Download Testosterone, Dehydroepiandrosterone, and Physical Performance

0021-972X/06/$15.00/0
Printed in U.S.A.
The Journal of Clinical Endocrinology & Metabolism 91(2):425– 431
Copyright © 2006 by The Endocrine Society
doi: 10.1210/jc.2005-1227
Testosterone, Dehydroepiandrosterone, and Physical
Performance in Older Men: Results from the
Massachusetts Male Aging Study
Amy B. O’Donnell, Thomas G. Travison, Susan S. Harris, J. Lisa Tenover, and John B. McKinlay
New England Research Institutes (A.B.O., T.G.T., J.B.M.), Watertown, Massachusetts 02472; Jean Mayer U.S. Department
of Agriculture Human Nutrition Research Center on Aging, Tufts University (S.S.H.), Boston, Massachusetts 02111; and
Emory University School of Medicine (J.L.T.), Atlanta, Georgia 30329
Objective: This manuscript examines the relationships of total testosterone (T), bioavailable T, dehydroepiandrosterone (DHEA), and
DHEA sulfate (DHEAS) to measures of physical performance in a
large, population-based, random sample of men.
Results: All hormones exhibited significant age-adjusted positive
association with PPT score below, but not necessarily above, the
thresholds. DHEA was positively associated with chair stand score
below, but was not above the threshold. None of the hormones studied
was significantly associated with grip strength.
Methods: In the most recent wave of the Massachusetts Male Aging
Study, measures of strength and physical performance [seven-item
physical performance test (PPT), timed chair stand test, and grip
strength] were made in 684 men, aged 55– 85 yr. Complete hormone
data were also obtained. Initial graphical exploration of performance
outcomes as a function of hormone levels showed linear increases in
physical performance up to certain threshold hormone concentrations, beyond which the associations were diminished. Regression
models were used to estimate threshold locations and standardized
regression coefficients quantifying the association between hormones
and strength.
Conclusion: Up to certain critical concentrations, elevated levels of
TT, total T, bioavailable T, DHEA, and DHEA sulfate are associated
with increased physical performance, as indicated by the PPT. However, levels beyond those critical concentrations, as might be achieved
through exogenous supplementation, do not appear to confer any
additional benefit. In general, hormone concentrations do not appear
to be meaningfully associated with grip strength or chair stand scores.
(J Clin Endocrinol Metab 91: 425– 431, 2006)
B
a randomized, controlled study of T supplementation in 61
hypogonadal, HIV-infected men, aged 18 –50 yr. They found
that T supplementation for 16 wk resulted in increases in
muscle volume and strength, whether accompanied by resistance training or not. Similarly, Wang et al. (12) found
positive effects over 90 d of a transdermal T gel on muscle
strength and lean mass in 227 hypogonadal men, aged 19 – 68
yr. Bakhshi et al. (13) demonstrated increases in grip strength
and functional independence in 15 ill older men given T
injections for up to 8 wk, and Page et al. (14), in a randomized
controlled study of 70 older men (ages 65– 83 yr), reported
improvement in muscle mass, grip strength, and timed physical performance tests (PPTs) in the men given T injections
for up to 3 yr.
By contrast, other studies have demonstrated modest or inconsistent effects of T supplementation on strength and function in
hypogonadal men (15–18). For example, Snyder et al. (17) found a
positive effect of T supplementation on lean mass, but not on
muscle strength, in 108 hypogonadal older men. Ly et al. (18) found
only limited improvement in lower limb muscle strength and no
change in physical function in 37 hypogonadal older men given T.
Bhasin et al. (19) demonstrated a positive effect of supraphysiological doses of T on muscle size and strength in 43 normal young
men. Bhasin et al. (20) later showed similar effects in older men,
albeit with elevated risk of adverse events, and reasonable effective
gain in muscle strength and fat-free mass with supplementation
consistent with high-normal T concentrations.
Few studies have examined the relationship of DHEA to
ECAUSE THE POPULATION of the United States is
aging rapidly, the prevalence of problems associated
with aging is also increasing. One of these is declining physical performance. Specific aspects of physical health, including mobility, strength, endurance, and coordination, are
known to decline with age (1, 2). These declines can lead to
a loss of independence (3) and an increase in medical difficulties, ranging from depression (4, 5) to obesity (6) to falls
and fractures (3, 7) to poor overall health (5, 8). Testosterone
(T), a steroid hormone required for the acquisition of lean
muscle mass, and dehydroepiandrosterone (DHEA), a weak
androgenic steroid that may increase muscle strength and
lean body mass, also decline with age, but the extent to which
these hormonal changes explain physical performance declines in older adults is unknown. T and DHEA are interrelated, because DHEA can be converted to androstenedione
or androstenediol and then, ultimately, to T.
