|
TESTOSTERONE AND MORTALITY IN OLDER MEN
Molly Shores
Veterans Affairs, Puget Sound Health Care System, Seattle, WA 910,
USA molly.shores@med.va.gov
Background: Due to the aging of our society, an increasing
proportion of the population is affected by age-associated low testosterone.
Low serum testosterone levels are associated with decreased muscle
mass and bone mineral density, the metabolic syndrome, type II diabetes,
and prevalent cardiovascular disease. In a small, preliminary study
in a Geriatric Rehabilitation unit, low testosterone was significantly
associated with six-month mortality.1 In a recent study, 2 in a
larger group of men (n=858), repeated low testosterone levels were
a significant risk factor for mortality in male veterans after controlling
for relevant covariates.
Study design: A computerized clinical database was used to
identify men >40 years with repeated testosterone levels from 1994-1999
and without diagnosed prostate cancer. Low testosterone was defined
as a total testosterone level <2.5 ng/ml or free testosterone <7.5
pg/ml. Men were classified as having low testosterone levels (n=166,
19.3%), equivocal testosterone levels (equal number of low and normal
levels, n=240 [28.0%]), or normal testosterone levels (n=452, 52.7%).
Cox proportional hazards regression models were used to estimate
risks for all-cause mortality, after adjusting for demographic and
clinical covariates over a follow-up period of up to 8 years.
Results: Mortality in men with normal testosterone was 20.1%
vs. 24.6% in men with equivocal testosterone vs. 34.9% in men with
low testosterone (p=.001). The increasing mortality rates found
in men with normal testosterone (20.1%) vs. equivocal (24.6%) vs.
low testosterone (34.9%) levels suggest a dose-response relationship.
After adjusting for age, medical morbidity, and other clinical covariates,
men with low testosterone levels had an increased mortality risk
with a Hazard Ratio (HR)=1.88, (95% Confidence Interval (CI), 1.34-2.63;
p<.001) compared to men with normal testosterone levels. Men with
equivocal testosterone levels had an increased mortality risk with
a HR=1.38, (95% CI, 0.99-1.92, p=.06) compared to men with normal
testosterone levels. In a sensitivity analysis, men who died within
the first year (n=50, 5.8%) were removed from the analysis, to minimize
the potential effect of acute illness on the results. After removing
men with early deaths, low testosterone continued to be associated
with an increased mortality risk (HR=1.68; 95% CI, 1.14-2.48; p=.009)
compared to normal testosterone. Limitations: The most significant
limitation of this study is the observational study design. Due
to the observational nature of the study, the possibility that residual
confounding occurred due to some other unmeasured factors cannot
be entirely excluded. Another limitation is that testosterone treatment
data were not available. However, it does not appear that this limitation
would invalidate the results, because testosterone treatment would
tend to bias the results towards the null, as it would minimize
the differences between men with low and normal testosterone levels.
Finally, the population is not representative of community-dwelling
men because it was a Veterans Affairs (VA) medical center-based
population. This limits the generalizability of the results because
veterans treated at VA medical centers tend to be of a lower socioeconomic
class and have greater medical morbidity.3 Also, the men in the
study had a very high overall mortality rate of 24.2% over a mean
follow-up period of 4.3 years. However, this is a mortality rate
consistent with other studies done of veterans treated at VA medical
centers with mortality rates ranging from 6-9%/yr.4-5 Thus, although
the mortality rate in this study was high, it is a mortality rate
that is comparable to that of other VA studies.
Conclusions: In adjusted analyses, men with low testosterone
levels had a significantly greater mortality risk (HR=1.88) than
men with normal testosterone levels. Men with equivocal testosterone
levels also had a greater mortality risk (HR=1.38) compared to men
with normal testosterone levels, although this did not reach statistical
significance (p=.06). In addition, after removing men from the study
who died within the first year, men with low testosterone continued
to have a 68% greater mortality risk compared to men with normal
testosterone levels. The persistence of elevated mortality risk
after excluding men who died within the first year, suggests that
the association between low testosterone and mortality is not simply
due to acute illness. Strengths of this study are the large sample
size, the ability to control for several covariates, and that the
classification of low testosterone was based on repeated low testosterone
levels and not solely upon a single measurement. Further prospective
studies are needed to examine the association between low-testosterone
and mortality in older men and to identify if there are particular
subgroups of hypogonadal men who are at greater risk for adverse
outcomes.
References
1. Shores MM, Moceri VM, Gruenewald DA, Brodkin KI, Matsumoto AM
and Kivlahan DR. Low testosterone is associated with decreased function
and increased mortality: A preliminary study among men in a geriatric
rehabilitation unit. J Am Geriatric Soc, 2004; 52(12):2077-81.
2. Shores MM, Matsumoto AM, Sloan KL, and Kivlahan DR. Low serum
testosterone levels and mortality in male veterans. Arch Intern
Med 2006;166:1660-1663.
3. Ahga Z, Lofgren RP, van Ruiswyk JV, Layde PM. Are patients at
Veterans Affairs Medical Centers sicker? A comparative analysis
of health status and medical resource use. Arch Intern Med. 2000;160:3252-3257.
4. Lavretsky H, Bastani R, Gould R, et al. Predictors of two-year
mortality in a prospective "UPBEAT" study of elderly veterans with
comorbid medical and psychiatric symptoms. Am J Geriatr Psychiatry.
2002;10:458-468.
5. Selim AJ, Kazis LE, Rogers W, et al. Risk-adjusted mortality
as an indicator of outcomes: Comparison of the Medicare Advantage
Program with the Veterans' Health Administration. Med Care. 2006:44:359-365.
|
