Aim

To investigate the prevalence of cognitive impairment and its predictors among the elderly in Bangladesh.

Subject and Methods

We use a cross-sectional survey of 1015 older people (≥60 years) in Bangladesh collected jointly by the Sir William Beveridge Foundation (SWBF) and Aging Support Forum (ASF), Bangladesh. The Mini-Mental State Examination (MMSE) scale, adapted to the Bengali language, was used for assessing cognitive impairment, which was available for 111 participants. Logistic regression analysis was used to identify predictors of cognitive impairment. Multiple imputation under the missing-at-random (MAR) mechanism was used to deal with missing data.

Results

Overall, 31 out of 111 (27.9%) participants had mild, moderate or severe cognitive impairment, with a higher proportion among women (33.3%, 17 out of 51) than men (23.3%, 14 out of 60). More precisely, 24 out of 111 (21.6%) and 7 out of 111 (6.3%) had mild/moderate and severe cognitive impairment, respectively, with a higher percentage among women (mild/moderate: 25.5%, severe: 7.8%) than men (mild/moderate: 18.3%, severe: 5.0%). Age (odds ratio, OR = 1.06, p = 0.046, 95% CI: 1.001 to 1.119) and social engagement (OR = 0.25, p = 0.033, 95% CI: 0.072 to 0.898) were found to be statistically significant predictors of cognitive impairment in this group of people. The effect of physical disability fell short of statistical significance at the 5% level but was significant at the 10% level (OR = 2.89, p = 0.099, 90% CI: 0.999 to 8.370).

Conclusion

The prevalence of cognitive impairment in Bangladesh is alarming and is comparatively higher among elderly women, which seems to increase with age and the presence of any physical disability. Also, prevalence appears to be lower in people who are more socially engaged or active.