A number of studies have demonstrated that administering doses of T to hypogonadal men can improve lean body
mass and muscle strength (9 –14). Bhasin et al. (11) conducted
First Published Online December 6, 2005
Abbreviations: BMI, Body mass index; BT, bioavailable testosterone;
DHEA, dehydroepiandrosterone; DHEAS, DHEA sulfate; MMAS, Massachusetts Male Aging Study; PPT, physical performance test; T, testosterone; TT, total T.
JCEM is published monthly by The Endocrine Society (http://www.
endo-society.org), the foremost professional society serving the endocrine community.
425
426
J Clin Endocrinol Metab, February 2006, 91(2):425– 431
O’Donnell et al. • Androgens and Physical Performance in Older Men
physical performance, muscle, and strength. Although Nestler et al. (21) demonstrated an increase in muscle mass in
healthy young men with DHEA therapy, these findings were
not upheld in later work on obese (22) and nonobese (23)
young men. Brown et al. (24) gave varying DHEA doses to
30 young men and found no difference in muscle strength
between the groups during resistance training.
The Massachusetts Male Aging Study (MMAS) is the first
study to examine associations of T and DHEA with a range
of physical performance measures in a large, populationbased, random sample of men.
Subjects and Methods
Study sample and in-home visit
The MMAS is an observational cohort study of health in aging men.
The design and field protocol have been described previously (25).
Briefly, communities in the Boston, Massachusetts, area were randomly
selected with probabilities proportional to the population within each of
six strata defined by community size and median income. Men born
between 1917 and 1946 were drawn at random from the Massachusetts
annual state census list, with sampling fractions adjusted to produce a
uniform distribution between ages 40 and 70 yr. A total of 1709 respondents completed a baseline visit between 1987 and 1989. Follow-up visits
were conducted after approximately 9 yr (1995–1997) and 15 yr (2002–
2004; Fig. 1). Measures of physical strength and performance were newly
introduced at the time of the second follow-up.
Of the 855 men who completed the second follow-up interview, 79
were interviewed by telephone because they had moved out of state, and
one subject began, but did not complete, the interview. These subjects’
physical performance measurements are therefore unavailable. Of the
remaining 775 men, 91 did not have valid hormone measures (they
refused the blood draw, did not have a blood draw due to health risks,
or were currently being treated with supplemental androgens or androgen suppressor medications). The remaining 684 subjects are considered in this analysis.
The MMAS field protocol involved a trained field technician/phlebotomist visiting each subject in his home, obtaining written informed
consent, and drawing two nonfasting blood samples. After administering an extensive set of questions concerning health status and behavior,
the interviewer conducted a series of physical performance measures,
including chair stands, grip strength, and a performance test.
Measure of T
Research design (e.g. time of day of blood sampling and method of
analysis) and the study sample composition (e.g. whether it is made up
of patients or volunteers) may impact T levels (25); accordingly, the
design of the MMAS accounted for potential biases in measuring hormone levels. Nonfasting blood samples were drawn within 4 h of the
subject’s awakening to control for diurnal variations in hormone levels.
Two samples were drawn 30 min apart and pooled for analysis in equal
aliquots to dampen the influence of episodic secretion. Blood was kept
in an ice-cooled container for transport and was centrifuged within 6 h.
Serum was stored in 5-ml scintillation vials at ⫺70 C until the time of
assay.
All hormone assays were performed in Dr. Christopher Longcope’s
laboratory at the University of Massachusetts Medical School (Worcester, MA). Total T (TT) was measured by RIA kit (Diagnostic Products
Corp., Los Angeles, CA). The intraassay coefficient of variation was
3.6%, and the interassay coefficient of variation was 8.3%. Bioavailable
T (BT; T not bound to SHBG) was calculated as BT ⫽ TT ⫻ (% free T ⫹
percent albumin-bound T)/100. DHEA and DHEAS were also measured
by RIA, with intraassay coefficients of variation of 2.1% and 4.2%, and
interassay coefficients of variation of 10.3% and 6.4%, respectively. Full
details regarding hormone measurements have been reported previously (26). As we have noted previously (25), the distribution of MMAS
serum total T values is comparable to those reported by other major
epidemiologic studies using similar RIA techniques and exhibits a similar rate of decline with age.
Measures of physical strength and function
The PPT was used to assess physical performance on functional
tasks (27) for 684 men. The PPT has been shown to be reliable (Cronbach’s ␣ ⫽ 0.79; interrater reliability ⫽ 0.93) (27) and has demonstrated concurrent validity with self-reported measures of physical
function. We administered the seven-item version that included writing a sentence, scooping a bean from a container (simulated eating),
putting on a jacket (simulated dressing), picking up a penny from the
floor, placing a book on a high shelf, turning 360°, and walking 50 ft.
