Oral Presentation ESA-SRB-NZSE-CaSR 2026 in conjunction with ENSA

Predicting circulating testosterone concentrations in individual men (143340)

Bu B Yeap 1 2 , Ross J Marriott 3 , Emma K Adams 3 , Leen Antonio 4 , Shalender Bhasin 5 , Adrian S Dobs 6 , Leon Flicker 1 , David J Handelsman 7 , Graeme J Hankey 1 , Alvin M Matsumoto 8 , Claes Ohlsson 9 , Eric S Orwoll 10 , Dirk Vanderschueren 4 , Robin Wilkening 11 , Gary A Wittert 12 , Frederick CW Wu 13 , Kevin Murray 14
  1. Medical School, University of Western Australia, Perth, WA, Australia
  2. Department of Endocrinology and Diabetes, Fiona Stanley Hospital, Perth, WA, Australia
  3. School of Population and Global Health, University of Western Australia, Perth, WA, Australia
  4. Laboratory of Clinical and Experimental Endocrinology, KU Leuven, Leuven, Belgium
  5. Brigham and Women's Hospital, Harvard Medical School, Boston, United States
  6. School of Medicine, Johns Hopkins University, Baltimore, United States
  7. ANZAC Research Institute, University of Sydney, Sydney, NSW, Australia
  8. Department of Medicine, University of Washington School of Medicine, Seattle, United States
  9. Centre for Bone and Arthritis Research at the Sahlgrenska Academy, University of Gothenburg, Goteborg, Sweden
  10. Oregon Health and Science University, Portland, United States
  11. School of Public Health and Preventive Medicine, Monash University, Melbourne, VIC, Australia
  12. School of Medicine, University of Adelaide, Adelaide, SA, Australia
  13. School of Medical Sciences, University of Manchester, Manchester, United Kingdom
  14. School of Population and Global Health, University of Western Australia, Perth, WA, Australia

Aims

Existing reference ranges for circulating testosterone concentrations in men are based on relatively small samples of healthy men, without considering variables such as age, height, weight, marital status, lifestyle behaviours, and medical comorbidities, all of which can influence testosterone concentrations. We aimed to develop a method of predicting total testosterone concentrations for individual men, taking these factors into account.

Methods

We devised a formula to calculate predicted total testosterone concentrations, based on meta-analyses of individual participant data from 21,074 men, aged 17-99 years, from nine cohort studies of community-dwelling men with testosterone measured using mass spectrometry (1). Inputs are a man’s age, height, weight, marital status, alcohol consumption, physical activity, smoking status and history of diabetes mellitus. The output is the predicted total testosterone concentration for that man, with estimated uncertainty shown as the prediction interval or PI (e.g. 80%PI).

Results

For a 30-year-old male, height 1.78 m, weight 75 kg, unmarried, with low alcohol consumption, high physical activity level, never smoked, no diabetes, predicted testosterone is 18.7 nmol/L, 80%PI 11.4-26.0 nmol/L. If aged 70 years, predicted testosterone is 18.4 nmol/L, 80%PI 11.2-25.6 nmol/L.

For the same 30-year-old male, except with weight 120 kg, predicted testosterone is 12.8 nmol/L, 80%PI 5.6-20.0 nmol/L. If with weight 120 kg, he was married, with high alcohol consumption, low physical activity level, ex-smoker, with diabetes, predicted testosterone is 9.8 nmol/L, 80%PI 2.6-17.0 nmol/L. If 70-years-old, with those unfavourable factors, predicted testosterone is 9.5 nmol/L, 80%PI 2.3-16.7 nmol/L.

Conclusion

Using our testosterone calculator, we demonstrate that on average, men with high BMI and other unfavourable sociodemographic and lifestyle factors, can expect to have testosterone concentrations <10 nmol/L in the absence of known pituitary or testicular problems. This could educate men, encouraging uptake of healthy lifestyle behaviours, and avoid potentially misleading diagnoses of “low” testosterone.

  1. 1. Marriott RJ, Murray K, Adams RJ, et al. Factors associated with circulating sex hormones in men: individual participant data meta-analyses. Ann Intern Med. 2023;176(9):1221-34.