Methylation clocks fail to generalize across genetically admixed individuals

Methylation clocks, which are based on DNA methylation patterns and used to predict the risk of age-related diseases like Alzheimer's disease, have shown limited effectiveness when applied to genetically diverse populations. Researchers have evaluated various first-, second-, and third-generation methylation clocks in 621 Alzheimer's disease patients and controls from African American, Hispanic, and White cohorts.

The clocks proved to be less accurate in predicting age for individuals from mixed genetic backgrounds, particularly those with significant African ancestry. This reduced accuracy was also observed in 2500 individuals of European and African descent from three separate datasets.

Furthermore, the methylation clocks were unable to consistently detect age acceleration in Alzheimer's disease cases compared to controls, especially among genetically admixed individuals. To investigate the reasons behind the clocks' lack of generalization, the researchers analyzed the relationship between clock CpGs (cytosine-phosphate-guanine sites) and various genetic and epigenetic factors.

They discovered that common methylation differences between African and European ancestry individuals often occurred at clock CpGs. While genetic variants rarely disrupted clock CpGs between populations, a significant number of these CpGs had methylation quantitative trait loci (meQTL) with significantly higher frequencies in individuals with African genetic ancestry.

The findings suggest that methylation clocks often fail to predict age and Alzheimer's disease risk when used across different populations. The study highlights the importance of considering genetic and epigenetic patterns unique to each human population to improve the portability of methylation clocks.