Chronic stress adds up in the body, and two famous studies need a second look

Key takeaways

• In the American Psychological Association’s August 2025 poll, 76% of U.S. adults reported at least one symptom of stress in the past month, most often feeling anxious, tired or headachy.
• Pooled data on 197,473 European workers linked job strain to a 23% higher risk of heart disease, a real but modest risk next to smoking.
• A widely cited 1999 study tied strained spouse caregiving to a 63% higher risk of death, but at least eight later population studies found that caregivers as a group live longer than similar non-caregivers.
• In a famous 2004 telomere study, mothers who felt the most stressed had shorter telomeres, yet caregiving and non-caregiving mothers didn’t differ as groups, and one testing lab advises against testing healthy people.
• In national data from 1999 to 2002, Black adults showed more signs of biological wear and tear than white adults at every age from 18 to 64, and poverty didn’t explain the gap.

For more than two decades, one number has followed family caregivers: 63%. The claim is that caregivers die at a 63% higher rate than people who don’t give care. Caregiving and aging websites have repeated it, and the U.S. Administration on Aging has cited it too³.

The figure traces to a 1999 study by researchers Schulz and Beach³⁴. They followed 392 older adults who cared for a spouse and 427 who didn’t, for about four years¹⁵.

Read the fine print and the claim shrinks. The 63% higher risk of death applied to caregivers who said they felt strained, compared with people whose spouse wasn’t disabled. Caregivers without strain didn’t show the same rise¹⁵. A finding about strained spouses had turned into a warning about all caregivers.

Then other researchers looked at large, general populations. In a 2015 review in The Gerontologist, David Roth of Johns Hopkins University and his colleagues found five population-based studies in which caregivers as a whole had lower death rates and longer lives than non-caregivers³⁴.

By 2020, Roth counted at least eight such studies on three continents, with 16% to 26% lower mortality among caregivers³. No study, he said, had found significantly higher death rates for caregivers as a group than for comparable people who don’t give care³.

That doesn’t mean caregiving is easy, or that stress is harmless. It means the effects of long-running stress are real but more specific than the headlines. Strain, not caregiving itself, seems to be the risk. Some of the most famous findings about stress deserve that kind of careful second look.

Your stress system is built for emergencies, not for months

In 1998, Rockefeller University scientist Bruce McEwen laid out a framework that still shapes the field. Writing in the New England Journal of Medicine, he used the word allostasis for the way the body stays stable by changing: shifting heart rate, hormones, blood sugar and immune activity to meet a challenge¹.

Those changes protect you in the short run. The trouble comes when they run too often or too long. McEwen called the resulting toll allostatic load: damage that accumulates when stress systems stay overactive, or underactive, for too long¹.

He described four ways it happens¹:

• Too many hits. Stress comes again and again, such as repeated spikes in blood pressure.
• No getting used to it. The same kind of stress keeps triggering a full hormone response, long after it should feel routine.
• No off switch. The body stays on alert after the stressful event is over.
• Too little response. One system responds too weakly, and others overwork to make up for it.

The National Institute of Mental Health (NIMH) notes that anxiety that lingers and interferes with life can affect sleep and the immune, digestive, heart and reproductive systems. It can also raise the risk of an anxiety disorder or depression¹⁰.

Our article on the mind-body connection covers how stress affects colds and immune cells. Our gut-brain article covers digestion, and an upcoming article covers stress and sleep.

Wear and tear can be counted, roughly

How do you measure years of stress? Researchers in the MacArthur Studies of Successful Aging built a score from ordinary checkup numbers¹.

The score counted how many measures fell in a risky range. They included blood pressure, waist-to-hip ratio, long-term blood sugar, cholesterol and stress hormones collected in overnight urine¹.

Among high-functioning older participants, those who started with higher scores were more likely to develop cardiovascular disease. Over three years, they were also more likely to decline in thinking and physical function¹.

McEwen also pointed to Britain’s Whitehall studies of civil servants. Illness and death rose step by step from the highest job grades to the lowest¹. Job stress and unstable work were tied to high blood pressure¹.

