Are ‘Biological Age’ Tests Dangerously Misleading?

People with an older biological age than their chronological age may be more at risk for dementia. (© svetazi – stock.adobe.com)

PHILADELPHIA — Many Americans have shelled out hundreds of dollars for biological age tests promising to reveal whether their bodies are aging faster or slower than their actual years. These methylation-based “aging clocks” have become the gold standard for evaluating anti-aging treatments, from supplements to lifestyle changes. But a controversial new paper argues these widely trusted tests might be fundamentally flawed — and potentially dangerous.

Independent researcher Dr. Josh Mitteldorf contends in a perspective published in the journal Aging that most methylation clocks fail to distinguish between two biologically opposite processes. Some age-related methylation changes reflect the body ramping up self-destructive programs, while others represent protective responses aimed at repairing damage. Current aging clocks treat both as equivalent signals of aging.

Mitteldorf, a theoretical biologist who runs the website AgingAdvice.org, suggests that treatments that make you appear “younger” on a biological age test might actually be shortening your life by shutting down crucial repair mechanisms your body has activated to fight damage.

Two Types of Aging Changes

Mitteldorf’s argument centers on a fundamental question dividing aging researchers: Is aging purely the result of accumulated damage, or does the body actually program itself to decline over time?

Most scientists today believe aging happens because cellular damage builds up over decades. Under this view, any age-related changes in gene expression must be the body’s attempt to combat damage by ramping up repair mechanisms.

Mitteldorf belongs to a minority faction believing aging is at least partially programmed. Just as puberty follows a genetic blueprint, so does aging. Some age-related changes represent the body turning on self-destructive processes like excessive inflammation or reduced repair capacity.

He calls these “Type 1” and “Type 2” changes, respectively. Type 1 changes are harmful: they’re part of a programmed self-destruction sequence. Type 2 changes are protective: they’re repair responses to accumulated damage.

Current methylation clocks indiscriminately combine both types. An intervention that reverses Type 1 changes (good for longevity) will look identical on current tests to one that reverses Type 2 changes (potentially bad for longevity).

The Smoking Gun Example

Mitteldorf points to a troubling example with the popular GrimAge clock, one of the most accurate predictors of death risk. A major component of this test is based on differences between smokers and non-smokers.

Why do smokers have different methylation patterns? “It is a reasonable conjecture that smokers’ bodies are constantly trying to repair their lungs,” Mitteldorf explains. Much of the smoking signature likely represents protective repair mechanisms working overtime.

But the GrimAge clock counts smoking as accelerated aging because smokers die earlier. If an intervention makes someone’s methylation profile look less like a smoker’s, is it because the body has successfully repaired lung damage and dialed down emergency repairs? Or has the intervention simply turned off protective mechanisms while leaving damage intact?

Such an intervention would score as “anti-aging” on the test while potentially shortening lifespan — a dangerous false positive.

Testing Random Genetic Changes

To investigate whether aging-related methylation changes are truly random or directed by biological programs, Mitteldorf attempted to build a “stochastic methylation clock” based on genuine genetic drift.

Using a database of 278 individuals aged 2 to 92, he identified methylation sites that remained partially active throughout life but showed increasing random variation with age. These represented genuine drift rather than directed changes.

Only about 10% of subjects showed genuine random methylation drift above background noise. When Mitteldorf tried to build an aging clock from this truly random drift, it performed poorly, correlating with age at just 0.38—far too weak to be useful.

This evidence, he argues, indicates that most consistent age-related methylation changes aren’t random at all; instead, they’re directed by biological processes. That, in turn, supports the idea that aging involves programmed changes rather than being driven solely by accumulated damage.

“An intervention that sets back the methylation age, as measured by Type 2, is deceiving us,” Mitteldorf writes. Such an intervention “nominally lowers ‘epigenetic age’, but it is likely to actually decrease life expectancy.”

Industry and Research Implications

The methylation clock industry has exploded in recent years, with companies like Elysium Health and TruDiagnostic offering direct-to-consumer tests. These companies, along with supplement makers and longevity clinics, routinely use methylation age as a biomarker to validate their products.

If Mitteldorf’s concerns prove valid, it could undermine confidence in a considerable portion of anti-aging research. Studies showing that certain interventions “reduce biological age” might need reinterpretation—or their results could be misleading if it’s unclear whether they’re affecting helpful or harmful processes.

The problem extends beyond consumer tests to clinical research. Pharmaceutical companies developing longevity drugs rely heavily on methylation clocks as endpoints in trials, since waiting decades for mortality data isn’t practical.

Mitteldorf acknowledges that separating Type 1 from Type 2 changes is extremely difficult with current methods. Most studies simply identify methylation sites that change consistently with age, without determining whether those changes are beneficial or harmful.

For researchers who believe aging is purely damage-driven, the implications are even more troubling. If all consistent methylation changes reflect repair, then any intervention that reduces “methylation age” could be harmful by suppressing protective responses.

While Mitteldorf’s study is limited by its small size and preliminary nature, it raises fundamental questions about tools that have become central to longevity research. As millions of people invest in anti-aging treatments validated by biological age tests, we need to ensure these tools are actually measuring what we think they’re measuring. Until we can separate the body’s self-destruction from its self-preservation efforts, methylation clocks might be leading us astray—making some interventions appear beneficial when they’re actually harmful.