Dementia prevalence estimates in sub-Saharan Africa: comparison of two diagnostic criteria

BACKGROUND We have previously reported the prevalence of dementia in older adults living in the rural Hai district of Tanzania according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) criteria. The aim of this study was to compare prevalence rates using the DSM-IV criteria with those obtained using the 10/66 diagnostic criteria, which is specifically designed for use in low- and middle-income countries. METHODS In phase I, 1,198 people aged 70 and older were screened for dementia. A stratified sample of 296 was then clinically assessed for dementia according to the DSM-IV criteria. In addition, data were collected according to the protocol of the 10/66 Dementia Research Group, which allowed a separate diagnosis of dementia according to these criteria to be established. RESULTS The age-standardised prevalence of clinical DSM-IV dementia was 6.4% (95% confidence interval [CI] 4.9-7.9%) and of '10/66 dementia' was 21.6% (95% CI 17.5-25.7%). Education was a significant predictor of '10/66 dementia', but not of DSM-IV dementia. CONCLUSIONS There are large discrepancies in dementia prevalence rates depending on which diagnostic system is used. In rural sub-Saharan Africa, it is not clear whether the association between education and dementia using the 10/66 criteria is a genuine effect or the result of an educational bias within the diagnostic instrument. Despite its possible flaws, the DSM-IV criteria represent an international standard for dementia diagnosis. The 10/66 diagnostic criteria may be more appropriate when identification of early and mild cognitive impairment is required.