U.S. cohort study finds yoga and Pilates participation associated with lower mortality risk, not proven longer life

Current assessment

The available evidence supports a narrower claim than the proposed headline: a large U.S. observational study found that people reporting yoga or Pilates use had lower all-cause mortality during follow-up after statistical adjustment for measured differences. It does not show that yoga or Pilates caused longer lifespans.

The paper, published online June 2, 2026 in the American Journal of Preventive Medicine, is a legitimate peer-reviewed research development involving Arizona State University researchers. But the strongest defensible interpretation is an association with lower mortality risk, not proof of a longevity benefit.

Reported facts

Athletech News reported on October 21 that a new Arizona State University-linked study connected yoga and Pilates with longer lifespans. The underlying research was already available online by June 2, and Arizona State University published an institutional explanation on September 4.

The study, led by Arizona State University researcher Connor M. Sheehan with co-authors Hyungmin Cha, Paul A. E. Espinoza and Yuan P. Xie, analyzed National Health Interview Survey data from 2007 and 2012. The researchers linked records for 55,023 U.S. adults to mortality data through 2019. The follow-up dataset included 5,530 deaths across 472,636.5 person-years.

After full adjustment and a false-discovery-rate correction across the practices examined, yoga use was associated with a hazard ratio of 0.73 for all-cause mortality, with a 95% confidence interval of 0.62 to 0.87. Pilates use had a hazard ratio of 0.64, with a 95% confidence interval of 0.49 to 0.85. Expressed as relative differences, those estimates correspond to 27% and 36% lower observed mortality hazards, respectively.

Supporting evidence

The study examined 21 complementary and alternative medicine practices, not yoga in isolation. Its abstract reports that most individual practices had null associations after adjustment and multiple-testing correction, while yoga and Pilates retained inverse associations with mortality. That makes the finding more substantial than a single unadjusted comparison, and the use of nationally representative survey data linked to death records gives the analysis population-level relevance.

The researchers also reported that the models accounted for demographic and health differences. In its September 4 summary, ASU said these included income, education, exercise, smoking, body-mass index and self-rated health. The paper’s abstract further identifies Cox proportional-hazards modeling, hierarchical covariate adjustment and Benjamini-Hochberg false-discovery-rate correction.

The available record also supports reporting that the work is exploratory. The authors wrote that broader complementary-and-alternative-medicine participation was associated with lower mortality, but that it remained unresolved whether the pattern reflected practice-specific benefits or general physical activity and health-conscious selection.

Limitations and competing evidence

The central limitation is causal inference. This was a prospective cohort study based on reported behavior, not a randomized trial assigning participants to yoga, Pilates or no practice. Statistical adjustment can reduce measured differences between participants and nonparticipants, but it cannot fully remove unmeasured differences in health status, access to care, diet, sleep, social connection, wealth or motivation.

ASU’s own account gives those concerns substantial weight. Sheehan said people who practice yoga or Pilates may already be more likely to engage in other health-supporting behaviors, while people in poorer health may be less able to participate. That explanation is consistent with the observed association and is not ruled out by the study.

The accessible abstract also does not establish a dose-response relationship for yoga or Pilates specifically, identify which styles or settings were involved, or show whether the findings apply equally across age groups, health conditions, practice frequency or home versus group practice. The article reports exploratory subgroup analyses as suggestive but largely nonrobust. Those details limit claims about who benefits and how.

Current conclusion

A cautious reading is that yoga and Pilates were associated with lower mortality in this nationally representative U.S. cohort, even after adjustment for several measured health and socioeconomic factors. That is a meaningful research finding and a credible yoga-industry news item.

It does not establish that taking yoga or Pilates causes a person to live longer, nor does it establish that one practice is superior to other forms of physical activity. Full-text review of the exposure definitions, model tables and sensitivity analyses would materially strengthen,or potentially narrow,the methodological account, but it would not change the basic need to frame this result as an observational association.