Measurement of Testosterone: How Important Is a Morning Blood Draw?

Presentation Number: THR-123
Date of Presentation: March 5th, 2015

E. David Crawford1, Wendy Poage2, Allen Nyhuis3, David Alan Price3, Sherie A Dowsett3 and David Muram*3
1University of Colorado, 2Prostate Conditions Education Council, 3Eli Lilly and Company



Serum testosterone levels exhibit a circadian variation, with peak levels in the morning. Evidence-based guidelines recommend that morning total testosterone (TT) levels be measured using a reliable assay as the initial diagnostic test for androgen deficiency (1). However, it is not known whether adherence to a morning blood draw is truly necessary. It has been argued that the circadian rhythm is diminished in older men such that TT measurement need not be limited to morning sampling, and patients who are seen later in the day can be sampled outside the recommended window, avoiding the need to return for a morning blood draw on a subsequent day. The purpose of this analysis was to determine whether it is appropriate to expand the 8-11 AM blood sampling window for the measurement of TT levels.


The study sample comprised men who elected to be screened during the September 2013 Prostate Cancer Awareness Week (PCAW), a US community-based prostate cancer screening program sponsored by the Prostate Conditions Educational Council (PCEC). Men undergoing screening provided informed consent, completed a series of health questionnaires, underwent a medical evaluation and, for a large subset, had blood drawn for laboratory assays, including measurement of TT levels. Analyses were performed using data from men with sufficient data for evaluation. TT levels measured in blood samples drawn from 8-11 AM (n=229) were compared with those measured during the rest of the day (n=442) (using the 2-sample t-test) as well as with those values from specific time windows (using ANOVA): 11 AM-2 PM (n=202), 2-5 PM (n=99), 5-8 PM (n=103), and 8 PM-8 AM (n=38). Analysis of variance methods were also used to adjust comparisons for age and BMI.


TT levels measured in blood drawn from 8-11 AM differed significantly from those drawn outside this time window (411.7 vs 368.3 ng/dl; p=0.0028). Differences in TT levels were evident across the 5 blood draw time windows (p=0.0006). These differences persisted after adjustment for age and BMI. TT levels in blood drawn from 2-5 PM (344.3 ng/dl) and 5-8 PM (334.4 ng/dl) differed significantly when compared directly with blood drawn from 8-11 AM (p<0.05), while TT levels in blood drawn from 11 AM-2 PM (396.5 ng/dl) and 8 PM-8 AM (373.4 ng/dl) did not (p>0.05).


In this analysis, TT levels in blood drawn from 11 AM-2 PM did not differ significantly from and, in fact, were numerically similar to TT levels in blood drawn from 8-11 AM. There may be clinical utility in expanding the blood draw window for TT measurement to early afternoon.


Disclosure: EDC: Consultant, Genomic Health, Consultant, Jansen Pharmaceuticals, Consultant, Dendreon, Consultant, Ferring Pharmaceuticals, Employee, Ferring Pharmaceuticals, Consultant, MDx, Consultant, Bayer, Inc.. WP: CME Programming, Independent Contractor, Bayer, Inc., Collaborator, Abbott Laboratories, Collaborator, Genomic Dx, Collaborator, Strand Diagnostics, Collaborator, MDx Health. AN: Employee, Eli Lilly & Company. DAP: Employee, Eli Lilly & Company. SAD: Employee, Eli Lilly & Company. DM: Employee, Eli Lilly & Company.