An allostatic load score is a research tool, built from many numbers. No single number on your lab report measures it. But several of the ingredients, like blood pressure, waist size and blood sugar, are ones your clinician already tracks.

Work is one well-studied source of long-running stress. In 2012, Kivimäki and colleagues pooled data on 197,473 workers from 13 European cohort studies in The Lancet. Job strain, defined as high demands with low control, was tied to a 23% higher risk of heart disease over an average of 7.5 years².

That risk is real but modest. Job strain accounted for about 3.4% of coronary heart disease cases. The authors compared that with about 36% for smoking, an estimate from a separate global study². Published studies also showed bigger effects than unpublished ones, a pattern that can signal publication bias².

Our article “Work stress has two classic recipes, and both show up in heart studies” covers the research in depth.

Caregiving strain is the risk, not caregiving itself

Caregiving is common. In the American Psychological Association’s 2025 Stress in America survey, 22% of adults said they currently care for a family member or friend, and another 22% had done so in the past¹⁴.

Why might caregivers live longer than similar non-caregivers? Roth points to selection first: healthier people may be more likely to take on caregiving³. He also suggests caregiving may help some people through purpose, social contact and the physical activity of daily care³.

None of this erases hardship. About 20% of caregivers report high strain³. Roth and his colleagues urge aiming support at that smaller group, rather than painting all caregiving as a health threat⁴.

If you’re caring for someone and feel stretched thin, that’s a signal, not a weakness. An upcoming article on caregiver strain goes deeper into practical help.

Telomeres: a landmark study with a quieter footnote

Telomeres are protective caps on the ends of chromosomes. In 2004, a University of California, San Francisco, team led by Elissa Epel and Elizabeth Blackburn reported a striking link in PNAS⁵⁶. Blackburn went on to share the 2009 Nobel Prize for her telomere research⁸.

They studied 58 healthy premenopausal mothers. Thirty-nine were caring for a chronically ill child, and 19 had healthy children⁵⁶.

The 14 women who felt the most stressed had telomeres shorter by the equivalent of about a decade of extra aging, compared with the 14 least stressed. They also had lower levels of telomerase, an enzyme that maintains telomeres, and more oxidative stress⁵⁶. Among caregivers, more years of caregiving went with shorter telomeres⁵.

This is the footnote. As groups, the caregiving mothers and the other mothers didn’t differ on these cell markers⁶. The link was with how stressed each woman felt⁶.

The study was small and measured each woman once, so it couldn’t show cause and effect. UCSF psychiatrist Owen Wolkowitz told Psychiatric News that proving stress shortens telomeres, or that easing stress reverses the change, would take longer studies⁵.

The business side. Blackburn was among the founders of Telome Health, a telomere-testing company, according to trade press⁸. In a 2017 post on their UCSF center’s website, Epel and Blackburn wrote that years earlier they had started a company to help run a crowdsourced telomere study. They later chose not to be involved and donated all their equity to a nonprofit. They also wrote that they had no ties to, or financial interest in, the testing companies listed on that page⁷.

That same 2017 page is cautious about consumer tests. Mistakes can enter anywhere from the blood draw to shipping, storage and the lab assay itself. Telomere length seems to rise and fall over months, and there are no firm norms for what’s typical at each age⁷.

A result is not a diagnosis, the authors wrote. One testing lab, Repeat Diagnostics, doesn’t recommend measuring healthy people at all⁷. Our article on biological age tests covers similar problems with other “how old are your cells” products.

Discrimination and “weathering” leave measurable marks

Researcher Arline Geronimus uses the term “weathering” for early health decline driven by repeated social and economic hardship and political marginalization. In a 2006 study in the American Journal of Public Health, her team tested the idea with national data⁹.

They used the National Health and Nutrition Examination Survey for 1999 to 2002, covering Black and white adults ages 18 to 64. Each person got an allostatic load score based on 10 biological markers, including blood pressure, body mass index, blood sugar and cholesterol⁹.

