0021-972X/89/6904-776$02.
00/0
Journal of Clinical Endocrinology and Metabolism
Copyright 1989 by The Endocrine Society
Increases in Bone Density During Treatment of Men
with Idiopathic Hypogonadotropic Hypogonadism*
JOEL S. FINKELSTEIN, ANNE KLIBANSKI, ROBERT M. NEER,
SAMUEL H. DOPPELT, DANIEL I. ROSENTHAL, GINO V. SEGRE, AND
WILLIAM F. CROWLEY, JR.
Reproductive Endocrine, Mineral Metabolism, and Thyroid Units, Department of Medicine and the Clinical
Research Center, Massachusetts General Hospital, Boston, Massachusetts 02114
ABSTRACT. To assess the effects of gonadal steroid replace-
ment on bone density in men with osteoporosis due to severe
hypogonadism, we measured cortical bone density in the distal
radius by 125I photon absorptiometry and trabecular bone density
in the lumbar spine by quantitative computed tomography in 21
men with isolated GnRH deficiency while serum testosterone
levels were maintained in the normal adult male range for 12-
31 months (mean SE, 23.7 1.1). In men who initially had
fused epiphyses (n = 15), cortical bone density increased from
0.71 0.02 to 0.74 0.01 g/cm2 (P < 0.01), while trabecular
bone density did not change (116 9 compared with 119 7
mg/cm3). In men who initially had open epiphyses (n = 6),
O STEOPOROSIS can cause significant morbidity in
men, and hypogonadism is a major risk factor for
its development (1-3). We have previously demonstrated
that men with idiopathic hypogonadotropic hypogonad-
ism (IHH), who are hypogonadal due to an isolated
deficiency of hypothalamic GnRH, have marked de-
creases in both cortical and trabecular bone density
compared to age-matched controls (4). Premature osteo-
porosis has also been reported in men with hypogonadism
associated with hyperprolactinemia (5), anorexia nervosa
(6), and Klinefelter's syndrome (7-9). Although few data
exist regarding the effects of androgen replacement on
the osteoporosis of hypogonadal men, cortical bone den-
sity increases in men with hyperprolactinemic hypogo-
nadism when normal testicular function is restored (5,
10). Furthermore, testosterone therapy is associated with
increases in bone formation in men with primary hypo-
gonadism (11,12). However, the effect of gonadal steroid
restoration on cortical and trabecular bone densities in
men with osteopenia due to isolated GnRH deficiency is
Received January 30, 1989.
Address all correspondence and requests for reprints to: Joel S.
Finkelstein, M.D., Mineral Metabolism Unit, Bulfinch 3, Massachu-
setts General Hospital, Fruit Street, Boston, Massachusetts 02114.
* This work was supported by NIH grants HD-15788, HD-21204,
and RR-01066 and from the Vincent Memorial Fund.
776
Vol. 69, No. 4
Printed in U.S.A.
\*
cortical bone density increased from 0.62 0.01 to 0.70 0.03
g/om2 (P < 0.01), while trabecular bone density increased from
96 13 to 109 12 mg/cm3 (P < 0.01). Cortical bone density
increased 0.03 0.01 g/cm2 in men with fused epiphyses and
0.08 0.02 g/cm2 in men with open epiphyses (P < 0.05). Despite
these increases, neither cortical nor trabecular bone density
returned to normal levels. Histomorphometric analyses of iliac
crest bone biopsies demonstrated that most of the men had low
turnover osteoporosis, although some men had normal to high
turnover osteoporosis. We conclude that bone density increases
during gonadal steroid replacement of GnRH-deficient men,
particularly in men who are skeletally immature. {J Clin Endo-
crinol Metab 69: 776, 1989)
unknown.
The osteopenia of hypogonadal men could be due to a
decrease in the peak bone mass achieved during skeletal
maturation, an accelerated loss of established bone, or
both. Histomorphometric analyses of bone can be useful
in assessing the mechanism of decreased bone mass in
patients with osteopenia. However, analyses of bone
biopsies from hypogonadal men have produced contra-
dictory results regarding the cause of their osteopenia
depending on the patients studied (11-13). These dis-
crepancies may be related to heterogeneity within the
patient populations, concommitant 1,25-dihydroxyvi-
tamin D3 deficiency in some of the study groups, or
variability in the mechanism for osteopenia in hypogo-
nadal men. Because men with isolated GnRH deficiency
are a relatively homogeneous population, they are an
excellent model in which to study androgen effects on
bone density and metabolism.
To assess the effects of gonadal steroid replacement
on cortical and trabecular bone density in 21 men with
GnRH deficiency, we made serial measurements of cor-
tical bone density in the shaft of the radius by 125I photon
absorptiometry and trabecular bone density in the lum-
bar vertebral bodies by quantitative computed tomogra-
phy (CT). To evaluate the relative importance of skeletal
 INCREASES IN BONE DENSITY IN IHH MEN
maturation in the alteration in bone density occurring
during gonadal steroid replacement, we compared the
effect of androgen replacement therapy on bone density
in 2 groups of GnRH-deficient men: those who had open
epiphyses when initially studied, and those who had
closed epiphyses. Finally, we performed quantitative
histomorphometric analyses of iliac crest biopsies in a
subset of 9 patients to determine the histological corre-
lates of the changes in bone density that were observed
during gonadal steroid replacement.