The PPT is scored by awarding between 0 and 4 points for each of the
seven tasks, yielding a range of 0 –28 for the total score. If the respondent was unable to perform a specific task, 0 points were given.
Otherwise, the score was based on the time taken to do the task, with
4 points for the fastest performance and 1 point for the slowest. For
the 360° turn, points were awarded based on qualitative features of
the turn: 2 points for continuous rather than discontinuous steps, and
2 points for steady rather than unsteady movement, giving possible
scores of 0, 2, or 4 points. The PPT test has been previously used in
other large epidemiological studies for community-dwelling and institutionalized older adults, demonstrating construct validity and
internal consistency (28).
Grip strength was measured using the Jamar hydraulic hand dynamometer (Sammons Preston, Bolingbrook, IL), which measures isometric grip force from 0 –200 lb. A peak-hold needle automatically retains
the highest reading until reset. The instrument can be adjusted to five
grip positions to accommodate different hand sizes. Reliability has been
assessed using the Pearson product-moment correlation and a one-way
ANOVA for repeated measures, such that all correlations were significant (P ⬍ 0.01), and the median and modal correlations were all 0.97 or
0.98 (29). For the grip strength measurement, respondents were seated
comfortably in straight back chairs with their arms at their sides and the
elbows of their dominant hands flexed at a 90° angle. The forearm and
wrist were in neutral alignment. Each respondent was asked to squeeze
the dynamometer for 3 sec once, rest 1 min, and then squeeze again for
3 sec. The higher of the two measurements was used in the analyses.
Three people did not complete the screening, 53 had pain/arthritis in
their hands, one had had surgery on the dominant hand, and four did
not believe they could do the grip strength test safely, leaving grip
strength measurements for 623 participants.
The repeated chair stand test, used to measure lower extremity
strength, was conducted according to a standard protocol (30). Participants were asked whether they thought it was safe to make five
attempts and, if they agreed, were asked to stand up straight as
quickly as they could and then sit back down without stopping in
between. Points were awarded based on the time needed to complete
the chair stands. The chair stand task was added to the research
protocol after 25 men had been interviewed, yielding 659 chair stand
scores.
FIG. 1. MMAS design.
O’Donnell et al. • Androgens and Physical Performance in Older Men
Statistical analysis
Scatterplots and locally weighted regression smoothers were used to
explore the relationship between hormone levels and performance. In
several specific cases, exploratory data analysis elicited a so-called
threshold effect as a feature of the association between hormone levels
and measures of physical function; that is, physical performance
was positively correlated with hormone levels up to certain concentrations, beyond which mean physical performance was relatively
stable or might, in fact, appear to decrease with increasing hormone
concentrations.
Because of this pattern, subsequent analyses conceptualized the presence of the threshold effects as nonlinearities in the relationship between
hormone levels and measures of physical performance. For each model,
the location of the nonlinearity was chosen from those consistent with
the exploratory graphical analysis in such a way as to minimize the
residual sum of squares of a regression of physical performance on
hormone concentration when controlling for age effects.
The resulting models allow the estimated slope of mean performance on hormone concentration below the threshold to differ from
that above. In many cases these differences appeared to be both
statistically and clinically significant. For all models, age and age
squared were included as covariates. Height was also included for the
grip strength model, because height has been shown to have a positive correlation to grip strength (31). Nonparametric rank-sum tests
were employed to compare physical performance among men with
hormone levels below the relevant threshold to that among men with
hormone levels above that threshold. These comparisons were supplemented with comparisons based upon a linear regression analysis
controlling for age and body mass index (BMI). Results were considered statistically significant if the null hypothesis of no difference
could be rejected at the 0.05 level.
Results
Table 1 presents the characteristics of the full analysis
sample (n ⫽ 684; mean ⫾ sd). The PPT score was 22.8 ⫾ 4.3
of a possible 28 points, indicating that our population possessed a level of performance comparable to other epidemiological studies of similar populations (32). The mean age of
the subjects considered in this analysis was 68 ⫾ 8.0 yr at the
time PPT measurements were made. Two hundred eightysix subjects (42%) were between the ages of 55 and 64 yr, 238
subjects (35%) were between the ages of 65 and 74 yr, and 160
subjects (23%) were between the ages of 75 and 85 yr.
As described above, threshold values were determined for
the association between each hormone and performance outTABLE 1. Sample characteristics (n ⫽ 684), Massachusetts Male
Aging Study, Third Wave 2002–2004
Variable
Mean (SD) or %
Age (yr)
Caucasian
Married
At least some college
Nonsmokers
Height (cm)
Weight (kg)
BMI
TT (ng/dl)
BT (ng/dl)
DHEA (ng/dl)
DHEAS (␮g/dl)
Chair stand test score (sec)
PPT score (points)
Grip strength (kg)
68.0 (8.0)
98
73
77
88
174.8 (7.1)
86.0 (15.5)
28.1 (4.7)
411.7 (164.8)
132.9 (52.2)
308.9 (177.4)
123.3 (125.1)
3.4 (1.2)
22.8 (4.3)
37.7 (8.8)
Nanograms per deciliter can be converted to nmol/liter by multiplying by 0.03467; ␮g/dl can be converted to nmol/liter by multiplying
by 34.67.
