07/16/2026
This may be of interest to gym bros....
The Pentagon is about to put soldiers on a hormone based on a single blood draw. Roughly a third of those readings will be wrong on a retest. Of the men who genuinely qualify, more than half get better on their own without treatment. This morning Pete Hegseth announced that every active-duty service member over 30 will be screened annually for low testosterone, framing it as keeping troops "on the leading edge of lethality" under a post reading "The High-T Department of War." There's a real medical program here, but most of what's being claimed around it isn't supported by the evidence.
About one in eighteen men aged 30 to 79 has low testosterone plus the symptoms that define an actual deficiency. Finding those men and treating them is standard, evidence-based medicine, and the rate may run higher in a military population. But that same number is the problem with screening everyone: the large majority of men getting tested are not deficient and never will be, and the guidelines are blunt about what happens when you test them anyway.
You don't screen healthy people for this, and the harm isn't hypothetical. Testing a man with no symptoms turns a lab value into a diagnosis he doesn't have, puts him on a drug he doesn't need, and commits him to years of therapy and its risks in pursuit of a number that was never causing him a problem. And most men don't stay on it. In a database study of more than 15,000 men who started testosterone gel, about 70% had stopped within a year. That's why the Endocrine Society makes a strong recommendation against routine testosterone screening in the general population, and states plainly that a low level without symptoms is not a diagnosis. Peter Snyder, who led the NIH testosterone trials, said it directly about this program: a low reading in a man with no symptoms is never a reason to treat.
A single reading is a weak thing to act on for a second reason. Over 30% of men who test in the deficient range read normal on a retest, so a large share of "low" results are just measurement noise. And among the men who do meet the full definition, low testosterone with symptoms, more than half saw it resolve on its own with no treatment when they were followed over time, most often in the younger and leaner ones. Screen on single numbers and much of what you find isn't a lasting problem.
The lethality argument rests on two ideas: that testosterone makes men more aggressive, and that it makes them physically better. Neither is supported by evidence.
On aggression: the belief that more testosterone means more aggression, and therefore more fighting capacity, is widely held and barely supported. The largest meta-analysis on the question found that baseline testosterone accounts for about one half of one percent of the variation in aggressive behavior in men. When researchers gave men testosterone directly to test cause and effect, the effect shrank to nothing statistically.
On performance: testosterone builds muscle and strength in a dose-dependent way across a wide range of doses. Within the normal male range, though, higher levels don't predict better performance. The androgen receptors in muscle are getting enough action across that range, so a man at the top of normal doesn't out-build a man in the middle of it. In a study of nearly 700 elite athletes, one in six of the men had testosterone below the standard male reference range and competed at the highest level anyway.
The most direct evidence comes from the Army itself. It ran two controlled trials giving soldiers testosterone during simulated field operations. In both, testosterone protected lean mass. In both, it failed to protect performance: strength, power, anaerobic and aerobic capacity all declined about the same as they did on placebo. No study has ever linked testosterone level, natural or supplemented, to marksmanship, target detection, or actual combat effectiveness.
Ranger School can reduce testosterone by half or more, but it's mostly from the Calorie deficit, and it comes back with food and rest. A low reading after a hard stretch isn't a disorder. Testosterone therapy treats the number instead of the conditions that produced it.
So, two very different things are riding under one announcement. Testing the small share of men who are genuinely deficient and symptomatic, the way the guidelines already lay out, is good medicine. Screening everyone else on a single blood draw, and promising sharper, more lethal soldiers by optimizing men who mostly aren't deficient to begin with, is a different claim. The Army put that claim to the test, and its own trials don't back it up.
Our book Signal (and its accompanying podcast series) digs into the whole evidence base on what testosterone does and doesn't do. Preorder link and citations for this post are in the comments.
What's the most confident testosterone claim you've heard that the data doesn't actually back up?
