More than one-tenth of adults worldwide now live with diabetes mellitus, and the number is projected to exceed 1.31 billion by 2050 if current trends hold. Yet one of the most stubborn obstacles to controlling this epidemic is not the discovery of new drugs but a deceptively simple question: are patients actually taking the medications they are prescribed? Nonadherence to medication for chronic diseases such as type 2 diabetes ranges from 30 to 50 percent in developed countries and may be even higher in developing ones. In Vietnam, studies have reported nonadherence rates of 23.9 percent among older outpatients in Hanoi and 31.9 percent in Hue. Skipping doses drives up glycated hemoglobin levels, raises the risk of microvascular and macrovascular complications, hospitalization, and mortality, and inflates health care costs.
A new cross-sectional study conducted at three provincial public hospitals in An Giang and Ca Mau, Vietnam, between January and March 2026, offers a systematic answer to why patients with diabetes fail to adhere. The researchers recruited 644 eligible patients aged 18 or older with a confirmed diabetes diagnosis who were using at least one antidiabetic medication, using a convenience-based consecutive sampling approach that enrolled the first five eligible patients per day at each hospital over 57 working days. Participants completed a structured self-administered questionnaire during outpatient visits, administered by trained research assistants who were not involved in their clinical care. The mean age of participants was 57.59 years, most were female (54.5 percent) and married (79.3 percent), 95.5 percent had health insurance, and 22.5 percent were on polypharmacy, defined as five or more prescribed medications.
The questionnaire’s centerpiece was a Vietnamese adaptation of the Identification of Medication Adherence Barriers Questionnaire, or IMAB-Q, a tool built on evidence about modifiable factors associated with treatment adherence and mapped to the Theoretical Domains Framework. Each item was rated on a 5-point Likert scale, with reverse coding applied so that higher scores always indicated greater barriers. Medication adherence itself was measured with the Brief Adherence Rating Scale, whose final item is a visual analog scale from 0 to 100 percent that served as the continuous outcome. Quality of life was assessed with the EQ-5D-5L questionnaire covering mobility, self-care, usual activities, pain or discomfort, and anxiety or depression. The English instruments were translated using a standardized forward-backward procedure, reviewed by a panel of three experts including a clinician, a clinical pharmacist, and a public health specialist, and pilot-tested among 30 patients with no further modifications required.
For psychometric validation, the full sample was randomly split into two equal subsamples of 322 each. Exploratory factor analysis in the first subsample, using principal axis factoring and Promax rotation, initially identified seven factors with eigenvalues greater than 1. Several two-item factors and weakly loading items were removed, yielding a stable three-factor solution that explained 56.3 percent of the total variance. The retained domains were labeled medication management barriers, psychosocial barriers, and motivational barriers, with Cronbach’s alpha coefficients of 0.91, 0.84, and 0.89 respectively. Confirmatory factor analysis in the second subsample supported the structure, with a comparative fit index of 0.913, RMSEA of 0.076, and McDonald’s omega coefficients all exceeding 0.70. Discriminant validity was well established, with all heterotrait-monotrait ratios below the 0.85 threshold.
The structural equation modeling results delivered the study’s most striking finding. When all three barrier domains were considered simultaneously, only motivational barriers (unstandardized coefficient of –0.297, standardized coefficient of –0.261, p < 0.001) and psychosocial barriers (–0.194 and –0.156, p < 0.001) were significantly associated with medication adherence. Medication management barriers, by contrast, showed no independent association (p = 0.637). Bootstrap analysis with 1,000 resamples confirmed the robustness of these paths: the confidence intervals for motivational and psychosocial barriers excluded zero, while the interval for medication management barriers included it. Correlation analysis echoed the pattern, with motivational barriers showing the strongest negative correlation with adherence (r = –0.270, p < 0.001), and only psychosocial barriers correlating significantly with quality of life (r = –0.180, p < 0.001).