Materials and Methods
Subjects
Patients with IHH. Twenty-one Caucasian men, 19-53 yr old
(mean SE, 29 2), were selected on the basis of the following
criteria (14); 1) failure to undergo puberty by age 18 yr (n =
20) or isolated idiopathic loss of gonadotropin secretion after
puberty (n = 1); 2) serum testosterone levels less than 3.5
nmol/L in the presence of low or normal circulating gonadotro-
pin levels; 3) normal free T4 index; 4) normal levels of TSH
and PRL at baseline and after the iv injection of 200 fig TRH;
5) normal GH and cortisol levels at baseline and in response to
insulin-induced hypoglycemia; and 6) normal CT scan of the
hypothalamic-pituitary region. None of the patients had a
history of hyperparathyroidism, Cushing's syndrome, liver dis-
ease, renal disease, or other chronic illness. No patients had
received corticosteroids, anticonvulsants, calcium, or vitamin
D supplements, and all patients had normal levels of serum
calcium, inorganic phosphate, and PTH. Fifteen of the 21
patients consumed alcohol. In 12 of these men, the amount of
alcohol ingested was less than 30 mL (1 oz)/day. The other 3
- men consumed about 60 mL (2 oz) alcohol/day. Ten men
smoked cigarettes, with a range of consumption from 7-25
pack-yr. No patients were involved in regular vigorous exercise.
Patients were divided into two groups on the basis of their
bone ages (15) at the time of their initial investigations. Group
I consisted of 15 patients with fused epiphyseal plates (bone
. age of 18 yr or more); group II consisted of 6 patients with open
epiphyses (mean bone age SE, 16.1 0.4 yr). The baseline
clinical, hormonal, and bone density data for these men have
been reported in our previous study (4).
Controls. For patients with fused epiphyses (group I), cortical
bone density measurements were compared with our previously
established normative data from chronological age-matched
men (4). For patients with open epiphyses (group II), cortical
bone density measurements were compared to those from con-
trol patients who were matched for bone age (16). Because the
published data for trabecular bone density in normal adoles-
cents (17) are insufficiently detailed to compare with our data
on the basis of bone age, trabecular bone density was compared
with normative data for adult men (18) for all patients (groups
I and II).
777
Protocol
Patients were admitted to the Mallinckrodt General Clinical
Research Center at the Massachusetts General Hospital for the
following studies: a complete history and physical examination,
radiographs of the hand and wrist for bone age using the method
of Tanner and Whitehouse (15), determination of cortical and
trabecular bone density, and measurements of serum concen-
trations of testosterone, PTH, 25-hydroxyvitamin D3, 1,25-
dihydroxyvitamin D3, and alkaline phosphatase. Initial serum
hormone levels were determined by previously described RIAs
(19-22) on pools of equal aliquots of blood specimens that were
drawn every 10 min for 24 h during an admission to the General
Clinical Research Center when the patients were receiving no
hormone replacement therapy. Subsequent levels of serum tes-
tosterone, PTH, 25-hydroxyvitamin D3,1,25-dihydroxyvitamin
D3, and alkaline phosphatase were determined during therapy
on blood samples drawn on the same day that bone density
measurements were made. Daily oral calcium intake was esti-
mated from a detailed retrospective dietary history performed
by a research dietitian at the center.
Seventeen of the 21 patients (all men in group I and 2 of 6
men in group II) had received prior androgen replacement
therapy and were sexually mature {i.e. Tanner stage V geni-
talia) at the time of their initial bone density studies. The mean
duration of prior therapy was 3.2 0.7 yr in group I and 0.2
0.1 yr in group II. Subsequent treatment in these patients and
in the 4 others was with im testosterone enanthate (3 patients),
hCG (8 patients), or pulsatile GnRH (10 patients) (23, 24) in
doses titrated to maintain serum testosterone levels within the
normal adult male range. The patients were followed prospec-
tively and underwent an identical set of studies after serum
testosterone levels had been sustained in the normal adult male
range for at least 1 yr (range, 12-31 months; mean, 23.7 1.1).
This protocol was approved by the Subcommittee on Human
Studies at the Massachusetts General Hospital, and all patients
provided written informed consent.
Bone density determinations
Cortical bone density was determined by 125I photon absorp-
tiometry (25, 26) using a Norland model 278A bone densitom-
eter (Norland Instruments, Fort Atkinson, WI). Quadruplicate
readings were obtained from the junction of the proximal two
thirds and distal one third of the nondominant radius, a site
that contains greater than 90% cortical bone. Our coefficient
of variation for replicate patient measurements on the same
day is 1.5-2.0%, and that for long term instrumental stability
(assessed by daily measurements of an aluminum cortical bone
phantom) is 1.0%. Previous studies have demonstrated that the
reproducibility of single photon absorptiometry at the midshaft
site is 2.4% for measurements made with this instrument over
a 3-yr period (27). Trabecular bone density was measured with
a General Electric model 8800 CT scanner (General Electric
Co., Schenectady, NY), as described previously (28), and results
are expressed in terms of milligrams of K2HPO4 standard per
cm3. Axial scans were obtained through the midbody of the
first four lumbar vertebrae, and their trabecular bone density
778 FINKELSTEIN ET AL.
was then measured and averaged. The likelihood that a dupli-
cate average would fall within 10 CT units was greater than
95%.
