Testosterone levels by age: 1987 versus today

A fall that ageing does not explain

The Massachusetts Male Aging Study (MMAS) took a random sample of men in the Boston area and measured them in three waves: 1987–89, 1995–97 and 2002–04. Because it is the same population each time, two things can be separated: the fall that happens as one man gets older, and the fall between generations when men of the same age are compared.

Feldman and colleagues (JCEM, 2002) first described the age-related fall in MMAS: total testosterone fell by about 1.6% a year, and free testosterone fell slightly faster because SHBG rose at the same time. Travison and colleagues (JCEM, 2007) then looked at the same data across time. Mean total testosterone in a 65-year-old measured in 1987–89 was about 501 ng/dL, while a 65-year-old measured in 2002–04 sat around 391 ng/dL. That corresponds to a fall of roughly 1% a year at population level, on top of ordinary ageing.

The authors tried to explain the fall with changes in weight, smoking and health. They only partly succeeded. Rising obesity explained some of it, but a substantial part of the difference remained. The study cannot say exactly why, only that the difference is not due to the men being older.

What does 1% a year mean?

One percent a year sounds small, but over 15 years it adds up to a noticeable difference. That is why a 65-year-old in 2004 was on average about a fifth lower than a 65-year-old in 1987. The difference applies to the average. Within both groups there were men with both high and low values, and many of the men in 2004 were still well within the laboratory range.

Not only an American finding

Andersson and colleagues (JCEM, 2007) compared Danish population surveys spanning several decades and found lower testosterone in later birth cohorts, together with rising SHBG and rising body weight. Perheentupa and colleagues (European Journal of Endocrinology, 2013) saw the same generational pattern in Finnish men: men born later had lower testosterone than men of the same age born earlier. Lokeshwar and colleagues (European Urology Focus, 2021) described falling total testosterone in American men aged 15–39 over the period 1999–2016, including men of normal weight.

The studies use different methods and populations, and none of them is perfect. But the direction is the same in several countries: men today are on average lower than men of the same age a generation ago.

Why is it happening?

Nobody knows for certain. The explanations most often mentioned in the literature are:

• more overweight and more abdominal fat, which lower testosterone and SHBG
• less physical activity and more sedentary work
• poorer sleep and more screen use at night
• more chronic disease and more medication use
• possible environmental factors, which researchers are still investigating and where the evidence is uncertain

It is important to keep the last two separate from the first ones. Overweight and sleep have solid support. The environmental explanations are hypotheses, not conclusions.

What it means for your own result

A laboratory reference range is based on the men in the reference group. If the whole population has shifted downwards, the range shifts with it. A number marked "normal" today may therefore be lower than what your father had at the same age. That does not make the number abnormal, but it means that "normal" does not always mean "optimal for you". You can read more in the article on why a normal testosterone result can miss the point.

At the same time, it is important not to over-interpret. The generational fall is an average for a population. It says nothing about where your particular value should be, and it is not a treatment target. The guidelines still use the same criteria for deficiency: symptoms plus repeatedly low morning measurements.

Age is still part of the picture

Within the individual man, testosterone falls slowly with age. Free testosterone falls fastest because SHBG rises. Harman and colleagues showed in the Baltimore Longitudinal Study of Aging (JCEM, 2001) that a much larger share of older men were low on the free testosterone index than on total testosterone. That is why it matters to measure SHBG and not just the total figure. You can find more in SHBG and free testosterone.

The European Male Ageing Study (EMAS, Wu et al., JCEM 2008) also found that overweight and poor general health were at least as strongly linked to lower testosterone as age itself. Many healthy older men have values comfortably within the young range.

What the study does not say

The Travison study is not proof that your result should sit where it did in 1987, and it is not a reason for treatment in itself. It is a reason not to read today's reference range as if it were your father's level, and to take both symptoms and numbers seriously.

A sensible next step

If you have symptoms such as low desire, fewer morning erections or persistent fatigue, a blood test is a good starting point. Take a fasting morning sample before 10 am with total testosterone, SHBG, albumin so that free testosterone can be calculated, and LH and FSH. The first test can be taken locally where you live. Repeat a low or borderline result on another morning, and go through the numbers with a doctor who looks at them together with your symptoms and history.