Testosterone Is Falling Worldwide: Why Every Generation Has Less

Perhaps you have sensed it yourself, or heard it from your father. Men today often seem more tired, softer, less resilient than the generation before. Part of that is story and nostalgia. Another part, though, can be measured. Men's testosterone levels really have fallen over recent decades, and not only because the individual man grows older.

In this article we look at what this cross-generational decline actually means, how large it is and why it is so easily overlooked. We understand why the reference range has dropped along with the average, which factors may be behind it and what you can do in your own everyday life, without sliding into panic or false promises.

The decline shows across generations, independent of age

Let us start with the key distinction. That testosterone slowly falls with age has long been known. This affects every man who grows older. Something different is the secular decline, the drop across calendar years and birth cohorts. It means: a man of the same age today has on average less testosterone than a man of the same age twenty or thirty years ago. Both effects add up.

This finding was at first an American observation. Travison and colleagues placed it in context in 2009 in Current Opinion in Endocrinology, Diabetes and Obesity and describe a possible secular decline of up to roughly one percent per calendar year, on top of the age trend. They name changes in body composition as part of the explanation and discuss environmental chemicals as a still unproven but conceivable contributor (doi:10.1097/med.0b013e32832b6348, PMID: 19396984).

For a long time it remained open whether this was a single finding. A very large review from 2025 closed this gap and confirmed the trend worldwide.

And now you know why this decline is more than a headline. It is visible in independent datasets across decades and across continents. The magnitude of around one percent per year sounds small. Exact figures vary by study and measurement method, the most commonly cited estimate is up to about one percent per calendar year, on top of the age trend. Summed across a generation, it adds up.

Why the reference range falls when the population falls

Here comes the thought that surprises many. When your doctor says your level is normal, they mean a statistical comparison with other men of your time. The reference range is calculated from large samples of present-day men. When the average of these men falls across generations, the lower limit of the normal range inevitably falls with it. The yardstick drifts along with what it is meant to measure.

This is no reproach to laboratory medicine. Reference ranges have to track the living population, otherwise they would not be practical. But it means the normal label is a relative statement, not an absolute one. A level can be in the green on paper and still be lower than the level your father or grandfather had at the same age.

The four KPNI lenses on the decline

In clinical psychoneuroimmunology, KPNI for short, we look not at an organ but at a network. The cross-generational decline can be understood well through four interwoven lenses. Each explains a part at the cellular level, and together they give the picture of why modern living presses on testosterone.

These four lenses explain why the decline is not the fate of a single hormone but the perceptible end of many threads. And now you know why it is worth looking, with testosterone, not only at the testes.

What can drive the decline: weight, sleep and environment

Let us look at the factors one by one. The most strongly documented is body weight. Obesity is the condition most closely associated with low testosterone in men, and across generations the average weight of the population has increased markedly.

These mechanisms also act on fertility, not only on the level. A review by Lien Davidson and colleagues in 2015 in Human Fertility describes how obesity can impair sperm production through the disturbed balance of testosterone and estrogen as well as through insulin and leptin, and places this in the global rise of obesity rates (doi:10.3109/14647273.2015.1070438, PMID: 26205254).

The second factor is sleep. A large part of testosterone production happens at night, coupled to sleep. Whoever sleeps too little over years could be sawing at one of its foundations. This too has grown over generations, because screens, shift work and constant availability have shortened many men's sleep.

That sleep and stress hormones together burden metabolism is shown by a controlled study from Peter Liu and colleagues in 2021 in the Journal of Clinical Endocrinology and Metabolism. In 34 young men, insulin resistance arose after four nights of only four hours of sleep. When cortisol and testosterone were held artificially at normal levels, this was about half as pronounced, a hint that the shift in these hormones could be one of the mechanisms involved. The raised blood sugar itself, however, could not be caught this way (doi:10.1210/clinem/dgab375, PMID: 34043794). In the population the picture is less uniform than one might think. An NHANES analysis by Jesus Hernández-Pérez and colleagues in Andrology found no simple link: in young men between twenty and forty, short sleep went along with higher levels, while in men between forty-one and sixty-four it was the longer sleep duration that was associated with low testosterone. For sleep quality the work found no association at all. That is a snapshot, not cause and effect. Above all it shows that findings from the sleep laboratory do not transfer one to one to the population (doi:10.1111/andr.13496, PMID: 37452666).

The third, still more open factor is environmental chemicals. Certain plasticizers in plastics and personal care products can act hormone-like. Here the research in humans is not yet conclusive, but there are first observations.

Here ends what studies show with certainty. That obesity and sleep play a large role is well documented. That environmental chemicals contribute is plausible and supported by observational and animal data, but not yet proven by large long-term human studies. This honest distinction is part of the matter.

What the data say about treatment, and what they do not

Before anyone concludes from the decline that all men would need hormones, a sober look is worthwhile. A single low value is not a diagnosis. And testosterone replacement therapy is not a lifestyle remedy. Testosterone is a prescription-only medicine, and the decision about it is a medical decision with benefits and risks.

What is well documented: lifestyle acts through weight. A large Australian study, the T4DM trial by Gary Wittert and colleagues in 2021 in The Lancet Diabetes and Endocrinology, examined more than a thousand overweight men with prediabetes over two years. In addition to a lifestyle program, one half received testosterone. In the testosterone group fewer men developed diabetes, 12 instead of 21 percent after two years. Important for context: in Germany testosterone is not approved for the prevention of diabetes. Such a use would be off-label, that is outside the marketing authorization, with all the consequences for liability and reimbursement. The authors themselves stress that long-term safety and cardiovascular effects of this use remain open. On top of that: in 22 percent of the treated men the hematocrit rose above the safety threshold, that is the thickening of the blood, compared with 1 percent on placebo. In 23 percent the PSA value reached the safety threshold, compared with 19 percent on placebo. Both are reasons to limit or stop a treatment (doi:10.1016/S2213-8587(20)30367-3, PMID: 33338415).

So the honest reading is: the decline shows across generations, but it is no carte blanche for hormones on suspicion. In most men it is worth looking first at the basics. And when a genuine deficiency with symptoms is present, the decision about a therapy belongs in medical hands, with a clear indication and monitoring.

Three levers you hold yourself

You will not reverse the trend of the whole population as an individual. But the factors that drive it are largely the same ones you can influence personally. These three levers are a start, not a treatment plan. You find your individual path with medical guidance.

And if symptoms remain despite good basics, diagnostics belong to it that look at the whole picture, not only at a single value. Testosterone should be measured in the morning and ideally more than once, together with the control hormones LH and FSH, with prolactin, blood count, thyroid, iron and ferritin as well as blood sugar. Prolactin definitely belongs in there, because a raised value can point to a treatable disorder of the pituitary gland. A markedly raised ferritin with high transferrin saturation can point to an iron storage disease, which is likewise treatable. Just as important is the conversation: earlier or current use of anabolic steroids, opioids or cortisone can dampen the control for years. This way treatable causes can be narrowed down, instead of attributing symptoms prematurely to one hormone or to a population trend.

Frequently asked questions about the testosterone decline

Connections to other topics