INTRODUCTION:
Low education and physical inactivity (PI) are modifiable risk factors for mild cognitive impairment (MCI). We aimed to investigate the association between education, PI, and MCI in Brazilians aged ≥50 years.

METHODS:
Cross-sectional study of the second wave of the Longitudinal Study of the Health of Brazilian Elderly (ELSI-Brasil, 2019-2021), with 9,949 participants (≥ 50 years). Those with complete data on education, physical activity (PA), and cognitive domains were included. Education was categorized as low (illiterate), medium (complete primary education), and high (secondary education or higher). PA was classified as physically inactive (<10 min/week moderate activity/walking and <10 min/week vigorous activity) and sufficiently active (≥150 min/week moderate activity or ≥75 min/week vigorous activity). MCI defined, through testing, with a global cognition z-score ≤ -1.5 standard deviations from the z-score of the normative sample. Associations were assessed using Pearson's chi-square test and Poisson regression with robust variance to estimate prevalence ratios (PR) and their 95% confidence intervals (CI).

RESULTS:
Prevalence of MCI = 5.25% (sample = 7,038). Low education level was associated with MCI (PR: 9.92; 95% CI: 6.79-14.51), as was PI (PR: 3.07; 95% CI: 2.43-3.87). In the multivariate analysis (reference: inactive with high education), inactive with low education showed a higher prevalence (PR: 11.27; 95% CI: 6.10-20.81), while inactive with medium education had a weaker association (PR: 4.26; 95% CI: 2.32-7.83). *p<0.001

CONCLUSIONS:
Inactive individuals with low education have a high risk for MCI. Although inactive individuals with medium education have a lower risk than inactive individuals with low education, the association remains significantly high and concerning in relation to inactive individuals with high education. These analyses reinforce preventive measures for MCI related to education and PA.