Why would motivation outweigh practical medication know-how? The researchers argue that motivational barriers reflect patients’ cognitive processes, emotional appraisal, and decision-making. Adherence is shaped by patients’ perceived need for treatment and their concerns about potential adverse consequences: when patients do not see medication as necessary, or harbor substantial worries about it, they are less likely to take it as prescribed. Motivation may improve when patients understand the necessity of medication and perceive its contribution to positive health outcomes. Notably, traditional sociodemographic and clinical variables failed to explain motivational barriers, as the regression model for this domain was not statistically significant (p = 0.259), suggesting that motivational barriers are shaped by more complex intrapersonal and contextual factors such as treatment beliefs, family support, or health literacy, which the current model did not capture.
Psychosocial barriers, reflected mainly in emotional burden, social influence, and lifestyle-related conflict, tell a story familiar from qualitative research on type 2 diabetes. After a prolonged treatment process and repeated life difficulties, patients may become exhausted with their lives and stop pursuing treatment and self-care. Fear of negative community reactions to the disease is a potent barrier: many patients conceal their condition, avoid taking medication in the presence of others, and even refuse monitoring by treatment supervisors. Family responsibilities, lifestyle preferences, and financial difficulties can also tip the decision against taking medication, while adhering to a strict regimen can disrupt daily life and make it hard to maintain important social roles with family and friends.
The regression models did identify predictors for the other two domains. Polypharmacy was associated with higher psychosocial barrier scores (p = 0.008), consistent with evidence that complex medication regimens are linked to poorer glycemic control, lower adherence, and reduced health-related quality of life, and that patients taking multiple medications experience psychological stress from frequent hospital visits, financial burden, and a reduced sense of independence. For medication management barriers, participants with health insurance had significantly lower barrier scores (p = 0.001), while a higher number of chronic comorbidities was associated with higher barrier scores (p = 0.009). Multimorbidity, the authors note, creates a persistent imbalance between illness and treatment burden and the patient’s capacity to manage that burden, complicating care coordination, self-care, and medication organization.
The authors caution that the three-factor structure should be viewed as a context-specific organization of the original IMAB-Q barriers rather than a direct replication of a fixed structure, since the original UK validation produced four scales whose associations did not follow the hypothesized Theoretical Domains Framework domains. They also acknowledge the limits of the cross-sectional design, the susceptibility of self-reported measures to recall and social desirability bias, and the potential selection bias from excluding 24.7 percent of returned questionnaires that were more than half blank, mainly due to questionnaire fatigue among older outpatients. Still, the policy implications are clear: diabetes management programs should go beyond medication provision and counseling, prioritizing individualized psychosocial support, patient-centered communication, and strategies that strengthen treatment motivation, with patients on polypharmacy singled out for targeted interventions. Motivational and psychosocial barriers, the study concludes, are the most promising targets for future adherence-improving interventions in diabetes care.
Subject of Research: Medication adherence barriers among patients with diabetes mellitus: barrier domains, associated factors, and associations with medication adherence
Article Title: Medication adherence barriers among patients with diabetes mellitus: barrier domains, associated factors, and associations with medication adherence
Article References: Le, M. H., Hang, Q. D., Trieu, T. B., Nguyen, M. C., Lam, Q. V., Tran, T. D., & Huynh, M. C. (2026). Medication adherence barriers among patients with diabetes mellitus: barrier domains, associated factors, and associations with medication adherence. Pharmacia, 73, Article e205741. https://doi.org/10.3897/pharmacia.73.e205741
Image Credits: AI Generated
DOI: 10.3897/pharmacia.73.e205741
Keywords: diabetes, medication adherence, Vietnam, psychosocial barriers, motivation, public health, Medication, adherence, barriers, patients, mellitus, barrier
News Source: Ophelia Keating. (October 17, 2026). Why diabetes patients skip their pills: motivation matters more than medication know-how. Scienmag.
