At least 72% of centenarian records in Greece were cases of pension fraud, where individuals were kept alive on paper solely so relatives could collect pension cheques. This systemic failure in data collection highlights a much larger problem: the science of human longevity is built on a foundation of inconsistent and often fraudulent paperwork.
The failure of mortality plateau modeling
Recent claims regarding a late-life mortality plateau—a pattern where the increasing risk of death slows down or stops at extreme ages—depend entirely on how researchers define middle age. In one study analyzing Italian data for individuals aged 105 and above, researchers tested over 850 possible model specifications. Only a single model produced a statistically significant plateau; any minor adjustment to the modeling parameters caused the significance to evaporate. This fragility suggests that "discoveries" in extreme-age mortality may be artifacts of specific modeling choices rather than biological realities.
Systemic errors in extreme-age datasets
Age measurement relies almost exclusively on administrative records, which are vulnerable to several undetectable error processes. Beyond direct pension fraud, researchers face identity substitutions where younger siblings hold valid paperwork belonging to older relatives. Because these individuals can often answer basic social questions, there is no reproducible method to verify if the age on the paperwork matches the biological reality of the person.
Other historical drivers of age inflation include low literacy rates, which led to clerical errors, and purposeful age increases to avoid military service or accelerate marriage eligibility. These errors are often undetectable, meaning demographers cannot accurately quantify how much of the "supercentenarian" population is real. When the data used to define the limits of human life is this compromised, the resulting conclusions about genetic hard-wiring and maximum lifespan remain speculative at best.
Addressing these data integrity issues is the only way to move beyond fragile statistical models and toward a genuine understanding of human biological limits.
Source: How long can humans live? We simply don't know
Domain: nature.com
Comments load interactively on the live page.