Iliac crest bone biopsies
Nine unselected patients (6 from group I and 3 from group
II) underwent bicortical iliac crest biopsy under local anesthesia
at a time when serum testosterone levels had been normal for
at least 1 yr. Eight of these men had received a double tetra-
cycline label before the biopsy. The undecalcified specimens
were embedded in methylmethacrylate, and 3-/*m sections were
stained with von Kossa, Toluidine blue, and Goldner trichrome.
Ten-micron unstained sections were used for quantitative flu-
orescent analysis. Histomorphometric analysis was performed
by Nichols Institute (San Juan Capistrano, CA) using a previ-
ously described method (29, 30). This same technique was used
to establish normal data in 21 healthy men, aged 20-42 yr
(kindly provided by Dr. Dennis Andress, University of Wash-
ington). The following histological indices were assessed: os-
teoid surface (as a percentage of the total surface), osteoblastic
surface (cuboidal or plump osteoblasts, as a percentage of the
total surface), bone apposition rate (microns per day), daily
rate of bone formation at the tissue level (square microns per
mm2 tissue/day), resorbing surface (as a percentage of total
bone surfaces), osteoid width (microns), osteoid area (as a
percentage of the total bone area), and mineralization lag time
(days). The bone formation rate at the tissue level was deter-
mined by multiplying the bone apposition rate (distance be-
tween the 2 tetracycline labels divided by the number of days
of tetracycline administration) by the total length of bone
surface occupied by double tetracycline labels, and then nor-
malizing by the total bone area. The mineralization lag time
was determined by dividing the osteoid width by the bone
apposition rate.
Statistical analyses
Comparisons of all parameters between patients and controls
or between groups I and II were made by Student's nonpaired
t test or, for data that were nonnormally distributed, a Mann-
Whitney test. Comparisons of all baseline and follow-up param-
eters within each group were made by Student's paired t test.
The amount of change in cortical and trabecular bone density
was compared between patients treated with hCG and pulsatile
GnRH using Student's nonpaired t test. Because only 3 men
were treated with testosterone enanthate, the amount of change
in bone density in these men was not compared statistically
with that in the other patients. Multiple linear regression
analyses with the method of least squares were used to assess
the relationship between the patient's chronological age, age at
the onset of therapy, duration of therapy, oral calcium intake,
and serum alkaline phosphatase level with changes in cortical
or trabecular bone density for the entire group of patients.
Parameters of bone histomorphometry were compared by Stu-
dent's nonpaired t test with data from 21 normal men. All
results are expressed as the mean SE.
JCE & M 1989
Vol 69 No 4
Results
Clinical characteristics
At the time of the initial bone density measurements
the 15 men with fused epiphyses (group I) were 24-52 yr
old (mean, 32 2), had serum testosterone levels of 0.6-
3.3 nmol/L (mean, 1.7 0.2; normal adult male range
11.4-33.6 nmol/L) (31), and had an adult bone age (4).
The 6 patients with open epiphyses (group II) were 19-
26 yr old (mean, 21 1; P < 0.01 compared with group
I), had serum testosterone levels of 0.5-3.1 nmol/L
(mean, 1.4 0.4), and had bone ages between 14.3-17.1
yr (mean, 16.1 0.4) at the time of the initial evaluation
(4). At the time of the final evaluation during therapy
serum testosterone levels had been in the normal adult
male range for 15-31 months (mean, 25.2 1.0 months)
in the patients in group I and for 12-25 months (mean,
20.0 2.4 months) in the patients in group II (P < 0.05).
Serum testosterone levels during therapy were similar in
both groups (18.4 2.6 vs. 23.9 4.2 nmol/L). At the
time of final evaluation 2 patients in group II still had
open epiphyses, with bone ages of 16 and 17 yr.
Cortical bone density
Initial determinations. The initial cortical bone density
was significantly below that of age- and sex-matched
controls both in group I with fused epiphyses (0.71
0.02 compared with 0.86 0.01 g/cm2; P < 0.0001) (4)
and in group II with open epiphyses (0.62 0.01 com-
pared with 0.79 0.02 g/cm2; P < 0.001) (4). When the
patients were initially evaluated, cortical bone density
was significantly higher in group I than in group II (P <
0.003; Fig. 1).
