How past testosterone levels were measured

There is no hormone record before the assay

When you read that men today have less testosterone than their fathers and grandfathers, it is fair to ask: how do we know? Testosterone could only be measured reliably in blood once radioimmunoassay arrived in the late 1960s. Anything before that is guesswork, however convincingly it is told. That is why we do not publish a number for "historical man". We stick to the data that actually exist, and we say where they are weak.

There are basically three kinds of source: repeated population surveys, cohorts that follow the same men over time, and frozen blood samples that can be measured again with newer methods. Each has strengths and pitfalls.

Repeated population surveys

The Massachusetts Male Aging Study (MMAS) examined the same population of men three times between 1987 and 2004. That allowed Travison and colleagues (JCEM, 2007) to separate two things that otherwise blur together: the decline that comes from an individual man getting older, and the decline seen between generations. A 60-year-old man in 2004 had lower testosterone than a 60-year-old man in 1987. The difference could not be explained by age alone, and it persisted after accounting for weight, smoking and health.

In Denmark, Andersson and colleagues (JCEM, 2007) showed a similar pattern in Danish population surveys: men born later had lower testosterone at the same age than men born earlier. Perheentupa and colleagues (European Journal of Endocrinology, 2013) found the same in Finnish population surveys. A finding that repeats across countries with different laboratories is more credible than a single study.

Cohorts that follow the same men

The Baltimore Longitudinal Study of Aging followed the same men for many years with repeated blood samples. Harman and colleagues (JCEM, 2001) used these data to describe how testosterone falls with age in the individual man, and how many eventually drop below the limits laboratories use. That kind of data tells you about ageing, but not by itself about generations. This is exactly why you need both cohorts and repeated cross-sections if you want to say anything about change over time.

Frozen serum banks

Some countries have stored blood samples for decades. Old samples can in principle be measured again on modern platforms, so all numbers come from the same method. But storage time, thawing and tube type can introduce errors, which is why such re-measurements are cautious.

Gislefoss and colleagues (2012) measured frozen blood from the Norwegian Janus serum bank, up to 29 years old, with both immunoassay and mass spectrometry (LC-MS/MS). The oldest samples were substantially higher than fresh ones. That matters because storage typically degrades hormones. If storage affected the result, it would have made the old numbers lower, not higher. Storage error cannot explain a decline over time; if anything, it understates it.

The method problem

Immunoassay and mass spectrometry do not always agree, especially at low values and in women and children. Laboratories have also changed equipment and calibration over the years. Part of any historical curve can therefore be method drift rather than biology. The best studies solve this by measuring stored samples on one single platform, or by only comparing within the same method. When we show a curve on this site, it rests on a named source, and we say what it can and cannot carry.

Why this matters for you

Reference ranges are built from the men a laboratory has measured recently. If the whole population sits lower than a generation ago, "normal" moves with it. A result printed as normal only says that you resemble the men around you today. It does not say whether the number fits your symptoms. We have written more about that in why your blood results are normal.

Total testosterone is not the whole picture either. A large share is bound to SHBG and is not freely available to tissue. Two men with the same total can have very different free testosterone. Read more in SHBG, the hormone that decides your free testosterone.

What historical numbers cannot do

A historical average is not your treatment target. Population data can explain why many men have symptoms and "normal" results at the same time, but they cannot say what is right for you. That is decided by your symptoms, repeated morning samples and a doctor who looks at everything together. This is also why we always look at several measurements rather than one number from one day. Other causes of low testosterone, such as weight, sleep, medication and pituitary disease, must also be considered.

Next step

Write down your symptoms and how long you have had them. Take a morning sample with total testosterone, SHBG and calculated free testosterone. The first test can be taken locally, where you live. The result is not a diagnosis in itself, but a starting point for a conversation with a doctor. Any decision about treatment belongs to the doctor.