Letter to the Editor: “Harmonized Reference Ranges for Circulating Testosterone Levels in Men of Four Cohort Studies in the United States and Europe”
Year: 2017
Type: letter
Abstract: The report by Travison et al. (1) entitled “Harmonized Reference Ranges for Circulating Testosterone Levels in Men of Four Cohort Studies in the United States and Europe,” touched a very important area in contemporary medical and health research. The reference ranges for androgen levels are highly needed by researchers and physicians alike to manage aging, male sexual and reproductive health, and many other health-related issues. From the description presented in their report, the process of research is valid, including the research design, sample selection, hormone measurement, measurement harmonization, and summarization analysis. However, we are deeply puzzled by the results that the estimated median and interquartile range (IQR) of the total testosterone level remain almost unchanged for men from age 50 to 99 years, a period of five decades. From the harmonized results, the median total testosterone level would be 531 ng/dL (IQR, 424 to 643) for men aged 13 to 39 years; 481 ng/dL (IQR, 286 to 608) for men aged 40 to 49 years; 477 ng/dL (IQR, 604 to 374) for men aged 50 to 59 years and 60 to 68 years; 477 ng/dL (IQR, 604 to 372) for men aged 70 to 79 years, and 476 (IQR, 604 to 362) for men aged 80 to 89 years. Although the authors did note this issue in their statement “Notably, results obtained in this study show a lesser age trend than that reported previously in cross-sectional analyses of men of different ages, underscoring the need for longitudinal studies of the effects of aging on sex steroid concentrations.” No additional explanation was provided. Similar results were also reported in another study in PloS One, entitled “A Validated Age-Related Normative Model for Male Total Testosterone Shows Increasing Variance but No Decline After Age 40 Years” (2), with no explanation of why testosterone does not decline with age. Such results are hardly acceptable in either theory or practice and are very misleading. It violates the basic knowledge of normal aging. To our understanding as epidemiologists, one important source of bias for the results reported in this and other studies is the cohort effect. Participants aged 80 years from a study cohort were born 80 years ago and participants aged 20 years were born 20 years ago. This cohort effect must be adjusted for to obtain unbiased estimates. One potential approach could be the age-period-cohort modeling method. For example, the newly established hierarchical age-period-cohort modeling method for sociology can be used to model individual level data (3). With this method, age-specific estimates of the total testosterone level can be estimated by adjusting for the cohort effect and the period effect if a researcher suspects that participants recruited from different periods will also be a confounding factor.
Authors: Xinguang Chen, Kai Wang
Institution: University of Florida
Language: English
Subfield: Endocrinology, Diabetes and Metabolism
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
Funder: University of Florida
OpenAlex ID: W2768468428