Diurnal rhythms of serum total, free and bioavailable testosterone and of SHBG in middle-aged men compared with those in young men

Michael J. Diver, Komal E. Imtiaz, Aftab M. Ahmad, Jiten P. Vora, William D. Fraser

Research output: Contribution to journalArticle

268 Citations (Scopus)


BACKGROUND: Conflicting views are reported on the association between advancing age and gradually diminishing concentrations of serum total testosterone in men. The putative loss of diurnal rhythm in serum total testosterone in older men is reported to be in part due to low concentrations in the morning when compared to concentrations found in young men. We have measured total, free and bioavailable testosterone along with SHBG in samples taken every 30 min throughout a 24-h period in 10 young and eight middle-aged men.

RESULTS: Both young and middle-aged men displayed a significant diurnal rhythm in all variables, with a minimum fall of 43% in total testosterone from peak to nadir in all subjects. Subjecting the data to a time series analysis by least squares estimation revealed no significant difference in mesor (P = 0·306), amplitude (P = 0·061) or acrophase (P = 0·972) for total testosterone between the two groups. Comparing bioavailable testosterone in the two groups revealed no significant difference in mesor (P = 0·175) or acrophase (P = 0·978) but a significant difference (P = 0·031) in amplitude. Both groups display a significant circadian rhythm (middle-aged group P < 0·001; young group P = 0·014). Free testosterone revealed a highly significant rhythm in both the young group (P < 0·001) and the middle-aged group (P = 0·002), with no significant difference between the groups in mesor (P = 0·094) or acrophase (P = 0·698). Although analysis of the SHBG data revealed a significant rhythm in the young group (P = 0·003) and the older group (P < 0·001), the acrophase occurred in the mid afternoon in both groups (15·12 h in the young and 15·40 h in the middle-aged). The older men had a significantly greater amplitude (P = 0·044) but again no significant difference was seen in mesor (P = 0·083) or acrophase (P = 0·477) between the two groups. Acrophases for total, bioavailable and free testosterone occurred between 07·00 h and 07·30 h; for SHBG the acrophase occurred at 15·12 h in the young group and 15·40 h in the middle-aged group.

CONCLUSIONS: The study suggests that the diurnal rhythm in these indices of androgen status is maintained in fit, healthy men into the 7th decade of life.
Original languageEnglish
Pages (from-to)710-717
Number of pages8
JournalClinical Endocrinology
Issue number6
Publication statusPublished - Jun 2003

Cite this