J Clin Endocrinol Metab, February 2006, 91(2):425– 431
427
come (Table 2). In most cases threshold values were between
the median and the 75th percentile of the hormone distribution (to accommodate distributional skew, log-transformed DHEA and DHEAS were considered, with little substantive impact on results; for ease of presentation, we focus
on untransformed hormone values here).
The standardized regression coefficients above and below
the estimated threshold for each of the hormones are listed
in Table 2. These coefficients may be interpreted as standardized slopes. For example, in Table 2 we note that the
model-estimated slope of PPT on BT below the relevant
threshold concentration, 141 ng/dl (4.89 nmol/liter), is 0.224
and is highly statistically significant (P ⬍ 0.001). This indicates that, on the average, a typical subject with a BT less than
or equal to 141 ng/dl would score 0.224 points higher on the
PPT than would another subject of the same age whose BT
concentration was 1 sd, roughly 52 ng/dl (1.80 nmol/liter),
lower.
Below their respective thresholds, both TT and BT were
positively associated with age-adjusted scores on the PPT.
DHEA and DHEAS measures followed the same trend below
their respective thresholds, but were inversely associated
with the PPT at a statistically significant level above their
thresholds. Neither TT nor BT was positively associated with
age-adjusted scores on the timed chair stand test (controlling
for age) or with grip strength (controlling for age and height).
Figure 2 depicts the predicted PPT score over a range of
possible TT values for 68-yr-old, 28 kg/m2 men (these are the
mean age and BMI for the analysis sample, respectively). A
typical such man with a total T concentration of 300 ng/dl
(10.40 nmol/liter) would be expected to have a PPT score of
approximately 23.0.
In terms of statistical significance, the strongest positive
association between hormones and physical measures when
controlling for age was that between DHEAS and a PPT score
below the DHEAS threshold of 97 ␮g/dl (2614 nmol/liter).
The only significant negative associations appear above the
threshold for DHEA and DHEAS with the PPT score and
above the threshold for DHEAS with chair stand test score.
DHEAS followed a similar pattern as DHEA below the
thresholds, showing a significant positive association with
the PPT score and no significant association with grip
strength. Only DHEA was significantly associated with chair
stand score, and no hormone concentrations were associated
with grip strength.
Components of PPT
Analyses revealed that hormone concentrations appear
to be more strongly associated with the PPT than with the
chair stand score or grip strength. When we examined the
relationship between T and DHEA/DHEAS with individual components of the total PPT, we observed similar
relationships to those observed with the overall PPT.
Can the association between physical performance and
hormone levels be summarized using the thresholds only?
We sought to determine whether the observed differences
persisted when subjects with hormone concentrations above
the thresholds were compared, as a group, to those subjects
whose concentrations were below the thresholds. Table 3
428
J Clin Endocrinol Metab, February 2006, 91(2):425– 431
O’Donnell et al. • Androgens and Physical Performance in Older Men
TABLE 2. Age-adjusted associations between hormones and PPT scores (n ⫽ 684), chair stand test scores (n ⫽ 659), and grip strength
scores (n ⫽ 623), Massachusetts Male Aging Study, Third Wave 2002–2004
PPT
TT
BT
DHEA
DHEAS
Chair stand test scores
TT
BT
DHEA
DHEAS
Grip strength
TT
BT
DHEA
DHEAS
Median
Threshold (SI)
Standardized regression
coefficient for hormone levels
below threshold (P values)a
Standardized regression
coefficient for hormone levels
above threshold (P values)a
390 ng/dl
128 ng/dl
280 ng/dl
80 ␮g/dl
451 ng/dl
141 ng/dl
464 ng/dl
97 ␮g/dl
0.141 (0.037)
0.224 (0.001)
0.222 (0.000)
0.684 (0.000)
0.023 (0.742)
⫺0.050 (0.432)
⫺0.280 (0.004)
⫺0.109 (0.013)
390 ng/dl
128 ng/dl
280 ng/dl
80 ␮g/dl
537
140
551
117
ng/dl
ng/dl
ng/dl
␮g/dl
0.086 (0.129)
0.118 (0.115)
0.164 (0.001)
0.288 (0.082)
0.024 (0.812)
⫺0.033 (0.629)
⫺0.057 (0.688)
⫺0.105 (0.036)
390 ng/dl
128 ng/dl
280 ng/dl
80 ␮g/dl
433
142
361
266
ng/dl
ng/dl
ng/dl
␮g/dl
0.054 (0.450)
0.126 (0.066)
0.103 (0.150)
0.121 (0.089)
⫺0.033 (0.611)
⫺0.031 (0.621)
⫺0.076 (0.275)
0.067 (0.374)
Nanograms per deciliter can be converted to nmol/liter by multiplying by 0.03467; ␮g/dl can be converted to nmol/liter by multiplying by
26.9485.
