We thank Dr Varghese (Reference Varghese2005) for his letter regarding our article (Reference Shaji, Bose and VergheseShaji et al, 2005). Lack of education is a potent predictor of poor performance across many items of the Chinese version of the Mini-Mental State Examination (MMSE; Reference Katzman, Zhang and Ouang-Ya-OuKatzman et al, 1988). There was no significant difference between total MMSE scores of those who were illiterate and those who were literate in the pilot study conducted with the Malayalam adaptation of the MMSE. Hence it was decided to use the same score for both groups.

We identified 55 cases of dementia among 327 people who scored at or below the cut-off on the MMSE. The one case identified from the 10% of the negatively screened population was counted as one among the ten cases in the negatively screened population of 1607 (i.e. 65 cases in 1934 people aged 65 years and above).

The assessment of risk factors based on retrospective accounts of the carers and an inadequate number of controls for calculating the odds ratios can be considered methodological limitations of the study. The prevalence of dementia increases proportionately with age (χ²=40.29, d.f.=5, P<0.001). This χ² value was not given in the text. The number of patients with Alzheimer's disease was 30. The error in the article is regretted.