Integrating Health Belief Model and Theory of Planned Behavior to Explain Complementary Therapy Use in Type 2 Diabetes: A Cross-Sectional Study in Denpasar, Indonesia

Ni Made Umi Kartika Dewi (Departement of Yoga and Health, Brahma Widya Faculty, I Gusti Bagus Sugriwa Denpasar Hindu State University, Denpasar, 80237, Indonesia. Doctoral Study Program, Faculty of Medicine, Universitas Udayana, Denpasar, 80232, Indonesia)
Ni Luh Seri Ani (Department of Public Health and Preventive Medicine, Udayana University, Denpasar, 80232, Indonesia)
Dinar Saurmauli Lubis (Department of Public Health and Preventive Medicine, Udayana University, Denpasar, 80232, Indonesia)
Agung Wiwiek Indrayani (Department Pharmacology and Therapy, College of Medicine, Universitas Udayana, Denpasar, 80232, Indonesia)
Pande Putu Januraga (PUI PT Center for Public Health Innovation, Udayana University)

Media Publikasi Promosi Kesehatan Indonesia (MPPKI)
Vol. 9 No. 4 (2026): April 2026

Introduction: Complementary therapy (CT) is widely practiced among individuals with type 2 diabetes mellitus (T2DM) in Indonesia. However, the psychosocial determinants influencing CT adoption and its safe integration into biomedical care remain insufficiently understood. This study aimed to determine the prevalence and psychosocial determinants of CT use among patients with T2DM using an integrated Health Belief Model (HBM) and Theory of Planned Behavior (TPB) framework. We hypothesized that stronger psychosocial constructs specifically self-efficacy, subjective norms, and positive attitudes would be positively associated with CT use, whereas higher levels of CT-specific rational knowledge would be inversely associated with CT adoption. Methods: A cross-sectional survey of 300 adult outpatients with T2DM was conducted at four community health centers in Denpasar, Indonesia, selected through multistage cluster sampling from May to July 2025. Interviewer-administered questionnaires collected sociodemographic and clinical characteristics, CT-specific rational knowledge, and HBM/TPB constructs. Data were analyzed using chi-square and Mann–Whitney U tests for bivariate comparisons, and backward stepwise binary logistic regression to identify independent predictors of CT use. Results: CT use within the past six months was reported by 43.3% of participants. Usage was significantly associated with treatment type (?²=9.458, p=0.024) and check-up consistency (?²=5.999, p=0.014). CT users demonstrated higher scores across all HBM and TPB constructs (all p<0.001) but lower rational knowledge (p<0.001). In multivariate analysis, self-efficacy (OR=13.46, p=0.002), subjective norms (OR=9.20, p=0.005), and attitudes (OR=3.86, p=0.021) independently predicted CT use, while rational knowledge was inversely related (OR=0.395, p<0.001). Conclusion:CT use among patients with T2DM was prevalent and predominantly influenced by psychosocial rather than demographic or knowledge-related factors. Strengthening culturally tailored communication, CT-specific health literacy, and provider readiness is crucial to promote safe, integrative diabetes management.