a
Obtained from a linear spline regression of PPT score on hormone, height, age and age-squared. Standardized regression coefficient indicates
the PPT increase, in units of SD, associated with 1 SD increase in hormone.
presents the mean time to perform PPT tasks for subjects
whose hormone concentrations were above the thresholds
for TT and DHEAS vs. those whose concentrations were not.
Subjects whose hormone levels were above thresholds consistently performed all tasks more rapidly than those whose
levels were below the thresholds, but in most cases the differences were small and did not rise to the level of statistical
significance. Whether TT levels were above or below the
relevant threshold was independently associated with time
taken to walk 50 ft, to simulate eating, and to picking up a
penny. Being above vs. below the DHEAS threshold was
independently associated with writing a sentence, simulated
eating, and walking 50 ft.
Age was more strongly associated with DHEAS than
with TT, whereas BMI was more strongly associated with
FIG. 2. Predicted PPT score as a function of TT for 68yr-old men.
TT than with DHEAS. The mean age (⫾sd) of subjects
whose TT levels were below the threshold 451 ng/dl was
68.01 (⫾8.19) yr, whereas the mean age of those whose TT
levels were above the threshold was 68.05 (⫾7.80) yr. By
contrast, the mean (⫾sd) age of subjects with DHEAS
below the threshold of 97 ␮g/dl was 69.73 (⫾8.05) yr,
whereas that of subjects whose DHEAS values were higher
were younger, on the average, was 65.50 (⫾7.34) yr. In
regression analyses with age and BMI effects controlled,
being above the TT threshold remained associated with
decreased time to pick up a penny and walk 50 ft, but
differences with respect to the DHEAS threshold were no
longer statistically significant.
Turning in a circle was omitted from this analysis, because
it was not a timed task.
O’Donnell et al. • Androgens and Physical Performance in Older Men
J Clin Endocrinol Metab, February 2006, 91(2):425– 431
429
TABLE 3. Time to complete physical performance tasks (n ⫽ 684), by hormone level
TT
Task
Write a sentence
Simulate eating
Walk 50 fte
Put on jacket
Pick up pennye
Lift book
Below 451 ng/dla
DHEAS
At or above 451 ng/dlb
Below 97 ␮g/dlc
At or above 97 ␮g/dld
Mean time (sec)
SD
Mean time (sec)
SD
Mean time (sec)
SD
Mean time (sec)
SD
12.63
12.46
16.77
13.43
2.85
2.14
4.62
3.32
6.19
5.48
1.59
1.64
12.18
12.00
15.47
12.90
2.53
2.13
4.22
3.09
4.13
4.49
1.38
1.43
12.97
12.54
16.81
13.24
2.83
2.19
4.82
3.49
4.90
5.32
1.63
1.74
11.72
11.92
15.54
13.22
2.53
2.06
3.82
2.80
4.91
4.88
1.36
1.26
Nanograms per deciliter can be converted to nmol/liter by multiplying by 0.03467; ␮g/dl can be converted to nmol/liter by multiplying by
26.9485.
a
n ⫽ 433 (63%); mean (SD) age (yr), 68.0 (8.2); mean (SD) BMI, 28.9 (5.0).
b
n ⫽ 251 (37%); mean (SD) age (yr), 68.0 (7.8); mean (SD) BMI, 26.7 (3.8).
c
n ⫽ 409 (60%); mean (SD) age (yr), 69.7 (8.1); mean (SD) BMI, 28.1 (4.5).
d
n ⫽ 275 (40%); mean (SD) age (yr), 65.5 (7.3); mean (SD) BMI, 28.1 (4.9).
e
Difference between subjects below and above thresholds, with age and BMI effects controlled, is statistically significant at the 0.05 level
(result obtained via linear regression analysis).
Discussion
We found positive associations in older men between hormone concentrations and physical performance as measured
by the PPT, although there were minimal, if any, significant
relationships between these hormones and physical performance as measured by chair stands or grip strength. Positive
associations between hormone concentrations and PPT performance were present only with hormone levels below certain threshold values. These threshold values are above the
median in all cases. The TT, BT, and DHEAS thresholds fall
in the mid-high range, whereas the DHEA threshold is in the
high range.
