Economic, cognitive, and behavioural dimensions of medication non-adherence and their association with blood-pressure control among rural Bangladeshi adults with hypertension: a cross-sectional analysis
Year: 2026
Type: article
Study design: Observational Study
Abstract: Medication non-adherence is multidimensional, yet adherence scales are usually analysed as a single total score, obscuring which type of barrier — economic, cognitive, or behavioural — is most burdensome and which relates to disease control. Using the three dimensions of the Bangladesh Medication Adherence Scale (BMAS), we examined, among rural Bangladeshi adults with hypertension, the relative burden of each barrier dimension, its association with blood-pressure (BP) control, and its sociodemographic distribution. Secondary cross-sectional analysis of a community-based survey in Kamalganj, north-east Bangladesh (461 adults with treated hypertension; analytic N = 446 currently taking medication). The 9-item BMAS yields three 3-item subscales — economic, cognitive (memory-related/unintentional) and behavioural (deliberate/intentional); higher scores denote fewer barriers. Barrier burden was compared with the Friedman test. Associations with BP control (< 140/90 mmHg) and with systolic BP were estimated by logistic and linear regression adjusted for age, gender, socioeconomic status (SES), education, body-mass index and diabetes, with Bonferroni correction across the three dimensions. Sensitivity analyses examined item assignment, inclusion of participants not taking medication, a < 130/80 mmHg threshold, and regimen and lifestyle factors. The factor structure was examined by confirmatory factor analysis (CFA). BP was controlled in 185 of 446 participants (41.5%). Economic barriers were the most burdensome (mean 5.31 ± 2.72 of 9), followed by cognitive (6.75 ± 1.85) and behavioural (7.45 ± 1.81) barriers (Friedman χ 2 (2) = 234.45, p < .001; Kendall's W = .26, modest concordance); 46.2% of participants had the maximum behavioural score. The overall logistic model for BP control was not significant (χ 2 (11) = 12.65, p = .32; Nagelkerke R 2 = .038). Within it, fewer cognitive barriers were associated with higher odds of control (adjusted odds ratio 1.14 per point, 95% CI 1.01–1.30, p = .037; Bonferroni-corrected p = .11) and with lower systolic BP (B = − 1.41 mmHg per point, 95% CI − 2.54 to − 0.27; corrected p = .045), a difference of under 3 mmHg for two subscale points. Estimates were similar across sensitivity analyses, but confidence intervals often included the null. Economic and behavioural barriers were not associated with BP. Higher SES was associated with fewer economic barriers (β = .36). CFA did not support the original three-factor structure (CFI = .87; RMSEA = .15); fit was acceptable with item 2 reassigned to the behavioural factor. In this rural Bangladeshi sample, the most burdensome adherence barrier — economic, concentrated among poorer patients — was not associated with BP control, and the overall model for control was not significant. A small, exploratory association between fewer memory-related barriers and lower BP needs confirmation in longitudinal studies using objective adherence measures and an independently confirmed factor structure. The concentration of economic barriers among poorer patients is an equity concern in its own right; it is not a finding of the BP models.
Source: BMC Public Health
Institutions: Bangladesh Medical University, University of Dhaka
Language: English
Subfield: Family Practice
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $3,290
OpenAlex ID: W7213191653
