This experiment examined the effects of delays in separation and freezing of whole blood components on analytes of interest in studies of
prostate cancer prevention, in order to evaluate the feasibility of centralized processing of blood for the multisite
Selenium and
Vitamin E Cancer Prevention Trial. Blood from 40 healthy men was subjected to four treatment protocols, allowing the contrast of immediate processing to delays of 32, 72, and 144 hours. At 32 hours, simulating refrigerated storage and overnight shipping, there was a 2.9% decrease (95% confidence interval, 0.7-5.1) in
insulin-like growth factor-I (
IGF-I) but no significant change in
carotenoids,
tocopherols,
testosterone, 3alpha-androstanediol
glucuronide (AAG),
sex hormone-binding globulin (SHBG) or
insulin-like growth factor binding protein 3 (IGFBP3). A 144-hour processing delay, simulating weekend blood collection or shipping delay, resulted in significant changes in
gamma-tocopherol (-1.5%),
IGF-I (-5.7%), IGFBP3 (-2.9%), SHBG (-4.0%),
testosterone (+4.7%), and AAG (+5.5%). The rank-order and intraclass correlations between analytes from blood processed immediately and those subjected to delayed processing were 0.96 or higher for
carotenoids,
tocopherols, AAG, and SHBG, and between 0.87 and 0.95 for
IGF-I, IGFBP3, and
testosterone. A 32-hour delay decreased lymphocyte viability from 82.5% to 75.0% (P = 0.45), but a 72-hour delay decreased viability to 36.8% (P < 0.001). Overnight shipping and centralized processing is an acceptable approach to blood collection in large multisite trials examining the
cancer-related measures proposed in the
Selenium and
Vitamin E Cancer Prevention Trial. Longer processing delays, however, have small but statistically significant effects on many analytes and substantially decrease lymphocyte viability.