We examined the possibility that the apparent nonlinearities in the associations we observed were the result of outcome ceiling effects, which might occur if, for instance, a
large percentage of the dependent variable values were clustered at the high end of the scale. This was not the case; for
instance, only 4% of the PPT scores were 28, the highest
possible value for that scale.
We observed differences in the linear relationship between
performance and hormone levels below relevant thresholds
vs. above. However, in an additional comparison of the collapsed group of subjects whose levels were above thresholds
to those below (without regard to their hormone level, except
in relation to the threshold cutoff), differences in time to
perform specific components of the PPT were not usually
significant when age effects were controlled, even though
subjects with higher hormone levels completed every task
more rapidly than their counterparts below the thresholds.
The crude comparison of subjects above vs. below the threshold appears to diminish the strength of the relationships
observed when the full spectrum of hormone values are
employed in analyses; generally speaking, the differences
observed do not persist when the subject groups above and
below the thresholds are collapsed, and the threshold itself
does not appear sufficient to account for the association
between androgens and performance.
In particular, although the relationship between PPT
scores and DHEAS levels was similar to that between PPT
scores and T concentrations, we observed that when age
effects were controlled, the crude differences between
DHEAS groups (⬎97 ␮g/dl vs. below) in time taken to
perform specific tasks were no longer statistically significant. The fact that in our sample those subjects with low
DHEAS levels are somewhat older than those whose
DHEAS levels are high suggests that the small, but consistent, differences between collapsed groups may be explained by age, and that the relationship between physical
performance and DHEAS must be viewed across the spectrum of possible concentrations, rather than as a simple
difference between those subjects below some threshold to
those above.
One of the major advantages of this study is that we were
able to measure overall physical performance, whereas much
of the earlier work has focused solely on strength. The PPT
(measuring physical performance) coupled with the grip
strength (measuring upper extremity strength) and chair
stands (measuring balance and lower extremity strength)
presents a more complete picture of performance and ability
than strength alone and may be more closely related to activities of everyday functioning and daily living, although
tests such as the repeated chair stand may be somewhat
ineffectual in discriminating among relatively healthy subjects. A second advantage of this study is that it was conducted in a large, community-based, random sample, rather
than a clinical-based population, thus making our results
more applicable to aging men in general.
Our results suggesting that beyond a certain threshold
there may be little association between androgens and physical performance stand somewhat in contrast to those of
Bhasin et al. (20), who found substantial gains in muscle mass
and strength among older men at high-normal and supraphysiological doses of T and stated that the “best trade-off”
between increased function and likelihood of adverse events
is obtained in the high-normal range. Because these researchers employed the leg press as a general measure of muscle
strength, with graduated resistance and weight-dependent
repetition of the exercise, and measure the per time clinical
effect of an induced change in T levels (whereas we measured
a cross-sectional association between T and strength in an
observational setting), some variance with respect to our
results is to be expected.
430
J Clin Endocrinol Metab, February 2006, 91(2):425– 431
Conversely, our results parallel those reported by Synder
et al. (33), who demonstrated that the effectiveness of T treatment for increasing bone mineral density is inversely related
to the level of pretreatment serum T.
An implication of our findings may be that men whose
T and DHEA levels are high are likely to receive little
benefit in physical performance from increased levels of
either hormone. In addition, they may incur unnecessary
risk. In a recent review, Tan and Culberson (34) list the
following as potential risks for using T replacement therapy: increase in hematocrit level, increase in sleep apnea,
sterility, erythrocytosis, edema, gynecomastia, and prostate stimulation. Others have suggested that potential
risks for T replacement therapy include increased prostate-specific antigen levels (35) and overdosage for lifestyle reasons (36). Thus, we concur with others who have
suggested that T replacement in men should be approached with caution (37). This caution should be applied
to DHEA and its sulfate form as well. Although DHEA is
currently taken by some in an effort to restore youth and
vitality, there is very little research to support an antiaging
effect (38, 39); in fact, the effects of oral administration of
DHEA on blood T levels are uncertain (21, 40). When
administering T or DHEA to improve physical performance, physicians should take the patient’s baseline hormone levels into account. If they are in the mid- to highnormal range, additional hormone supplementation may
be useless and even potentially harmful.
O’Donnell et al. • Androgens and Physical Performance in Older Men
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Acknowledgments
We thank Sharon Tennstedt for helpful discussions, and William B.
Simpson for statistical assistance.
21.
22.
Received June 2, 2005. Accepted November 23, 2005.
Address all correspondence and requests for reprints to: Dr. John B.
McKinlay, New England Research Institutes, 9 Galen Street, Watertown,
Massachusetts 02472. E-mail: [email protected].
This work was supported by the National Institute of Diabetes and
Digestive and Kidney Disorders (Grant DK-44995).
