There’s a difference between feeling stressed and being under chronic stress, even though the two get used interchangeably. Feeling stressed is a subjective state, something you can usually name: a hard week, a difficult conversation, a looming deadline. Chronic stress is something else. It’s a prolonged, repeated activation of the body’s stress-response systems, and research suggests it leaves physical traces that don’t always line up with how a person says they feel.
This isn’t a claim that a blood test can tell you whether you’re mentally well, or diagnose anxiety, depression, or burnout. It can’t, and no biomarker panel currently does that. What a specific, well-studied set of biomarkers can do is offer a rough physiological readout of whether a prolonged stressful period has started to register in the body, as one more data point alongside how someone actually feels, not a replacement for it.
The research behind “wear and tear”
The relevant concept here is allostatic load, introduced by neuroscientist Bruce McEwen and developed with sociologist Teresa Seeman in the 1990s. It describes the cumulative physiological cost of repeatedly adapting to stressors, the biological “wear and tear” that builds up when the body’s stress systems are activated too often, for too long, without adequate recovery.
In one of the foundational studies, published in the Proceedings of the National Academy of Sciences, Seeman and colleagues tracked nearly 1,200 older adults and found that higher allostatic load scores at baseline predicted a meaningfully increased risk of death, and of cognitive and physical decline, seven years later, independent of socioeconomic status and existing health conditions. That finding has since been replicated and extended across dozens of cohort studies internationally.
What the biomarkers actually are
Allostatic load isn’t measured by a single number. Researchers typically combine markers across several physiological systems, most commonly the cardiovascular, metabolic, immune, and neuroendocrine systems. A 2023 meta-analysis in Psychoneuroendocrinology, pooling data from thirteen cohort studies, found that a compact five-marker index, C-reactive protein (an inflammation marker), resting heart rate, HDL cholesterol, waist-to-hip ratio, and HbA1c (a measure of average blood sugar), predicted long-term mortality about as well as far more elaborate biomarker panels. Other commonly studied markers include cortisol and DHEA-S, both produced by the adrenal glands, where research has focused particularly on the ratio between them rather than either hormone in isolation, since flattened or dysregulated patterns (rather than simply “high” or “low”) tend to show up more consistently in chronically stressed populations.
Why this doesn’t mean what it sounds like it means
It’s worth being explicit about the limits here, because it’s easy to oversell this kind of research. Allostatic load is a population-level epidemiological construct. It was built and validated to predict group-level outcomes across cohorts of thousands of people, not to hand any one individual a diagnosis from a single blood draw. A single elevated marker, on its own, says very little. Context, trends over time, and clinical interpretation all matter more than any one number.
It’s also worth noting that cortisol and inflammatory markers rise and fall for reasons that have nothing to do with chronic stress: illness, poor sleep the night before, recent exercise, medication, and the time of day a sample is taken can all move these numbers. None of this is a stand-in for a clinical assessment of mental health, and it isn’t a tool for self-diagnosing anxiety or depression. Those still need a qualified clinician.
Takeaway
Where this becomes genuinely useful is for the person whose subjective stress doesn’t seem to resolve with rest, who wants to know whether a prolonged period of pressure, a demanding job, a caregiving role, a difficult year, has started to show up anywhere measurable. Getting a handful of these markers checked through an advanced blood test, reviewed by a doctor rather than self-interpreted, is one practical way to do that.
Used this way, alongside how someone actually feels and whatever psychological support they’re already engaged with, physiological markers add a layer of information that self-report alone can miss. Used any other way, as a mental health diagnosis, a replacement for therapy, or proof that nothing is wrong because the numbers look fine, they’re being asked to do something the research was never designed to support.










