The Joint Annual Scientific Meetings of the Endocrine Society of Australia and the Society for Reproductive Biology 2018

Plasma estradiol concentration is independently associated with longer leucocyte telomere length in older men (#99)

Bu B Yeap 1 2 , Jennie Hui 3 , Matthew W Knuiman 4 , David J Handelsman 5 , Leon Flicker 1 6 , Mark L Divitini 4 , Gillian M Arscott 3 , Susan V McLennan 7 , Stephen M Twigg 7 , Osvaldo P Almeida 1 6 , Graeme J Hankey 1 , Jonathan Golledge 8 , Paul E Norman 1 , John P Beilby 3
  1. The Medical School, University of Western Australia, Perth, WA, Australia
  2. Department of Endocrinology and Diabetes, Fiona Stanley Hospital, Perth, WA, Australia
  3. PathWest Laboratory Medicine, Sir Charles Gairdner Hospital, Perth, WA, Australia
  4. School of Population and Global Health, University of Western Australia, Perth, WA, Australia
  5. ANZAC Research Institute, University of Sydney, Sydney, NSW, Australia
  6. WA Centre for Health and Ageing, University of Western Australia, Perth, WA, Australia
  7. Department of Endocrinology, University of Sydney, Sydney, NSW, Australia
  8. Queensland Research Centre for Peripheral Vascular Disease, James Cook University, Townsville, Queensland, Australia


Telomeres protect chromosomes from damage, and shorter leucocyte telomere length (LTL) is a marker of advancing biological age. Men have shorter LTL compared with women, and experience a decline in circulating testosterone (T) with age. However the association between T and its bioactive metabolites, dihydrotestosterone (DHT) and estradiol (E2) with telomere length, particularly in older men, is uncertain.


To determine associations of sex hormones with leucocyte telomere length (LTL) in older men.

Participants and methods

2,913 community-dwelling men aged 70-84, mean (±SD) 76.7±3.2 years. Early morning blood samples were assayed for T, DHT, E2 using mass spectrometry, and luteinising hormone (LH) and sex hormone-binding globulin (SHBG) using immunoassay. LTL was measured using PCR in triplicate with median values used for analysis, expressed as the T/S ratio. Cross-sectional analyses utilised multivariable linear regression with adjustment for age, cardiometabolic risk factors and prevalent cardiovascular disease (CVD).


Average difference per decade of age was T -0.46 nmol/L, DHT -0.11 nmol/L, E2 -7.5 pmol/L, SHBG +10.2 nmol/L, LH +2.9 IU/L and LTL (T/S ratio) -0.065. E2 correlated with T/S ratio (r=0.038, p=0.039) and SHBG was inversely correlated (r=-0.053, p=0.004). After adjusting for age, cardiometabolic risk factors and prevalent CVD, E2 remained associated with T/S ratio (per 1SD increase in E2: coefficient 0.011, p=0.043). T, DHT and LH were not associated. When E2 and SHBG were simultaneously included in the model, E2 remained positively (coefficient 0.014, p=0.014) and SHBG inversely (coefficient -0.013, p=0.037) associated with T/S ratio.


In older men, neither T nor DHT are associated with LTL. E2 is independently associated with LTL and SHBG is inversely associated, implicating gonadal axis activity with biological age in this demographic group. Further research is needed to explore causality and the feasibility of interventions to slow biological ageing.