Skip to main content
. 2016 Feb 11;6:21306. doi: 10.1038/srep21306

Table 3. Effect of genetically predicted testosterone (nmol/L) on cognitive function among men (50+ years), Guangzhou Biobank Cohort Study, recruitment 2003–2008 and follow up till Dec 31, 2012.

Outcome Mean(SD) n Model Beta-coefficienta 95% CI P value
Delayed 10-word 5.5 4160 1 −0.02 −0.06–0.02 0.23
10-word recall (1.9) 4160 2 −0.03 −0.07–0.01 0.17
    4129 3 −0.02 −0.06–0.02 0.27
    4123 4 −0.02 −0.06–0.02 0.30
    4093 5 −0.02 −0.06–0.02 0.30
    3470 6 −0.02 −0.06–0.02 0.30
MMSE 27.5 4122 1 0.06 −0.002–0.12 0.06
  (2.7) 4122 2 0.05 −0.007–0.11 0.08
    4083 3 0.06 0.0003–0.12 0.05
    4078 4 0.06 −0.0004–0.12 0.05
    4048 5 0.06 −0.001–0.12 0.06
    3443 6 0.06 −0.01–0.13 0.10

Abbreviations: SD, standard deviation; CI, confidence interval.

Model 1 showed the association of genetically predicted testosterone with cognitive function without any covariates;

Model 2 adjusted for age;

Model 3 additionally adjusted for education, smoking, and use of alcohol;

Model 4 additionally adjusted for body mass index (BMI);

Model 5 additionally adjusted for Framingham score.

Model 6 replicated the analysis in Model 1 excluding men in another cognition project.

aBeta coefficient refers to the average change in cognitive function (the delayed 10-word recall score and MMSE score) with each unit (nmol/L) increase in genetically predicted testosterone.