The post-workout testosterone spike depends on how you measure it

Testosterone, cortisol and the ratio between them are often treated as indicators of anabolic status: a shorthand for whether the body is in a building-up or breaking-down state. Blood sampling made this awkward to study, but the arrival of salivary testosterone analysis made samples easy to collect, and published research multiplied rapidly as a result. The question this meta-analysis set out to answer is deceptively simple: how much do salivary testosterone, cortisol and their ratio actually change after a single bout of exercise? Given how often these hormones are invoked in training talk, pinning down the real size of the response matters.

The authors searched PubMed, Web of Science, Science Direct and SPORTDiscus up to 31 December 2013 for studies combining saliva, testosterone and exercise. From 933 records they selected 21 studies: 2 randomised controlled trials, 18 uncontrolled trials and 1 control trial. Each study involved an acute bout of aerobic, resistance or power exercise in men, with salivary testosterone and cortisol measured within 30 minutes of finishing. Changes were pooled as standard differences in means using a random-effects model, with subgroup analyses separating study designs and baseline sampling approaches.

Across exercise modes, salivary testosterone rose after exercise, with standardised effects of 0.891 for aerobic, 1.061 for resistance and 0.509 for power work; salivary cortisol rose too, most steeply after aerobic exercise at 3.041. The testosterone-to-cortisol ratio moved inconsistently, falling after aerobic and power exercise and barely shifting after resistance work. Crucially, study design changed the answer: in aerobic studies, randomised trials produced a larger effect than uncontrolled ones (1.337 versus 0.446), and in power studies the direction of the effect flipped depending on whether baseline saliva was taken 24 hours before or immediately before exercise. The authors concluded the current body of evidence is weak.

This analysis complicates the popular idea that a workout delivers a clean, reliable hormonal surge that can be read off a saliva sample. The hormones do respond, but the measured size, and in some cases the direction, of the response depended as much on methodology as on the exercise itself. That undermines confident use of the salivary testosterone-to-cortisol ratio as a simple anabolic or recovery marker. The authors call for more randomised controlled trials and a standardised measurement protocol before firmer conclusions can be drawn about how exercise modality shapes these hormones.

If you have ever seen post-workout testosterone or cortisol figures quoted as proof that one training style is more 'anabolic' than another, this is a reason for scepticism: much of that literature comes from uncontrolled studies whose results shift with sampling logistics. It is also a reason to treat single saliva readings, increasingly available through consumer testing, as noisy snapshots rather than verdicts on your hormonal state. The acute hormone response to exercise is real, but the evidence describing it is not yet strong enough to hang training decisions on.