Coexisting Anemia and Elevated C-Reactive Protein in Relation to Memory-Related Cognitive Outcomes: Evidence From CHARLS With Exploratory Support From ELSA.

AIM: To examine whether coexisting anemia and elevated C-reactive protein (CRP), reflecting a potential dual vascular-inflammatory burden, are associated with memory-related cognitive outcomes in middle-aged and older adults, using CHARLS as the primary cohort and ELSA for exploratory cross-cohort support. METHODS: Longitudinal data from 5166 CHARLS participants were analyzed. Anemia was defined using sex-specific hemoglobin thresholds, and elevated CRP was defined as CRP > 3 mg/L. Participants were categorized into four anemia-CRP exposure groups. Multivariable linear regression models examined 2015 cognitive scores and cognitive change from 2011 to 2015. The primary outcomes were 2015 immediate recall, delayed recall, and memory scores. Other cognitive measures and cognitive change were secondary outcomes, whereas low cognitive performance and the CHARLS-ELSA pooled analysis were exploratory. Robustness was assessed using multiple sensitivity analyses, including bootstrap resampling, mixed-effects models, inverse probability weighting, and false discovery rate correction. RESULTS: In CHARLS, participants with both anemia and elevated CRP had poorer 2015 memory-related cognitive performance than those with neither condition. After false discovery rate correction, associations remained significant across the memory-related outcomes, including immediate recall, delayed recall, and memory. The strongest associations were observed for delayed recall (β = -0.541, 95% CI: -0.888 to -0.194; FDR-adjusted p = 0.024) and memory (β = -0.983, 95% CI: -1.579 to -0.387; FDR-adjusted p = 0.024). Associations with global cognition and cognitive change were directionally consistent but did not remain statistically significant after FDR correction. CONCLUSIONS: Coexisting anemia and elevated CRP were associated with poorer memory-related cognitive performance in middle-aged and older adults. Longitudinal cognitive changes require cautious interpretation.