Endogenous sex steroid hormones and risk of liver cancer among US men: Results from the Liver Cancer Pooling Project

Zeni Wu, Jessica L. Petrick, Andrea A. Florio, Chantal Guillemette, Laura E. Beane Freeman, Julie E. Buring, Gary Bradwin, Patrick Caron, Yu Chen, A. Heather Eliassen, Lawrence S. Engel, Neal D. Freedman, J. Michael Gaziano, Edward L. Giovannuci, Jonathan N. Hofmann, Wen Yi Huang, Victoria A. Kirsh, Cari M. Kitahara, Jill Koshiol, I. Min LeeLinda M. Liao, Christina C. Newton, Julie R. Palmer, Mark P. Purdue, Thomas E. Rohan, Lynn Rosenberg, Howard D. Sesso, Rashmi Sinha, Meir J. Stampfer, Caroline Y. Um, Stephen K. Van Den Eeden, Kala Visvanathan, Jean Wactawski-Wende, Anne Zeleniuch-Jacquotte, Xuehong Zhang, Barry I. Graubard, Peter T. Campbell, Katherine A. McGlynn

Research output: Contribution to journalArticlepeer-review


Background & Aims: Incidence rates of liver cancer in most populations are two to three times higher among men than women. The higher rates among men have led to the suggestion that androgens are related to increased risk whereas oestrogens are related to decreased risk. This hypothesis was investigated in the present study via a nested case-control analysis of pre-diagnostic sex steroid hormone levels among men in five US cohorts. Methods: Concentrations of sex steroid hormones and sex hormone-binding globulin were quantitated using gas chromatography–mass spectrometry and a competitive electrochemiluminescence immunoassay, respectively. Multivariable conditional logistic regression was used to calculate odds ratios (ORs) and 95% CIs for associations between hormones and liver cancer among 275 men who subsequently developed liver cancer and 768 comparison men. Results: Higher concentrations of total testosterone (OR per one-unit increase in log2 = 1.77, 95% CI = 1.38–2.29), dihydrotestosterone (OR = 1.76, 95% CI = 1.21–2.57), oestrone (OR = 1.74, 95% CI = 1.08-2.79), total oestradiol (OR = 1.58, 95% CI=1.22–20.05), and sex hormone-binding globulin (OR = 1.63, 95% CI = 1.27–2.11) were associated with increased risk. Higher concentrations of dehydroepiandrosterone (DHEA), however, were associated with a 53% decreased risk (OR = 0.47, 95% CI = 0.33–0.68). Conclusions: Higher concentrations of both androgens (testosterone, dihydrotestosterone) and their aromatised oestrogenic metabolites (oestrone, oestradiol) were observed among men who subsequently developed liver cancer compared with men who did not. As DHEA is an adrenal precursor of both androgens and oestrogens, these results may suggest that a lower capacity to convert DHEA to androgens, and their subsequent conversion to oestrogens, confers a lower risk of liver cancer, whereas a greater capacity to convert DHEA confers a greater risk. Impact and implications: This study does not fully support the current hormone hypothesis as both androgen and oestrogen levels were associated with increased risk of liver cancer among men. The study also found that higher DHEA levels were associated with lower risk, thus suggesting the hypothesis that greater capacity to convert DHEA could be associated with increased liver cancer risk among men.

Original languageEnglish (US)
Article number100742
JournalJHEP Reports
Issue number7
StatePublished - Jul 2023


  • Androgen
  • Liver cancer
  • Male
  • Oestrogen
  • Sex steroid hormone

ASJC Scopus subject areas

  • Internal Medicine
  • Immunology and Allergy
  • Hepatology
  • Gastroenterology


Dive into the research topics of 'Endogenous sex steroid hormones and risk of liver cancer among US men: Results from the Liver Cancer Pooling Project'. Together they form a unique fingerprint.

Cite this