Travison · The Journal of clinical endocrinology and metabolism 2017 · Cross-sectional cohort harmonization study · n=9054

Harmonized Reference Ranges for Circulating Testosterone Levels in Men of Four Cohort Studies in the United States and Europe.

Cited 325 times in the scientific literature.

Level 3 - non-randomized controlled study

Multi-cohort cross-sectional assay calibration and normative range study

PubMed 28324103 · doi:10.1210/jc.2016-2935 · record verified 2026-08-30

What was done

Samples from 9,054 community-dwelling men across four cohorts (Framingham Heart Study, European Male Aging Study, Osteoporotic Fractures in Men Study, and Male Sibling Study of Osteoporosis) were evaluated. Aliquots from 100 participants per cohort were re-measured at the CDC using a reference method. Generalized additive models, Bland-Altman analyses, and Passing-Bablok regression were used to construct normalizing equations and generate harmonized, age-specific reference percentiles.

What was found

Assay differences accounted for substantial baseline variation between cohorts. Harmonization reduced intercohort differences. In healthy nonobese men aged 19 to 39 years, the harmonized percentiles were: - 2.5th: 264 ng/dL - 5th: 303 ng/dL - 50th (median): 531 ng/dL - 95th: 852 ng/dL - 97.5th: 916 ng/dL The resulting 95% reference interval (2.5th to 97.5th percentile) for healthy young nonobese men is 264 to 916 ng/dL. Testosterone concentrations were consistently higher in nonobese men than in the unselected cohort population across age groups.

Why it matters

Standardizes diagnostic thresholds for male hypogonadism across clinical and research laboratories by providing universal, CDC-calibrated reference intervals for total testosterone.

Limits

Normalizing equations were derived from re-testing 100 men per cohort rather than directly measuring all 9,054 participants with the CDC reference method. The study sample was restricted to European and American cohorts, limiting generalizability to other racial and ethnic groups. Clinical correlation with symptomatic hypogonadism was not evaluated in the abstract.

Cited by