Response to androgen replacement. After serum testos-
terone levels were normalized for at least 1 yr, cortical
bone density increased in 14 of 15 patients in group I,
with mean bone density levels increasing from 0.71
0.02 to 0.74 0.01 g/cm2 (P < 0.01; Fig. 1). Cortical bone
density increased in all 6 patients in group II, with mean
bone density levels increasing from 0.62 0.01 to 0.70
0.03 g/cm2 (P < 0.01; Fig. 1). Cortical bone density
increased more in group II than in group I patients (0.08
0.02 us. 0.03 0.01 g/cm 2 ; P < 0.05; Fig. 1). After
treatment, mean cortical bone density remained lower
than that in controls in both groups (P < 0.001). Cortical
bone density was still higher in group I than in group II
after treatment, but this difference was no longer statis-
tically significant (P = 0.13). The change in cortical bone
density was roughly proportional to the initial alkaline
phosphatase level (r = 0.52; P < 0.02). There was no
significant relationship between the change in cortical
bone density and the patients' chronological age, age that
treatment was begun, duration of therapy, or dietary
 INCREASES IN BONE DENSITY IN IHH MEN
* p<0.01
o.g
t 0.8
0.7
0)
I 8
0.6
ij
0.5
Initial Final Initial Final
GROUP I GROUP I I
FIG. 1. Cortical bone density before and after treatment of men with
isolated GnRH deficiency who initially had fused epiphyses (group I)
or who initially had open epiphyses (group II). The solid lines represent
the mean cortical bone density, which increased significantly after
treatment in both groups I and II (P < 0.01). Cortical bone density
increased more in group II than in group I (P < 0.05).
calcium intake. There was no significant difference in
the amount of change in cortical bone density between
the patients who were treated with HCG and those who
were treated with pulsatile GnRH.
Trabecular bone density
Initial determinations. Initial and follow-up trabecular
bone density measurements in the patients in groups I
and II are shown in Fig. 2. The initial trabecular bone
density for all patients was significantly less than that
in controls (110 8 compared with 177 4 mg/cm3; P
< 0.0001) (4), and there was no significant difference in
initial trabecular bone densities between the two groups
(116 9 vs 96 13 mg/cm3; P = 0.28).
Response to androgen replacement. After serum testos-
terone levels had been normal for at least 1 yr, mean
trabecular bone density did not change in group I with
fused epiphyses (116 9 vs. 119 7 mg/cm3; P = 0.60),
although several patients had dramatic increases (Fig.
2). Trabecular bone density increased in all six patients
in group II, with mean levels increasing form 96 13 to
109 12 mg/cm3 (P < 0.01; Fig. 2). After treatment,
trabecular bone density for all patients remained below
that in controls (116 6 vs. Ill 4 mg/cm3; P < 0.0001)
and was similar in groups I and II, (119 7 vs. 109 12
mg/cm3; P = 0.43). There was no significant relationship
between the change in trabecular bone density and the
779
s; 200 * p<0.01
S 180
160
JJ
140
to 120
c:
100
80
60
40
f- Initial Final Initial Final
GROUP I GROUP I I
FlG. 2. Vertebral bone density before and after treatment of men with
isolated GnRH deficiency in patients who initially had fused epiphyses
(group I) and in patients who initially had open epiphyses (group II).
The solid lines represent the mean vertebral bone density, which
increased significantly (P < 0.01) after treatment only in group II.
patients' chronological age, age that treatment was be-
gun, duration of therapy, dietary calcium intake, or ini-
tial serum alkaline phosphatase level. There was no
significant difference in the amount of change in trabec-
ular bone density between the patients treated with hCG
and those treated with pulsatile GnRH. Furthermore,
there was no significant relationship between the change
in trabecular bone density and the change in cortical
bone density.
Serum studies and dietary calcium intake
Mean serum alkaline phosphatase levels were higher
in the patients with open epiphyses (group II) than in
the patients with fused epiphyses (group I) at the time
of the initial studies (41 4 vs. 27 2 U/L; P < 0.005)
(4) and at the time of the follow-up evaluation (37 6
vs. 25 2 U/L; P < 0.05). Mean serum alkaline phos-
phatase levels did not change significantly in either group
during the study. Initial serum PTH (15 1 ng/L), 25-
hydroxyvitamin D3 (46 4 nmol/L), and 1,25-dihydroxy-
vitamin D3 (95 6 pmol/L) levels were normal in all
patients and did not change significantly during therapy
(16 1 ng/L, 51 5 nmol/L, and 93 8 pmol/L,
respectively). Dietary calcium intake was also similar in
the two groups of patients, at the time of both the initial
(807 1 1 6 vs. 2016 736 mg/day) and follow-up bone
density studies (1008 101 vs. 1205 317 mg/day).
780 FINKELSTEIN ET AL. JCE & M 1989
Vol 69 No 4

Bone histomorphometry suggests that once trabecular osteoporosis is established

Table 1 shows the results of quantitative histomor-
phometry in nine patients. Two distinct histological pic-
tures were seen (Fig. 3). Seven patients had low turnover
osteoporosis, with decreases in bone resorption surfaces
and indices of bone formation (percent osteoid surfaces,
percent osteoid area, and tissue-based bone formation),
and their mineralization lag time was prolonged (Table
1 and Fig. 3). In two of these seven men the daily rate of
bone formation was so low that no bone surfaces took up
both tetracycline labels. The other two patients had
normal to high turnover osteoporosis with high normal
percent resorption surfaces and osteoid surfaces, elevated
percent osteoblastic surfaces, and a normal rate of bone
formation (Table 1 and Fig. 3). Both groups of patients
had wide osteoid seams.
