news
Epigenetic clocks give wildly conflicting biological ages across different generations
Researchers analyzing DNA methylation data from the UK Household Longitudinal Study found substantial discrepancies between first, second, and third-generation epigenetic clocks. First-generation tools like Horvath consistently overestimated age in younger adults and underestimated it in older participants. The choice of algorithm completely changes the biological age number you get back.
Researchers analyzing DNA methylation data from the UK Household Longitudinal Study found substantial discrepancies between first, second, and third-generation epigenetic clocks. First-generation tools like Horvath consistently overestimated age in younger adults and underestimated it in older participants. The choice of algorithm completely changes the biological age number you get back.
The architecture take
I used to think buying a commercial biological age test gave you a hard number to build your life around. Then I watched researchers test Horvath, PhenoAge, and DunedinPoAm side-by-side on thousands of participants, and the certainty dissolved. First-generation clocks track chronological calendar years through DNA methylation, but they skew heavily depending on how old you are when you spit in the tube. Newer clocks try to gauge the actual pace of decay, catching systemic inflammation, metabolic drift, and cell turnover before symptoms surface. But if two algorithms give you an eight-year spread on the same vial of blood, you are not measuring biological truth. You are measuring how that specific developer weighted their training cohort. I stopped treating these scores like lab-grade facts last year. Watch direct-to-consumer longevity brands quietly pivot toward tracking single organ systems instead. For now, resting heart rate tells you more about your week than an algorithmic score ever will.