Black adults had higher average scores than white adults at every age, and the gap widened with age⁹. Among people 55 to 64, 77% of Black adults had a high score, compared with 60% of white adults⁹.

Black women fared worst. Half had a high score by age 45, and more than 80% did by 64⁹.

Poverty didn’t explain the racial gap. In fact, the gap was largest among people who weren’t poor⁹. The data were snapshots rather than people followed over time, and they lacked key stress hormones such as cortisol, so the study can’t prove cause⁹.

Discrimination is a common stressor. In the 2025 Stress in America survey, 25% of adults called it a significant source of stress in their lives¹⁴. This research points to conditions people live in, not to personal weakness or poor coping.

What the evidence says, rated

Claim	Evidence level	What studies found
Long-running stress is linked to poorer health	Strong evidence	Consistent links with heart, metabolic and mental health measures12910
Job strain raises heart disease risk	Strong evidence	A modest 23% higher risk in nearly 200,000 workers; a small share of all cases2
Caregivers as a group die sooner	Evidence says no	At least eight population studies found longer lives; healthier people may take on care34
Highly strained caregivers face higher death risk	Promising but early	Best known from one 1999 study of older spouses; reviewers urge focus on strained caregivers415
Feeling stressed shortens telomeres	Promising but early	Small snapshot study; cause unproven56
A consumer telomere test tells healthy people their cell age	Not enough evidence	Measurement error and no firm age norms7
Racism-related stress adds biological wear	Promising but early	Large national snapshot; poverty didn’t explain the gap; racism itself not measured9

A stress load audit you can fill in

Natural health has long asked about root causes, and chronic stress is often one of them. This audit treats stress the way McEwen did: as a load that builds from many sources and can be lightened from several directions¹. It’s our own tool, not a validated questionnaire.

Source of ongoing stress	What you control	One small step	Who can help
Work demands	How you ask for clearer priorities	List your top three tasks and ask your manager to confirm them	Manager, HR, union
Caregiving	How much you take on alone	Name one task someone else could do this week	Family, friends, faith community
Money or housing	Which bills you tackle first	Make one call about a payment plan	A nonprofit credit or housing counselor
Discrimination	Who you talk to about it	Tell one trusted person what happened	Community groups, a therapist12
Loneliness	One contact a week	Text one friend today	Friends, clubs, volunteer groups
Health worries	Getting answers	Book the checkup you’ve been putting off	Your primary care clinician

McEwen urged clinicians to help people build coping skills, know their limits and relax, and to take isolation and lack of control at work seriously¹. NIMH’s coping list includes regular exercise, healthy meals, steady sleep, less caffeine, a journal and reaching out to supportive friends or family¹⁰.

Connection may deserve extra attention. In the 2025 Stress in America survey, 54% of adults said they felt isolated at least some of the time¹³. People with high loneliness were far more likely than those with low loneliness to report feeling depressed or sad, 65% versus 15%¹³.

Common stress symptoms, and when to get help

The 2025 Stress in America poll asked which symptoms people had in the past month because of stress¹⁴. The most common were:

• feeling nervous or anxious (37%) or depressed or sad (34%);
• fatigue (36%) and headaches (35%);
• irritability or anger (33%);
• an upset stomach or a change in appetite (22% each);
• muscle tension (19%);
• teeth grinding, chest tightness or dizziness (13% each).

These symptoms can have other causes, so don’t assume stress is the explanation. If they won’t go away, NIMH says that’s a reason to talk with a professional¹⁰.

Talk with a doctor or mental health professional if:

• stress gets in the way of your everyday life¹⁰¹¹;
• it makes you avoid things you need or want to do¹⁰;
• it seems to be there all the time¹⁰;
• you’re struggling to cope, or your symptoms won’t go away¹⁰¹¹;
• sadness lingers, which can be a sign of depression¹¹.

NIMH suggests asking a possible therapist about their training, approach, evidence base and fees¹².

