Bentmar Holgersson M, Landgren F, Rylander L, Lundberg Giwercman Y. Mortality Is Linked to Low Serum Testosterone Levels in Younger and Middle-aged Men. Eur Urol.
http://www.sciencedirect.com/science/article/pii/S0302283816308958
We collected data on all testosterone measurements conducted in men between November 1987 and December 1992 at the Department of Clinical Chemistry, Skane University Hospital Malmo, Sweden (n = 4278). Data were linked to the national Cancer Registry and the national Cause of Death Registry as of 2013.
Since no information was available regarding the indication for testosterone analysis, to exclude individuals on testosterone replacement and those seriously ill at the time of blood sampling, men with more than two testosterone measurements within 2 yr and those who died within 1 yr from measurement were excluded.
In addition, men aged <20 yr at measurement and those with a diagnosis of prostate cancer before measurement were excluded (n = 1199), leaving 3079 men (median age 47 yr, range 20–87 yr). The time of the earliest blood sample in our database was defined as the baseline.
Subjects were categorized as dead or alive on December 31, 2013 or at the time of emigration, whichever event occurred first. To take the age-related decline in testosterone levels into account, we calculated age-normalized z-scores and stratified the results by 10th and 90th percentiles.
The hazard ratio and 95% CI for mortality was calculated for men aged >/=50 yr and <50 yr at measurement using SPSS 22 software (SPSS, Chicago, IL, USA), and the 10–90th percentile group was used as a reference. The median follow-up length was 21 yr for those aged <50 yr at blood sampling, and 16 yr for the older group.
After >20 yr of follow-up, for men aged 20–49 yr at baseline with a testosterone level below the 10th percentile, the risk of all-cause mortality was almost doubled.
The risk of cardiovascular diseases (ICD9 codes 390–459 and ICD10 codes I00–I99; Fig. 1D) was more than doubled for these men. No significant association was found for men aged >50 yr at baseline.
In conclusion, men aged <50 yr presenting with low serum testosterone have significantly lower life expectancy. Our finding could indicate that testosterone deficiency per se is, at least in part, the underlying cause of the association between low testosterone and higher risk of mortality.
