During the past two decades, testosterone levels in American men have rapidly declined.
This information comes from a long-term prospective study that evaluated changes in serum testosterone on a population-wide basis.
The study was published in the Journal of Clinical Endocrinology and Metabolism.
“The interesting thing we discovered was that, on average, when we measured the testosterone in the blood of a 60-year-old in 1989 it was higher than that in a different 60-year-old measured in 1995,” said Thomas Travison, PhD, of the New England Research Institutes, Watertown, Mass. “We observed the same phenomenon over a wide range of ages.”
Although testosterone loss is common as men age, it is often associated with diabetes, abdominal obesity, sexual dysfunction, depression and other adverse conditions.
The Massachusetts Male Aging Study was composed of randomly selected men (aged 45-79 years) living in the Boston area. The men were initially sampled between 1987 and 1989 (n=1,374). There were two follow-up periods: 1995 to 1997 (n=906) and 2002 to 2004 (n=489). There were 2,769 total observations from 1,532 men.
The age-matched study was conducted to measure participants at different times and generations, according to Travison.
In each wave, the researchers took blood samples and conducted extensive interviews, Travison said. Participants were asked about demographic and lifestyle characteristics, self-assessed general health, smoking and alcohol habits, BMI, waist-hip ratio, physical activity and other factors.
At baseline, the median serum testosterone level was 501 ng/d; at the first follow-up it was 435 ng/dL and at the second follow-up it was 391 ng/dL.
The estimated cross-sectional decline in total testosterone level was 0.4% per year of age (95% CI, –0.6% to –0.2%). The longitudinal within-person decline was about 1.6% per year (95% CI, –1.8% to –1.4%). The age-matched time trend was 1.2% lower per year (95% CI, –1.4% to –1.0%).
The decline was age-independent. “It is a little troubling,” Travison said. “The average differences are not very large, but they are big enough and occurring over a short enough time period to be the cause of some concern.”
These demonstrated population-level declines are greater than the cross-sectional declines typically associated with age, according to the researchers.
The researchers concluded that the decline in testosterone “does not appear to be attributable to observed changes in explanatory factors, including health and lifestyle changes such as smoking and obesity.” These factors are potential explanations both because they are associated with testosterone levels and also because they have changed across generations, Travison said.
For instance, there were increases in obesity and medication use, but a decrease in smoking. By the last phase of the study, the researchers found that 719 (52%) of the participants reported at least one chronic illness, 340 (25%) were current smokers, 296 (22%) were obese (BMI >30) and 300 (22%) reported using at least three prescription medications.
The researchers wondered whether these factors are contributing to the generational decline. “We know that people in all age groups are heavier today than they used to be. Similarly, they’re taking a lot more medication, which has an influence on endocrine health, and they are smoking a lot less, which is good. Smoking is associated with higher testosterone levels; if you stop smoking that can bring levels down,” Travison said.
Thus far, however, it does not appear that these factors can explain the declines in testosterone, Travison said.
Source: J Clin Endocrinol Metab. 2007;92:196-202.
The findings of another study by Travison and colleagues (J Clin Endocrinol Metab. 2006:doi10.1210/jc.2006-1859) indicated that health and lifestyle changes, such as developing diabetes, weight gain or loss of a spouse, can have as much impact on testosterone levels as normal aging during a short to mid-length period of about 10 years.
Changes such as weight gain, unemployment and inactivity, can be prevented and treated. Population-level changes, as demonstrated by Travison et al, are outside of an individual’s control.
According to Travison, “It’s the same old song and dance. For an individual man, a healthy lifestyle, being active, getting exercise may help in the long run.”
“This population-level decline in testosterone concentrations in men is not explained fully by the usual suspects: increasing BMI and prevalence of obesity, certain other co-morbid conditions or decreasing incidence of smoking. Although the analysis by Travison et al did reveal significant age-related increases in adiposity and medication use and a welcome decline in smoking, the age-matched decline in testosterone concentrations persisted even after adjusting for these variables,” Shalender Bhasin, MD, of the section of endocrinology, diabetes, and nutrition, Boston University School of Medicine, wrote in an accompanying editorial.
He voiced concern over the decline and its public health impact on American men.
“This magnitude of change during such a short period is disquieting,” Bhasin wrote. “Although increasing adiposity and lifestyle factors that were recorded in the [Massachusetts Male Aging Study] could not account for the secular trends in testosterone level, it is possible that other lifestyle factors, such as increasing use of tight-fitting underwear, increasing room temperatures in American homes and offices during the past three decades, decreased physical activity with increased body mass indices and decreased smoking could have contributed to the declining testosterone levels in men.”
There are several limitations of this study, including a small age-matched trend and limited geographic region surveyed, but the researchers said these data should not be ignored. –by Katie Kalvaitis
For more information:
- Travison TG, Araujo AB, O’Donnell AB, et al. A population-level decline in serum testosterone levels in American men. J Clin Endocrinol Metab. 2007;92:196-202.
- Bhasin S. Secular decline in male reproductive function: is manliness threatened? J Clin Endocrinol Metab. 2007;92:44-45.