Discussion
In this study we have demonstrated that gonadal ste-
roid replacement is associated with increases in bone
density in men with hypogonadism due to isolated GnRH
deficiency and that the effect of testosterone replacement
on bone density is related to the initial degree of bone
maturation. In men who initially had open epiphyses,
both cortical and trabecular bone density increased. Dur-
ing androgen replacement in men who initially had fused
epiphyses, only cortical bone density increased, and this
increase was less pronounced than that in the men with
open epiphyses. Because increases in bone density were
more apparent in the hypogonadotropic men with im-
mature bone ages, these results suggest that most of the
increase in bone density that occurs during treatment of
pubertal hypogonadotropic men reflects the normal proc-
ess of bone accretion that occurs during sexual matura-
tion (17, 32-34). The failure of trabecular bone density
to increase in hypogonadotropic men with mature bones
TABLE 1. Bone histomorphometry in nine men with isolated GnRH deficiency
in skeletally mature hypogonadotropic men, the defect
may not be reversible by androgen replacement over the
short term. In addition, it is likely that prior androgen
therapy limited the increases in bone density in these
men.
Several studies have investigated the effects of nor-
malization of gonadal steroids on osteopenia. Restoration
of normal gonadal function increases cortical bone mass
in both men and women with hyperprolactinemic hypo-
gonadism (5,35). In women with hypogonadism and bone
loss due to the biochemical castration accompanying
GnRH analog administration, trabecular bone density
returns to normal once ovarian function is rapidly re-
stored (36, 37). Finally, although most studies have dem-
onstrated that estrogen replacement can only arrest fur-
ther bone loss in postmenopausal women (38), some
investigators have demonstrated small increases in bone
density during estrogen replacement therapy (39-42).
The reasons why cortical bone density may show a
greater response to gonadal steroid replacement than
trabecular bone are unclear. Trabecular bone is more
sensitive than cortical bone to estrogen deficiency in
women (37, 43-46). It is possible that the differential
responses of cortical and trabecular bone to androgen
replacement in men with fused epiphyses reflect a differ-
ence in the time that peak bone mass is reached; peak
trabecular bone density normally occurs by age 16-18 yr
(17), whereas peak cortical bone density is not achieved
until early adulthood (4, 16, 47). Our inability to detect
changes in trabecular bone density in men with fused
epiphyses may also reflect differences in the precision of
bone density measurements, because our trabecular bone
density technique is less precise than our technique for
measuring cortical bone density. Further studies are
needed to distinguish among these possible explanations.
Prior studies have produced conflicting data concern-

GnRH-deficient men
Low turnover Others All patients (n = 21)
(n = 7) (n = 2) (n = 9)
Osteoid surface (%) 10.8 2.26 29.2 2.4 14.9 3.2C 23.8 2.0
Osteoblastic surface (%) 2.4 0.6 9.0 1.5 3.9 1.1 3.4 0.6
Osteoid width (/tin) 11.8 1.6e 16.7 1.5 12.9 1.5* 7.9 1.7
Osteoid area (%) 1.6 0.56 2.6 0.2 1.8 0.46 3.6 0.3
Rate of bone apposition (/um/day) 0.5 0.1 1.0 0.1 0.7 0.1 0.6 0.1
Rate of bone formation (^rnVmrn2 tissue-day) 50 8 6 198 19 99 32* 286 32
Mineralization lag time (days) 27.2 5.3d 17.5 0.7 24.0 3.0c 13.4 0.7
Resorption surface (%) 1.2 0.56 6.8 0.6 2.8 1.0c 4.7 0.6
All data are the mean SE.
" From Dr. Dennis Andress, University of Washington (with permission).
b
P< 0.01 vs. normal men.
c
P < 0.05 vs. normal men.
d
P< 0.08 vs. normal men.
 INCREASES IN BONE DENSITY IN IHH MEN
FIG. 3. Light micrographs stained with
Toluidine blue {left) and tetracycline flu-
orescence studies (right) of bone biopsies
in a GnRH-deficient man with low turn-
over osteoporosis (A and B) and a
GnRH-deficient man with the less com-
mon normal to high turnover osteopo-
rosis (C and D). The arrows in A indicate
the inactive bone surface. No tetracy-
cline uptake is visible in B. Active bone
turnover is indicated in C by the pres-
ence of two osteoclasts within a How-
ship's lacunae and a row of osteoblasts
making osteoid. Tetracycline labels in D
are indicated by the arrows. OS, Osteoid;
OB, osteoblasts; OCL, osteoclasts; HL,
Howship's lacunae; Tr, trabecular bone;
BM, normal bone marrow.
