BACKGROUND: Dementia is increasingly prevalent due to rapid aging of the population, but under-recognized among people with low education levels. This is partly due to a lack of appropriate and precise normative data, which underestimates cognitive aging in the use of screening tools for dementia.

OBJECTIVE: We aimed to improve the precision of screening for cognitive impairment, by characterizing the patterns of cognitive aging and derived normative data of the Mini-Mental State Examination (MMSE) for illiterate and low-educated populations.

METHODS: This community-based study included data from 2,280 individuals aged 40 years or older from two rural areas. Multiple linear modeling examined the effect of aging on cognition reflected by the MMSE, stratified by education level and gender. Threshold effect of age on cognition was performed using a smoothing function.

RESULTS: The majority of participants (60.4%) were illiterate or had attended only primary school (24.6%). The effect of aging on cognition varied by gender and education. Primary-school educated females and males remained cognitively stable up to 62 and 71 years of age, respectively, with MMSE score declining 0.4 and 0.8 points/year in females and males thereafter. Illiterates females scored 2.3 points lower than illiterate males, and scores for both declined 0.2 points/year. According to these results, normative data stratified by age, education and gender was generated.

CONCLUSION: This study suggests gender and educational differences exist in cognitive aging among adults with limited or no formal education. To improve screening precision for cognitive impairment with the use of MMSE in low-educated population, age, gender, and education level should be considered.