Enzyme immunoassay may be inadequate for measuring salivary testosterone in older men.

Several correlations have been reported between men's testosterone (T) and other characteristics. Stalwart findings are a decline in T during the day, decline with obesity, and decline with advancing age. Here seven previously reported correlations are tested among older American men in the National Social Life, Health and Aging Project (NSHAP), their salivary T measured by enzyme immunoassays (EIA). Few significant correlations are found, with most tests producing weak or null results. These findings, overall, suggest that T does not "work" much among older men. However, a threat to this conclusion is raised by Welker et al. namely that EIA of salivary T may contain large errors, invalidating the tests. To check this possibility, these correlations from the literature were tested among older British men whose salivary T was measured by liquid chromatography-tandem mass spectroscopy (LC-MS/MS), a technique noted for its specificity. Not all could be tested, but the relationships of T with age, obesity, and health are significant among British men, indicating that T remains active in older men, and throwing doubt on the adequacy of EIA for measuring salivary T.