ing the effects of androgen administration on bone his-
tomorphometry in hypogonadal men. Some investigators
have suggested that testosterone replacement is associ-
ated with an increase in bone formation (11, 12), while
others have suggested that testosterone replacement de-
creases the rate of bone formation (13). Most of our
patients had low turnover osteoporosis at a time when
they had been eugonadal for at least 1 yr, although bone
turnover was normal or increased in two men. The pro-
longed mineralization lag time and increased osteoid
seam width both suggest a concomitant mineralization
defect. Our data suggest that the osteoporosis of men
with IHH can result from a variety of defects in bone
metabolism, although, as suggested previously (4), a de-
fect in bone formation appears to be the most common
abnormality. Because the bone biopsies were performed
when the patients were eugonadal, it is possible that the
bone histomorphometry was altered by androgen replace-
ment. The reason for the mineralization defect and the
different histological types of osteoporosis in these men
is unclear and warrants further study.
Although bone density increased during gonadal ste-
roid replacement, it failed to reach normal adult levels
in these men over 1-3 yr of treatment. This finding may
be due to the defect in bone formation in treated patients
demonstrated by the bone histomorphometry. The fail-
ure of bone density to reach normal adult levels may also
indicate that normal bone development requires an in-
terplay among several different factors, only one of which
has been replaced by our therapy; that puberty must
occur at the appropriate time in order to achieve a normal
adult bone mass; or that a more prolonged period of
781
androgen replacement is needed.
Other investigators have also demonstrated that bone
density is diminished in patients who undergo abnormal
puberty. For example, cortical bone mineral content is
decreased in boys with delayed puberty (32), and scoliosis
and stress fractures are common in ballet dancers with
delayed menarche (48). Similarly, cortical bone density
is decreased in men with Klinefelter's syndrome (7, 8)
and in women with Turner's syndrome, even when com-
paring the patients with bone age-matched controls (49).
These findings are all consistent with the hypothesis
that adolescent gonadal steroid deficiency causes a defect
in bone development. However, because IHH, Klinefel-
ter's syndrome, and Turner's syndrome are genetic or
congenital disorders that can be associated with other
somatic defects, we cannot exclude the possibility that
such patients have an independent predisposition for
osteoporosis.
A limitation of the present study deserves mention.
Because treatment is indicated in young men with hy-
pogonadism to stimulate pubertal development and
maintain normal sexual function, an untreated control
group was not included for comparison. In a cross-sec-
tional analysis of normal adult men, cortical bone density
increased slightly until age 25 yr (4, 16, 47). However,
since osteopenia occurs in men who acquired hypogo-
nadism in adult life (1,5), bone density would be expected
to decrease in adult men with untreated hypogonadism.
Therefore, we feel that the increase in cortical bone
density observed in our men with mature bones is related
to their androgen therapy. In a cross-sectional analysis
of normal adolescent boys, cortical bone density in-
782 FINKELSTEIN ET AL.
creased dramatically between the ages of 16 and 18 yr
(16), as it did in our patients with open epiphyses whose
bone age before therapy averaged 16.1 0.4 yr. This
observation suggests that the increase in bone density in
our men with immature bones may well reflect the com-
pletion of the normal process of pubertal bone accretion.
We conclude that bone density increases during go-
nadal steroid replacement in men with GnRH deficiency,
although cortical and trabecular bone may have different
responses to androgen replacement. Cortical bone den-
sity increases regardless of the degree of bone matura-
tion, while trabecular bone density increases only in men
who have open epiphyses. Histomorphometric analyses
of bone during therapy suggest that the osteopenia of
hypogonadotropic men is usually due to a persistent
defect in bone formation, although some men have nor-
mal or increased bone turnover. These data underscore
the importance of early diagnosis and consistent replace-
ment with gonadal steroids in hypogonadal men to max-
imize their peak bone density and thereby decrease their
risk of future fractures. Further studies are required to
determine whether cortical bone density can be normal-
ized by more prolonged therapy and if the osteoporosis
of hypogonadotropic men can be entirely prevented if
treatment is begun early.
Acknowledgments
The authors thank Donna Peltier-Saxe, R.N., and Denise Heavern,
R.N., for their excellent clinical assistance; Ellen Williams, R.D., and
Margie McCullough, R.D, for their help with the dietary histories;
William Mayo-Smith, M.D., for helping to evaluate the CT scans of
the spine; and the staff of the Mallinckrodt General Clinical Research
Center for their dedicated care of the patients.
References
1. Seeman E, Melton III LJ, O'Fallon WM, Riggs BL. Risk factors
for spinal osteoporosis in men. Am J Med. 1983;75:977-83.
2. Albright F, Reifenstein Jr EC. Metabolic bone disease: osteopo-
rosis. In: Albright F, Reifenstein Jr EC, eds. The parathyroid glands
and metabolic bone disease. Baltimore: Williams and Wilkins;
1948:145-204.
3. Swartz CM, Young MA. Male hypogonadism and bone fracture. N
Engl J Med. 1988;318:996.
4. Finkelstein JS, Klibanski A, Neer RM, Greenspan SL, Rosenthal
DI, Crowley Jr WF. Osteoporosis in men with idiopathic hypogo-
nadotropic hypogonadism. Ann Intern Med. 1987;106:354-61.