References
1. Wolfson L 2001 Gait and balance dysfunction: a model of the interaction of age
and disease. Neuroscientist 7:178 –183
2. Brooks S, Faulkner JA 1994 Skeletal muscle weakness in old age: underlying
mechanisms. Med Sci Sports Exerc 26:432– 439
3. Brach J, VanSwearingen JM 2002 Physical impairment and disability: relationship to performance of activities of daily living in community-dwelling
older men. Phys Ther 82:752–761
4. Penninx B, Deeg DJ, van Eijk JT, Beekman AT, Guralnik JM 2000 Changes
in depression and physical decline in older adults: a longitudinal perspective.
J Affect Disord 61:1–12
5. Penninx BW, Guralnik JM, Ferrucci L, Simonsick EM, Deeg DJ, Wallace RB
1998 Depressive symptoms and physical decline in community-dwelling older
persons. JAMA 279:1720 –1726
6. Cress ME, Buchner DM, Questad KA, Esselman PC, deLateur BJ, Schwartz
RS 1999 Exercise: effects on physical functional performance in independent
older adults. J Gerontol A Biol Sci Med Sci 54:M242–M248
7. Edelberg HK 2001 Falls and function. How to prevent falls and injuries in
patients with impaired mobility. Geriatrics 56:41– 45, 49
8. Wei JY, Li Y, Lincoln T, Grossman W, Mendelowitz D 1989 Chronic
exercise training protects aged cardiac muscle against hypoxia. J Clin Invest
83:778 –784
9. Bhasin S, Storer TW, Berman N, Yarasheski KE, Clevenger B, Phillips J, Lee
WP, Bunnell TJ, Casaburi R 1997 Testosterone replacement increases fat-free
mass and muscle size in hypogonadal men. J Clin Endocrinol Metab 82:407–
413
10. Brodsky IG, Balagopal P, Nair KS 1996 Effects of testosterone replacement on
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
muscle mass and muscle protein synthesis in hypogonadal men–a clinical
research center study. J Clin Endocrinol Metab 81:3469 –3475
Bhasin S, Storer T, Javanbakht M, Berman N, Yarasheski KE, Phillips J, Dike
M, Sinha-Hikim I, Shen R, Hays R, Beall G 2000 Testosterone replacement
and resistance exercise in HIV-infected men with weight loss and low testosterone levels. JAMA 293:763–770
Wang C, Swedloff RS, Iranmanesh A, Dobs A, Snyder PJ, Cunningham G,
Matsumoto AM, Weber T, Berman N 2000 Transdermal testosterone gel
improves sexual function, mood, muscle strength, and body composition
parameters in hypogonadal men. Testosterone Gel Study Group. J Clin Endocrinol Metab 85:2839 –2853
Bakhshi V, Elliott M, Gentili A, Godschalk M, Mulligan T 2000 Testosterone improves rehabilitation outcomes in ill older men. J Am Geriatr Soc
48:550 –553
Page ST, Amory JK, Bowman FD, Anawalt BD, Matsumoto AM, Bremner
WJ, Tenover JL 2005 Exogeneous testosterone (T) alone or with finasteride
increases physical performance, grip strength, and lean body mass in older
men with low serum T. J Clin Endocrinol Metab 90:1502–1510
Morley JE, Perry III HM, Kaiser FE, Kraenzle D, Jensen J, Houston K,
Mattammal M, Perry Jr HM 1993 Effects of testosterone replacement therapy
in old hypogonadal males: a preliminary study. J Am Geriatr Soc 41:149 –152
Sih R, Morley JE, Kaiser FE, Perry III HM, Patrick P, Ross C 1997 Testosterone
replacement in older hypogonadal men: a 12-month randomized controlled
trial. J Clin Endocrinol Metab 82:1661–1667
Snyder PJ, Peachey H, Hannoush P, Berlin JA, Loh L, Lenrow DA, Holmes
JH, Dlewati A, Santanna J, Rosen CJ, Strom BL 1999 Effect of testosterone
treatment on body composition and muscle strength in men over 65 years of
age. J Clin Endocrinol Metab 84:2647–2653
Ly LP, Jimenez M, Zhuang TN, Celermajer DS, Conway AJ, Handelsman DJ
2001 A double-blind, placebo-controlled, randomized clinical trial of transdermal dihydrotestosterone gel on muscular strength, mobility, and quality of
life in older men with partial androgen deficiency. J Clin Endocrinol Metab
86:4078 – 4088
Bhasin S, Storer TW, Berman N, Callegari C, Clevenger B, Phillips J, Bunnell
TJ, Tricker R, Shirazi A, Casaburi R 1996 The effects of supraphysiologic doses
of testosterone on muscle size and strength in normal men. N Engl J Med
335:1–7
Bhasin S, Woodhouse L, Casaburi R, Singh AB, Mac RP, Lee M, Yarasheski
KE, Sinha-Hikim I, Dzekov C, Dzekov J, Magliano L, Storer TW 2005 Older
