The cold-shower immunity claim: an honest read

Cold showers do not boost the immune system in any way an immunologist would use the word 'boost.' The one large randomised trial everyone quotes, the 2016 Dutch PRESSURE study, found that people who ended their morning shower with cold water reported roughly 29 percent fewer sickness absences from work over three months than people who did not. The number of illnesses did not change. That is a real, useful, and much smaller finding than what the internet has done with it.

The rest of the immunity claim is inference: cold-water swimming self-reports, a few small studies that tracked white-cell counts for a couple of hours after a plunge, and a lot of confidence borrowed from mechanism papers that never followed the participants long enough to know whether the counts meant anything. PRESSURE is the closest thing to a clinical answer, and even PRESSURE did not measure immunity. It measured whether people felt well enough to show up at work.

What the PRESSURE trial actually found

The paper: Buijze and colleagues, published in PLOS ONE in September 2016. Roughly three thousand Dutch adults were randomised into a control group (normal showers as they liked them) and three cold-shower groups that ended a normal warm morning shower with thirty, sixty, or ninety seconds of cold water. The mandated intervention ran thirty days, followed by a sixty-day continuation period in which cold showering was voluntary. The primary outcome was self-reported sickness absence from work across the ninety-day period.

The result: the cold-shower groups took about 29 percent fewer sick days across the three months than the control group did. Statistically, the three cold durations were indistinguishable from each other, meaning thirty seconds of cold was as effective as ninety. Adherence was high; roughly two thirds of the people randomised into the cold groups kept doing it voluntarily after the trial ended, which is the sort of finding that gets less press than the headline number and probably deserves more.

The second, quieter finding was that the number of illness days did not differ between the groups. Cold-shower participants got sick as often as controls; they just went to work more of the days on which they were sick. This is not the same as boosting an immune system. It is closer to a story about a discipline effect: people who committed to a thirty-day cold habit tended to also commit to a slightly higher I-can-handle-a-Tuesday threshold.

What the paper did not measure

No blood work. No immune-cell counts, no antibody titres, no cytokine panels, no infection markers of any kind. Sickness absence was self-reported by participants, meaning the raw ingredient for the headline effect is whether a human being decided a given morning was a work morning or a couch morning. That is a real endpoint that matters to employers and to household budgets. It is not what a clinical immunologist means by the phrase 'boost your immune system,' and PRESSURE never claimed it was. The people quoting PRESSURE later did.

There is also the unblinded-trial problem, which is not really a problem, more a fact of life for shower research. Everyone in the cold-shower groups knew they were in a cold-shower group; it is genuinely hard to fool someone about whether they took a cold shower. That means a portion of the sick-day effect is compatible with a placebo or a self-discipline confound. It does not make the finding fake. It does mean the causal mechanism is ambiguous, and 'reduces sickness absence' is a different sentence from 'improves immune function' even if you personally believe they mean the same thing.

So does the immune system actually get anything?

Two things can be true at once. Regular cold-water contact appears to reduce the number of sick days people take from work, in one large randomised trial, with an effect that survived several sensitivity analyses. And the biological pathway from a cold shower to a measurably better clinical outcome has not been demonstrated at the population level anywhere in the published record. Small mechanism studies show acute shifts in white-cell counts and stress hormones for a couple of hours after cold-water exposure. Those shifts have never been convincingly linked to anyone getting sick less often, which is a familiar shape across the wider cold-shower literature: acute changes are easy to measure, and chronic outcomes are much harder.

The reader-useful version: 'cold showers reduce the sick days you take' has one large study behind it. 'Cold showers boost your immune system' is a phrase used mostly by people who did not read the study.

So how long does the cold part need to be?

PRESSURE found thirty seconds of cold at the end of a warm shower worked as well statistically as ninety. That does not mean thirty is a target dose; it means the trial did not have the resolution to tell the three durations apart. What is defensible from the paper is that a very short cold finish counted as the intervention. If a longer duration feels better, no evidence in this literature says it does more; and no evidence says it does less either. The practical version of how long a cold shower needs to be lives in its own piece.

Does the effect keep working after you stop?

PRESSURE did not follow people long enough to answer that with confidence. What it did find, almost incidentally, is that about two thirds of the cold-shower group kept doing it voluntarily once the study ended. That does not prove anything about immunity. It suggests something the study was not designed to measure, which is that people who tried cold showers for thirty days generally decided they liked cold showers. Whatever that means for immune outcomes stays unknown; it does say something about how well the habit sticks once the shock is gone.

Is a cold plunge better than a cold shower for this?

Nobody has run PRESSURE with plunges. The available immunity-adjacent research is dominated by cold-shower self-reports and cold-water-swimming observational cohorts; ice-bath-specific immune data is a handful of small mechanism studies with sample sizes in the low double digits. If the reader is chasing fewer sick days, what has been tested at scale is a cold shower. A plunge might do the same, more, or nothing extra on top; the papers do not know yet.

For a week-by-week version of what actually happens when someone commits to a month of cold showers, the 30-day cold shower experiment piece walks the arc. The immunity headline is one reason people start; it is rarely the reason they finish.