Peter Attia is not a man who changes his mind lightly. He's a physician who runs his own longevity practice, reads primary literature obsessively, and has made a career out of being rigorous about what the evidence actually shows versus what people want it to show. So when he says he's "way way way more bullish on sauna" than he used to be, that's worth sitting with for a moment.
His original position was reasonable: sauna feels good, maybe helps sleep, probably isn't doing anything profound for your lifespan. That's a defensible read if you haven't gone deep on the data. But the Finnish prospective studies — 2,000 participants tracked over two decades — produced numbers that are genuinely hard to dismiss. Thirty to forty percent relative risk reduction in mortality for people using sauna more than once per week. An 18 percent absolute risk reduction. These aren't marginal signals. These are the kinds of numbers that show up in pharmaceutical trials and make headlines.
Attia's honest about the limitation: this is prospective observational data, not randomized controlled trials. People who use saunas regularly might also exercise more, sleep better, stress less. Confounders exist. But as he notes, the burden of evidence in the non-randomized data has become difficult to ignore — and this tracks with everything Rhonda Patrick has published on the same Finnish cohorts. She found a 63 percent reduction in sudden cardiac death and a 66 percent lower risk of Alzheimer's for the highest-frequency users. Multiple researchers, same dataset, consistent direction of effect.
The mechanisms Attia points to — heat shock proteins and nitric oxide — are well-supported. Heat shock proteins clear misfolded cellular debris, the molecular junk that accumulates as we age and clusters into the plaques associated with neurodegeneration. Nitric oxide drives vasodilation, improving blood flow, dropping blood pressure, and replicating many of the cardiovascular adaptations of aerobic exercise. Neither mechanism is exotic or contested. Both are well-characterized in the literature.
The disagreement isn't about whether sauna works — the evidence on that is increasingly unified. The real debate is about why it works, and which mechanisms are doing the heavy lifting. Attia says "it's many things," which is honest and probably correct. Huberman emphasizes the growth hormone response. Patrick focuses on heat shock proteins and inflammatory markers. Some researchers think the cardiovascular mimicry of exercise is primary; others think the neurological effects of the heat-induced endorphin-dynorphin cascade matter more for mental health outcomes.
This isn't a problem. It's actually a feature. Multiple mechanisms converging on similar outcomes means the intervention is robust. You don't need to pick the right theory — the practice itself works regardless of which pathway you're most interested in.
Frequency matters more than duration. The Finnish data shows the dose-response relationship clearly: more than once per week is better than once per week, and four to seven times per week produces the most significant outcomes. If you're deciding between a single long heroic session per week and four shorter sessions, take the four. Consistency compounds. That's true of exercise, sleep, and apparently, heat.
And Attia's caution about alcohol is worth repeating: sauna while drinking is a cardiovascular risk. The heat is already stressing your vasculature in productive ways. Alcohol layered on top can tip that stress from adaptive to dangerous.
What strikes me most here isn't the physiology — it's the epistemology. Attia built his reputation on demanding randomized controlled trials, and yet here he is, updating his clinical recommendations based on prospective observational data because the signal is simply too strong to ignore. That's intellectual honesty in practice. The lesson for how we approach any wellness protocol: don't dismiss non-randomized data when the effect size is large and the mechanistic story is coherent. The absence of a perfect study isn't the same as the absence of evidence.