men are as responsive as young men to the anabolic effects of graded doses
of testosterone on the skeletal muscle. J Clin Endocrinol Metab 90:678 – 688
Nestler JE, Barlascini CO, Clore JN, Blackard WG 1988 Dehydroepiandrosterone reduces serum low density lipoprotein levels and body fat but does not
alter insulin sensitivity in normal men. J Clin Endocrinol Metab 66:57– 61
Usiskin KS, Butterworth S, Clore JN, Arad Y, Ginsberg HN, Blackard WG,
Nestler JE 1990 Lack of effect of dehydroepiandrosterone in obese men. Int J
Obes 14:457– 463
Welle S, Jozefowicz R, Statt M 1990 Failure of dehydroepiandrosterone to
influence energy and protein metabolism in humans. J Clin Endocrinol Metab
71:1259 –1264
Brown GA, Vukovich MD, Sharp RL, Reifenrath TA, Parsons KA, King DS
1999 Effect of oral DHEA on serum testosterone and adaptations to resistance
training in young men. J Appl Physiol 87:2274 –2283
O’Donnell AB, Araujo AB, McKinlay JB 2004 The health of normally aging
men: The Massachusetts Male Aging Study (1987–2004). Exp Gerontol 39:975–
984
Feldman HA, Longcope C, Derby CA, Johannes CB, Araujo AB, Coviello
AD, Bremner WJ, McKinlay JB 2002 Age trends in the level of serum
testosterone and other hormones in middle-aged men: longitudinal results
from the Massachusetts male aging study. J Clin Endocrinol Metab 87:
589 –598
Reuben DB, Siu AL 1990 An objective measure of physical function of elderly
outpatients. The Physical Performance Test. J Am Geriatr Soc 38:1105–1112
Sherman SE, Reuben D 1998 Measures of functional status in communitydwelling elders. J Gen Intern Med 13:817– 823
Bohannon RW 1986 Test-retest reliability of hand-held dynamometry during
a single session of strength assessment. Phys Ther 66:206 –209
Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer
DG, Scherr PA, Wallace RB 1994 A short physical performance battery
assessing lower extremity function: association with self-reported disability
and prediction of mortality and nursing home admission. J Gerontol 49:
M85–M94
Everett P, Sills S 1952 The relationship of grip strength to stature, somatotype
componenets, and antropometric measurements of the hand. Res Q 23:161–166
Schroll M, Bjornsbo-Scroll K, Ferrt N, Livingstone MB 1996 Health and
physical performance of elderly Europeans. SENECA Investigators. Eur J Clin
Nutr 50(Suppl 2):S105–S111
Snyder P, Peachey H, Hannoush P, Berlin JA, Loh L, Holmes JH, Dlewati A,
Staley J, Santanna J, Kapoor SC, Attie MF, Haddad Jr JG, Strom BL 1999
Effect of testosterone treatment on bone mineral density in men over 65 years
of age. J Clin Endocrinol Metab 84:1966 –1972
O’Donnell et al. • Androgens and Physical Performance in Older Men
34. Tan RS, Culberson JW 2003 An integrative review on current evidence of
testosterone replacement therapy for the andropause. Maturitas 45:15–27
35. Guay AT, Perez JB, Fitaihi WA, Vereb M 2000 Testosterone treatment in
hypogonadal men: prostate-specific antigen level and risk of prostate cancer.
Endocr Pract 6:132–138
36. Oettel M 2003 Testosterone metabolism, dose-response relationships and receptor polymorphisms: selected pharmacological/toxicological considerations on benefits versus risks of testosterone therapy in men. Aging Male
6:230 –256
37. Juul A, Skakkebaek NE 2002 Androgens and the ageing male. Hum Reprod
Update 8:423– 433
J Clin Endocrinol Metab, February 2006, 91(2):425– 431
431
38. Legrain S, Girard L 2003 Pharmacology and therapeutic effects of dehydroepiandrosterone in older subjects. Drugs Aging 20:949 –967
39. Wolf OT, Neumann O, Hellhammer DH, Geiben AC, Strasburger CJ,
Dressendorfer RA, Pirke KM, Kirschbaum C 1997 Effects of a two-week
physiological dehydroepiandrosterone substitution on cognitive performance
and well-being in healthy elderly women and men. J Clin Endocrinol Metab
82:2363–2367
40. Morales AJ, Haubrich RH, Hwang JY, Asakura H, Yen SS 1998 The effect of
six months treatment with a 100 mg daily dose of dehydroepiandrosterone
(DHEA) on circulating sex steroids, body composition and muscle strength in
age-advanced men and women. Clin Endocrinol (Oxf) 49:421– 432
JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the
endocrine community.