If you or someone you know is in crisis or thinking about suicide, call or text the 988 Suicide & Crisis Lifeline at 988, or chat at 988lifeline.org¹². For emergency warning signs, see our guide, “The symptoms that mean call 911 now”.

Questions readers ask

The bottom line

Chronic stress is not just a feeling. When the body’s alarm system runs too often or too long, the toll shows up in blood pressure, blood sugar and other measures that predict disease. Job strain and the stress of discrimination track with real costs.

The famous numbers need context, though. Caregivers as a group don’t die sooner; strain is the risk. The telomere study that made headlines was small, and consumer telomere tests aren’t ready to guide healthy people. Lighten the load where you can, and get help when it won’t lift.

Related reading

• Stress can reach your immune cells, but positive thinking still can’t cure cancer
• Your gut and brain really talk, but many mood claims started in mice
• Biological age tests can disagree with themselves by years
• Work stress has two classic recipes, and both show up in heart studies
• Venting on a punching bag fed anger in a lab, and calmer skills work better

Sources

1. McEwen BS. “Protective and Damaging Effects of Stress Mediators.” New England Journal of Medicine, 1998;338(3):171–179 (University of Wisconsin course copy). psychiatry.wisc.edu
2. Kivimäki M, Nyberg ST, Batty GD, et al. (IPD-Work Consortium). “Job Strain as a Risk Factor for Coronary Heart Disease: A Collaborative Meta-Analysis of Individual Participant Data.” The Lancet, 2012;380:1491–1497 (London School of Hygiene & Tropical Medicine repository copy). lshtm-eprints-repo-prod.lshtm-eprints.cdl.cosector.com
3. Roth DL. “Recent Updates on the Health Effects of Family Caregiving Stress” (presentation slides). Johns Hopkins Medicine, The Academy, February 2020. hopkinsmedicine.org
4. Roth DL, Fredman L, Haley WE. “Informal Caregiving and Its Impact on Health: A Reappraisal From Population-Based Studies.” The Gerontologist, 2015;55(2):309–319 (abstract reprinted by the USC Leonard Davis School of Gerontology). tingstad.usc.edu
5. Psychiatric News. “Can Stress Reduction Fight Some Signs of Aging?” (on Epel ES, et al., “Accelerated Telomere Shortening in Response to Life Stress,” PNAS, 2004). American Psychiatric Association, 2005. psychnews.psychiatryonline.org
6. University of California, San Francisco. “Blackburn Explains Award-Winning Research.” UCSF News, March 2005. ucsf.edu
7. Epel E, Blackburn E. “Background Information and Our Views on Telomere Testing.” UCSF Aging, Metabolism, and Emotions Center, July 3, 2017. amecenter.ucsf.edu
8. GenomeWeb. “Telome Health Launches qPCR-Based Tool Measuring Telomere Length.” GenomeWeb (trade press). genomeweb.com
9. Geronimus AT, Hicken M, Keene D, Bound J. “‘Weathering’ and Age Patterns of Allostatic Load Scores Among Blacks and Whites in the United States.” American Journal of Public Health, 2006;96(5):826–833. pmc.ncbi.nlm.nih.gov
10. National Institute of Mental Health. “I’m So Stressed Out! Fact Sheet.” NIH Publication No. 20-MH-8125. nimh.nih.gov
11. Centers for Disease Control and Prevention. “Managing Difficult Emotions.” CDC, October 1, 2024. cdc.gov
12. National Institute of Mental Health. “Psychotherapies.” NIH, last reviewed February 2024. nimh.nih.gov
13. American Psychological Association. “APA Poll Reveals a Nation Suffering From Stress of Societal Division, Loneliness” (Stress in America 2025). APA, November 6, 2025. apa.org
14. American Psychological Association and The Harris Poll. “Stress in America 2025: Topline Data.” APA, November 2025. apa.org
15. Huffman GB. “Caregiving as a Risk Factor in Mortality” (Tips From Other Journals, summarizing Schulz R, Beach SR, JAMA, December 15, 1999). American Family Physician, 2000;61(10):3112–3115. aafp.org