ObjectiveType II diabetes is a recognized risk factor of declining cognitive function in high-income countries. However, there is limited research on this association across low- and middle-income countries. We aimed to examine and compare the relationship between type II diabetes and cognition amongst adults aged 60 years and older for two of the largest LMICs: India and China.MethodsCross-sectional data was analyzed from population-based Harmonized Cognitive Assessment Protocols studies in India (n = 4,062) and China (n = 9,741). Multivariable-adjusted linear regression models examined the relationship between diabetes (self-reported or biomarker HbA1c ≥6.5%) and general cognition. Interaction testing assessed effect modification based on urban versus rural residence and educational attainment.ResultsType II diabetes was not associated with general cognitive scores in India or China in fully adjusted models. Interaction testing revealed a positive association in rural but not urban residences in India, however this was not seen in China. Both countries showed effect modification by education attainment. In India, diabetes was associated with higher average cognitive scores among those with none or early childhood education, while the relationship was null among those with at least an upper secondary education. In China, diabetes was inversely related to average cognitive scores among those with less than lower secondary education, while the relationship was null among the remainder of the study sample.ConclusionThe type II diabetes and cognitive function association in India and China differs from that observed in high-income countries. These findings suggest epidemiologic and nutrition transition variations. In India, health care access, urbanization and social differences between urban and rural areas may influence this relationship. In both countries, epidemiologic and nutrition patterns may adversely impact individuals from socially and financially vulnerable populations with less than lower secondary education. Longitudinal research using harmonized cognitive scores is encouraged to further investigate these findings.