5. Greenspan SL, Neer RM, Ridgway EC, Klibanski A. Osteoporosis
in men with hyperprolactinemic hypogonadism. Ann Intern Med.
1986;104:777-82.
6. Rigotti NA, Neer RM, Jameson L. Osteopenia and bone fractures
in a man with anorexia nervosa and hypogonadism. JAMA.
1986;256:385-8.
7. Smith DA, Walker MS. Changes in plasma steroids and bone
density in Klinefelter's syndrome. Calcif Tissue Res.
1977;22(Suppl):225-8.
8. Foresta C, Ruzza G, Mioni R, Meneghello A, Baccichetti C. Tes-
tosterone and bone loss in Klinefelter syndrome. Horm Metab Res.
1983;15:56-7.
9. Foresta C, Zanatta GP, Busnardo B, Scanelli G, Scandellari C.
Testosterone and calcitonin plasma levels in hypogonadal osteo-
JCE & M 1989
Vol 69 No 4
porotic young men. J Endocrinol Invest. 1985;8:377-9.
10. Greenspan SL, Oppenheim DS, Klibanski A. Importance of go-
nadal steroids on bone mass in men with hyperprolactinemic
hypogonadism. Ann Intern Med. 1989;110:526-31.
11. Baran DT, Bergfeld MA, Teitelbaum SL, Avioli LV. Effect of
testosterone therapy on bone formation in an osteoporotic hypo-
gonadal male. Calcif Tissue Res. 1978;26:103-6.
12. Francis RM, Peacock M, Aaron JE, et al. Osteoporosis in hypogo-
nadal men: role of decreased plasma 1,25-dihydroxyvitamin D,
calcium malabsorption, and low bone formation. Bone. 1986;7:261-
8.
13. Jackson JA, Kleerekoper M, Parfitt AM, Rao DS, Villanueva AR,
Frame B. Bone histomorphometry in hypogonadal and eugonadal
men with spinal osteoporosis. J Clin Endocrinol Metab.
1987;65:53-8.
14. Crowley Jr WF, Filicori M, Spratt DI, Santoro NF. The physiology
of gonadotropin-releasing hormone (GnRH) secretion in men and
women. Recent Prog Horm Res. 1985;41:473-531. ,
15. Tanner JM, Whitehouse RH, Camerson N, Marshall WA, Healy
MJR, Goldstein H. Assessment of skeletal maturity and prediction
of adult height (TW2 method), 2nd ed. London: Academic Press;
1983; 1-106.
16. Mazess RB, Camerson JR. Bone mineral content in normal U.S.
whites. In: Mazess RB, ed. Proceedings, International Conference
on Bone Mineral Measurement. Washington DC: NIAMDD;
DHEW publication NIH 75-683; 1974;228-38.
17. Gilsanz V, Gibbons DT, Roe TF, et al. Vertebral bone density in
children: the effect of puberty. Radiology. 1988; 166:847-50.
18. Cann CE, Genant HK, Kolb FO, Ettinger B. Quantitative com-
puted tomography for prediction of vertebral fracture risk. Bone.
1985;6:l-7.
19. Rao PN, Moore PH. Synthesis of new steroid haptens for radio-
immunoassay. I. 15/3-Carboxymethylmercaptotestosterone-bovine
serum albumin conjugate: measurement of testosterone in male
plasma without chromatography. Steroids, 1976;28:101-9.
20. Cholst IN, Steinberg SF, Tropper PJ, Fox HE, Segre GV, Bilezikian
JP. The influence of hypermagnesemia on serum calcium and
parathyroid hormone levels in human subjects. N Engl J Med.
1984;310:1221-5.
21. Belsey R, DeLuca HF, Potts JT. Competitive binding assay for
vitamin D and 25-OH vitamin D. J Clin Endocrinol Metab.
1971;33:554-7.
22. Reinhardt TA, Horst RL, Orf JW, Hollis BW. A microassay for
1,25-dihydroxyvitamin D not requiring high performance liquid
chromotography: application to clinical studies. J Clin Endocrinol
Metab. 1984;58:91-8.
23. Hoffman AR, Crowley WF. Induction of puberty in men by long-
term pulsatile administration of low-dose gonadotropin-releasing
hormone. N Engl J Med. 1982;307:1237-41.
24. Spratt DI, Finkelstein JS, O'Dea LStL, et al. Long-term adminis-
tration of gonadotropin-releasing hormone in men with idiopathic
hypogonadotropic hypogonadism. Ann Intern Med. 1986;1O5:848-
55.
25. Cameron JR, Sorenson J. Measurement of bone mineral in vivo:
an improved method. Science. 1963;142:230-2.
26. Cameron Jr, Mazess RB, Sorenson JA. Precision and accuracy of
bone mineral determination by direct photon absorptiometry. In-
vest Radiol. 1968;3:141-50.
27. Siemenda C, Hui SL, Longcope C, Johnston CC. Sex steroids and
bone mass. A study of changes about the time of menopause. J ^
Clin Invest. 1987;80:1261-9.
