakfasts (Coefficient: 0.838; OR: 2.311). Improved GDM knowledge (Coefficient: -0.344; OR: 0.709) and social support levels (Coefficient: -0.072; OR: 0.931), unemployment (Coefficient: -0.935; OR: 0.392), and obstetric history (Coefficient: -0.980; OR: 0.375) were protective factors against poor compliance. The multifactorial logistic regression model was formulated as follows: Logit (P) = β0 + β1 (Educational Level) + β3 (Employment Status) + β4 (Obstetric History) + β5 (Breakfast Source) + β6 (GDMKQ Scores) + β7 (Social Support). The model demonstrated robust predictive power with an area under the curve (AUC) of 0.854 in internal validation and 0.972 in external validation. Calibration plots indicated good agreement between predicted and observed outcomes, supporting the model's reliability and clinical utility. CONCLUSION: The study identified key demographic, behavioral, and social determinants affecting dietary compliance in GDM patients. Critical factors include education levels, household income, employment, breakfast source, GDM knowledge, and social support. These insights can inform interventions to enhance dietary adherence and optimize GDM management strategies in clinical settings. Our multifactorial logistic regression model displays high predictive accuracy and serves as a practical tool for assessing dietary compliance risks, facilitating personalized patient care.