2.1 percent, or 20 percent. The European Male Ageing Study looked at 2966 men and required both a low level and matching sexual symptoms before counting anyone. It found late onset hypogonadism in 2.1 percent. The Baltimore Longitudinal Study of Aging looked at 890 men and judged on blood level alone. It found low testosterone in about 20 percent of men over 60.
Same hormone, same era, wildly different answers. What separates those two figures is the entire point of this page.
Two large studies asked how common low testosterone is in aging men. One came back with 2.1 percent. The other came back with about 20 percent. Neither was sloppy. Neither was lying. They were not measuring the same thing.
Understanding why they disagree is the most useful minute you will spend on this subject, because the disagreement is not really about testosterone. It is about what counts as having something.
The strict answer
The European Male Ageing Study followed 2966 men and used a deliberately demanding definition. A man was counted only if his testosterone was low and he had the sexual symptoms that go with it. Both, in the same man, at the same time.
Split by age, that headline flattens something important.
- 40 to 49: 0.1 percent
- 50 to 59: 0.6 percent
- 60 to 69: 3.2 percent
- 70 to 79: 5.1 percent
Read that at fifty and it is reassuring. Read the same table at seventy and it is a different document entirely.
The loose answer
The Baltimore Longitudinal Study of Aging followed 890 men and asked a simpler question: is the blood level low? Not whether the man felt anything. Just the number.
- Over 60: about 20 percent
- Over 70: about 30 percent
- Over 80: about 50 percent
Same hormone. Same era. One study says two men in a hundred, the other says one in five.
Where the gap comes from
The difference is made almost entirely of men whose blood test looks low and whose life is fine. They exist in very large numbers, and the strict study excluded every one of them.
Reference ranges are built from populations, not from you. A value below a cutoff is a statistical statement about where you sit in a distribution. It is not a verdict, and on its own it is not a reason to do anything at all.
The trap runs the other way as well
It is tempting to take that and conclude the number is meaningless. It is not. The strict definition cut in both directions: a man with all the symptoms and a normal level was also excluded, correctly, because something else was producing his symptoms. Fatigue, low mood, lost drive and a thickening waist have a long list of causes and most of them are not hormonal.
So the symptom questionnaire on its own overcounts. The blood test on its own overcounts. Each is wrong in a different direction, which is precisely why using one without the other produces the two numbers at the top of this page.
What is happening underneath both studies
They are sampling the same slow slope. Total testosterone falls by roughly 0.4 percent a year in men aged 40 to 70, beginning gradually from around age 35. The free fraction, the portion not locked to carrier proteins and therefore the portion your body can use, falls by roughly 1.3 percent a year, more than three times faster.
That second figure is why some men with a total testosterone sitting comfortably inside the normal range still feel exactly what they feel, and why a report showing only a total is showing you part of the page.
What this means for your result
Three lines, and they are the whole argument.
- A number without symptoms is not a diagnosis.
- Symptoms without a number are not a diagnosis either.
- The two read together, by someone who knows what else produces the same picture, is the only thing that settles it.
Everything else is somebody selling you their preference. A business that treats on the number alone will never run short of patients. A doctor who refuses to test because you are only tired will miss the men who genuinely have this. Both are cheating, in opposite directions, and you can usually tell which one you are sitting in front of inside ten minutes.
The honest limits
What these figures cannot tell you. Both studies are observational. They describe populations, not individuals, and a prevalence in a research cohort is not a probability for the man reading this. Definitions of low testosterone still vary between guidelines and between laboratories, which is part of why the figures move around so much in the first place.
Anyone who quotes you one confident percentage has read one of these studies and not the other. What a correct assessment involves, and why the hour at which you give blood changes the answer, is set out in how it is diagnosed.