Background In the context of rapidly ageing populations worldwide, this study aims to investigate a cross- country comparison of the relationship between various socioeconomic markers such as education, household wealth, and urbanicity with cognitive decline over 8 years in England and China.

Methods We used data from 7,200 adults aged 50+ from the English Longitudinal Study of Ageing (ELSA) and Chinese Health and Retirement Longitudinal Study (CHARLS), national representative samples of English and Chinese populations. ELSA and CHARLS are similar in their study design and have comparable measures at each wave. For these analyses, we used the available data spanning over 8 years across ELSA wave 5 (20010/11) to wave 9 (2018/19); and CHARLS waves 1 (2011) to 4 (2018). The outcome was the change in memory assessed over time using immediate and delayed 10-word recall tests (max score 20) over 4 waves at every two-year follow-up within each of the two cohorts. We measured socioeconomic status at baseline, including individual-level (education and wealth) and area-based characteristics (urban/rural) with similar comparable measures within each cohort. Educational attainment was similarly classified into three categories: below A-level, A-levels or equivalent, and university degree. Total housed wealth was classified into quintiles. The associations between each SES marker and cognitive decline over an 8-year follow-up were examined by linear mixed models assessed comparatively within each country.

Results In English adults, we found a significant protective association between higher levels of education and baseline memory (intercept beta (β) =7.8, standard errors (SE)=0.15) with a β =1.9, SE=0.9 higher memory scores for those with vocational training and β =2.9, SE=0.2 for those with a degree. A slower decline in memory over time was observed for those with intermediary education β =0.6, SE=0.2 independent of all covariates. These associations were similar across countries but with more robust protections for those with intermediate levels of education in the Chinese population compared to England. Lastly, there was a significant positive association between living in an urban area and higher baseline memory with β =0.9, SE=0.6 and slower memory decline over time β =0.6, SE=0.1, particularly in China, but not in England. No associations were found with the level of wealth.

Conclusion These results imply that a socioeconomic advantage in terms of higher education at the individual level was associated with a slower memory decline over almost a decade both in England and China, with a most pronounced difference in participants living in rural China. Public health strategies for preventing cognitive decline should target socioeconomic gaps to reduce health disparities and protect those particularly disadvantaged.