28. Rosenthal DI, Ganott MA, Wyshak G, Slovik DM, Doppelt SH,
Neer RM. Quantitative computed tomography for spinal density
measurement: factors affecting precision. Invest Radiol.
1985;20:306-10.
29. Sherrard DJ, Baylink DJ, Wergedal JE, Maloney NA. Quantitative
histological studies on the pathogenesis of uremic bone disease. J
Clin Endocrinol Metab. 1974;39:119-35.
30. Andress DL, Endress DB. Maloney NA, Kopp JB, Coburn JW,
Sherrard DJ. Comparison of parathyroid hormone assays with
bone histomorphometry in renal osteodystrophy. J Clin Endocrinol
Metab. 1986;63:1163-9.
 INCREASES IN BONE DENSITY IN IHH MEN
31. Spratt DI, O'Dea LStL, Schoenfeld D, Butler J, Rao PN, Crowley
WF. Neuroendocrine-gonadal axis in men: frequent sampling of
LH, FSH, and testosterone. Am J Physiol. 1988;254:E658-66. 41.
32. Krabbe S, Christiansen C, Rodbro P, Transbol I. Effect of puberty
on rates of bone growth and mineralisation: with observations in
male delayed puberty. Arch Dis Child. 1979;54:950-3.
33. Krabbe S, Christiansen C. Longitudinal study of calcium metabo- 42.
lism in male puberty. I. Bone mineral content, and serum levels of
alkaline phosphatase, phosphate, and calcium. Acta Paediatr
Scand. 1984;73:745-9. 43.
34. Krabbe S, Hummer L, Christiansen C. Longitudinal study of
calcium metabolism in male puberty. II. Relationship between
mineralization and serum testosterone. Acta Paediatr Scand. 44.
1984;73:750-5.
35. Klibanski A, Greenspan SL. Increase in bone mass after treatment
of hyperprolactinemic amenorrhea. N Engl J Med. 1986;315:542- 45.
6.
36. Cann CE, Henzl M, Burry K, et al. Reversible bone loss is produced
by the GnRH agonist nafarelin. In: Cohn DV, Martin TJ, Meunier 46.
PJ, eds. Calcium regulation and bone metabolism: basic and clinical
aspects. Amsterdam: Elsevier; 1987;9:123-7. 47.
37. Johansen JS, Riis BJ, Hassager C, Moen M, Jacobson J, Chris-
tiansen C. The effect of a gonadotropin-releasing hormone agonist
analog (nafarelin) on bone metabolism. J Clin Endocrinol Metab.
1988;67:701-6. 48.
38. Riggs BL, Melton III LJ. Involutional osteoporosis. N Engl J Med.
1986,314:1676-86.
39. Christiansen C, Christensen MS, Transbol I. Bone mass in post-
menopausal women after withdrawal of oestrogen/gestagen re- 49.
placement therapy. Lancet. 1981;1:459-61.
40. Horsman A, Jones M, Francis R, Nordin C. The effect of estrogen
783
dose on postmenopausal bone loss. N Engl J Med. 1983;309:1405-
7.
Lindsay R, Aitken JM, Anderson JB, Hart DM, MacDonald EB,
Clarke AC. Long-term prevention of osteoporosis by oestrogen.
Evidence for an increased bone mass after delayed onset of oestro-
gen treatment. Lancet. 1976;1038-41.
Nachtigall LE, Nachtigall RH, Nachtigall RD, Beckman EM.
Estrogen replacement therapy I: a 10-year prospective study in the
relationship to osteoporosis. Obstet Gynecol. 1979;53:277-81.
Drinkwater BL, Nilson K, Chesnut CH, Bremner WJ, Shainholtz
S, Southworth MB. Bone mineral content of amenorrheic and
eumenorrheic athletes. N Engl J Med. 1984;311:277-81.
Marcus R, Cann C, Madvig P, et al. Menstrual function and bone
mass in elite women distance runners: endocrine and metabolic
features. Ann Intern Med. 1985;102:158-63.
Cann C, Genant HK, Ettinger B, Gordan GS. Spinal mineral loss
in oorphorecomized women: determination by quantitative com-
puted tomography. JAMA. 1980;244:2056-9.
Cann C, Martin MC, Genant HK, Jaffe RB. Decreased spinal
mineral content in amenorrheic women. JAMA. 1984;251:626-9.
Wolanski N. Changes in bone density and cortical thickness of the
second metacarpal between the ages of 3 and 74 years as a method
for investigating bone mineral metabolism. Acta Anat. 1967;67:74-
93.
Warren MP, Brooks-Gunn J, Hamilton LH, Warren LF, Hamilton
WG. Scoliosis and fractures in young ballet dancers. Relation to
delayed menarche and secondary amenorrhea. N Engl J Med.
1986;314:1348-53.
Shore RM, Chesney RW, Mazess RB, Rose PG, Bargman GJ.
Skeletal demineralization in Turner's syndrome. Calcif Tissue Int.
1982;34:519-22.