Men of the same age in a Boston cohort had testosterone levels about 1.2 per cent lower for each later year of measurement between 1987 and 2004, a trend that ageing on its own did not explain.
The figure comes from Travison and colleagues, published in the Journal of Clinical Endocrinology and Metabolism in 2007 and indexed on PubMed as A population-level decline in serum testosterone levels in American men. The prospective cohort study covered randomly selected men aged 45 to 79 in greater Boston, Massachusetts, across three data collection waves, 1987 to 1989, 1995 to 1997 and 2002 to 2004.
"Testosterone levels among men in the Boston area declined among age-matched men from 1987-1989 to 2002-2004," a Healio summary of the study said.
What the study measured
The study was observational. Men were randomly selected from the greater Boston area rather than recruited through a clinic, and the research team returned to them in two later waves.
Data came from 1374 men at the first wave, 906 at the second and 489 at the third, totalling 2769 observations on 1532 men. The two outcome measures were serum total testosterone and calculated bioavailable testosterone.
The stated objective was to establish the size of any population-level change in serum testosterone concentrations and how far it could be explained by secular changes in relative weight and other factors.
Healio's report of the study sets out the baseline picture. Of the men seen in the first phase, 719 (52 per cent) reported at least one chronic illness, 340 (25 per cent) were current smokers and 296 (22 per cent) were obese with a body mass index above 30. Median serum testosterone was 501 ng/dL at baseline, 435 ng/dL at the first follow-up and 391 ng/dL at the second.
Three declines, one headline number
Healio's report separates three estimates, and they answer different questions.
The cross-sectional decline, which compares older men with younger men measured at the same time, was 0.4 per cent per year of age, with a 95 per cent confidence interval of 0.2 to 0.6 per cent lower. The longitudinal within-person decline, which follows the same men as they age, was about 1.6 per cent a year, with a 95 per cent confidence interval of 1.4 to 1.8 per cent lower.
The third estimate is the one in the headline. The age-matched time trend, which compares men of the same age measured in different calendar years, was 1.2 per cent lower per year, with a 95 per cent confidence interval of 1.0 to 1.4 per cent lower. Healio reports that the decline was age-independent.
Travison, quoted by Healio, compared a 60-year-old measured in 1989 with a different 60-year-old measured in 1995. "We observed the same phenomenon over a wide range of ages," he said.
The paper's context section says age-specific estimates of mean testosterone appear to vary by year of observation and by birth cohort, and that estimates of longitudinal declines typically outstrip cross-sectional decreases. Those observations are what prompted the test for a change over calendar time, independent of chronological ageing.
The abstract puts the result this way: the authors observe "a substantial age-independent decline in T that does not appear to be attributable to observed changes in explanatory factors, including health and lifestyle characteristics such as smoking and obesity." The paper also says the estimated population-level declines are greater in magnitude than the cross-sectional declines in testosterone typically associated with age.
The paper does not identify a cause. Its conclusions describe a decrease "potentially attributable to birth cohort differences or to health or environmental effects not captured in observed data."
Checking the video's 25 per cent
The accompanying VegOut video, What Your Diet Is Doing To Your Testosterone, includes a claim of close to 25 per cent. That figure is not a finding of this study.
Neither the 2007 paper nor Healio's summary reports a 25 per cent fall in testosterone. The 25 per cent that does appear in Healio's write-up is the 340 men who were current smokers at the baseline phase, a share of participants rather than a change in hormone levels.
A cumulative percentage cannot be built by multiplying 1.2 per cent by the number of years between the first and last wave either. The 1.2 per cent is an age-matched time trend, and that is the form in which Healio reports the result.
Diet is not among the factors named in the abstract. The study did not test whether any particular way of eating raises or lowers testosterone, so its results cannot settle that question in either direction. The authors state only that the decline does not appear to be attributable to the explanatory factors they observed, including smoking and obesity.
The study has limits. It covers one metropolitan area and men aged 45 to 79, and the number of men measured fell from 1374 at the first wave to 489 at the third. The dates need care too. Some secondary text, including the machine-readable abstract on PubMed, gives the baseline wave the wrong end year. The baseline wave ran from 1987 to 1989, which matches Healio's sentence on the age-matched comparison.
Other VegOut nutrition-research explainers include the Physicians' Health Study II multivitamin trial in 14,641 US male doctors, where the 0.8 g per kg protein RDA came from, and a survey of how many nutrition hours US medical schools taught.
Photo: Hospital corpsmen check a blood sample (illustrative U.S. Navy photograph, not from the Massachusetts Male Aging Study). U.S. Navy photo by Photographer's Mate 2nd Class Jeffery Russell, public domain, via Wikimedia Commons.




