Abstract
BACKGROUND: Cognitive impairment has a major impact on health, quality of life and survival and its increasing burden presents a critical global health challenge. Empirical population-based studies of cognitive function and its association with demographic, socioeconomic, health and behavioural factors among older adults in low-resource setting are rare. This study describes the burden of cognitive impairment and associations with demographic, health and behavioural factors among older adults in rural Bangladesh. METHODS: We conducted a population-based cross-sectional study of a random sample of men and women aged 60 years and above in 96 rural villages in Bangladesh. Cognitive function was measured using the Bangla Adaptation of the Mini-mental State Examination (BAMSE), where higher score indicates higher function. Blood pressure, height and weight were measured using standard protocols and fasting glucose and 2-h oral glucose test were used to identify diabetes risk. Interviewer administered survey questionnaires assessed depressive symptoms, anxiety and self-reported health behaviours. Analyses were carried out separately for men and women and examined associations between sociodemographic, health and behaviour factors with BAMSE scores using robust Poisson regression. FINDINGS: Data were gathered from 403 (216 female, 187 male) eligible participants. More than 50% of the population had at least mild cognitive impairment and women had lower cognitive scores than men. Younger age, higher education, wealth, and literacy were significantly associated with higher BAMSE scores among women and men. Associations with marital status varied between men and women, with being married having a positive association with BAMSE among women, i.e. higher cognitive function (relative score ratio (95% CI) 1.08 (1.02, 1.15), p = 0.013), but no association among men (0.94 (0.87, 1.02), p = 0.13). No clear associations were observed with diabetes or hypertension, but overweight and obesity were associated with an increased BAMSE score among women (1.10 (1.02, 1.19), p = 0.011) but not men (1.01 (0.94, 1.10), p = 0.70). Moderate and severe depressive symptoms were associated with lower BAMSE scores among men (0.90 (0.82, 0.99), p = 0.037), but not women (0.94 (0.83, 1.06), p = 0.31). Physical activity was associated with a relative increase in BAMSE score (1.08 (1.01, 1.16), p = 0.020) among women, though no association was seen in men (1.01 (0.95, 1.07), p = 0.76). The consumption of betel was associated with lower BAMSE among women (0.94 (0.89, 1.00), p = 0.056), but there was no evidence of association among men (1.01 (0.96, 1.07), p = 0.61). INTERPRETATION: A large proportion of older adults in rural Bangladesh have impaired cognitive function and there are important gender differences in the distribution of cognitive scores and their association with demographic, socioeconomic, health and behavioural factors in this population. Improvement in health and social care systems, taking into account the specific social, economic and gender dimensions of the context, are needed to prevent and manage the burden of cognitive decline in rural Bangladesh. FUNDING: This survey and analysis were part of the Bangladesh Diabetes Community-Led Awareness, Response and Evaluation (DClare) study funded by UKRI/MRC (MR/T023562/1) under the Global Alliance for Chronic Diseases Scale-Up Programme.