Objective: We investigated the risk of amnestic mild cognitive impairment (aMCI) in relation to serum bioavailable (BT) and total testosterone (TT) levels in older men.
Design, setting and subjects: A cross-sectional study in an ambulatory setting, with older men aged 55-93 years with normal cognition, aMCI and Alzheimer's disease (AD).
Measurements: Morning serum BT and TT levels were determined. AD was diagnosed by the Neurological and Communicative Disorders and Stroke/Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA) criteria for probable AD and aMCI by the Petersen criteria.
Results: We recruited 203 Chinese older men (48 aMCI, 66 AD and 89 with normal cognition). Mean serum BT, but not TT, levels were significantly lower in the aMCI (mean BT +/- SEM 1.06 +/- 0.10 nmol/l) and AD (0.99 +/- 0.08 nmol/l) groups than in the normal controls (1.82 +/- 0.12 nmol/l) (P < 0.001, one-way anova) with no significant difference between the aMCI and AD groups. After adjustment for education, age and apolipoprotein E (apoE) genotype, logistic regression analyses showed that the serum BT level [adjusted odds ratio (OR) = 0.52, 95% confidence interval (CI) 0.32-0.85] was an independent protective factor for aMCI. For the combined outcome of aMCI and AD, the serum BT level was an independent protective factor but age and apoE epsilon4 were independent risk factors. There was no interaction between BT and age.
Conclusions: In older men, serum BT, but not TT, levels were associated with a lower risk of aMCI and AD.