Objective: The principal aim of this pilot study was to evaluate the performance of a Gujarati version of the MMSE as a screening instrument for dementia. The effect of ethnicity on MMSE performance was also examined.
Design: Two-stage cross-cultural survey.
Setting: Elderly immigrant Gujarati and British-born white communities in Leicester.
Subjects: First stage: 149 Gujaratis and 148 whites. Second stage: 27 Gujaratis and 42 whites.
Measures: English and Gujarati versions of the MMSE, validated against clinical diagnosis following psychiatric interview (ICD-10 criteria).
Results: Mean MMSE scores were lower in the Gujarati group, but most of this difference was due to the effects of age, education and visual impairment. Ethnic group had an independent effect on three orientation items; when these were omitted, there was no difference in the mean MMSE scores of the groups after adjustment for age, education and visual impairment. The MMSE performed comparably in both groups as a screen for moderate-severe dementia, but was less efficient at detecting milder and less certain cases in the Gujarati group. The estimated prevalence of confirmed dementia was higher in the Gujarati group, but this was not statistically significant.
Conclusions: This Gujarati version of the MMSE performed adequately as a screen for dementia in this immigrant community population. Further evaluation of its performance is required in larger community samples, clinical samples and in native Indian Gujaratis.