Growth hormone treatment in growth hormone-deficient adults. I. Effects on muscle mass and strength ROSS C. CUNEO, RICHARD HESP,
FRANC0 SALOMON, C. MARK AND PETER H. SijNKSEN
WILES,
Divisions of Medicine and Neurology, United Medical and Dental Schools of Guy’s and St. Thomas’ Hospitals, St. Thomas’ Hospital, London SE1 7EH; and Division of Radio-Isotopes, Medical Research Council, Northwick Park, Harrow, Middlesex HA1 3UJ, United Kingdom
latter group reported feeling “stronger,” but objective testing showed no consistent change (22). Short-term treatment with hGH in trained healthy adults has shown an increase in fat-free mass and a decrease in fat mass The effect of recombinant DNA human growth hormone (7), but the effect on muscle strength has not been re(rhGH) treatment in adults with growth hormone (GH) defi- ported. Jorgensen et al. (15) recently reported a small but ciency wasstudiedin 24 patients in a double-blind placebo-consignificant increase in quadriceps force trolled trial. The dosewas 0.07 U/kg body wt daily. After 6 mo not statistically after 4 mo of recombinant DNA hGH (rhGH) treatment of treatment, significant increases were noted in the rhGH adults. In adults with GH defigroup for total cross-sectionalarea of thigh muscle(+11.2 k 3.1 in young GH-deficient vs. -0.5 rt 3.0 cm2;P = 0.015vs. placebo) and quadricepsmus- ciency, we have reported major increases in lean body cle (+4.1 t 0.8 vs. +0.4 t, 1.2 cm2; P = 0.031) measuredby mass after 6 mo of treatment with rhGH (23). Here, we computerized tomography. Strong correlations were noted be- describe the effect on skeletal muscle mass and strength tween lean body mass(measuredas total body potassium)and following rhGH treatment in adults with GH deficiency.
CUNEO, Ross C., FRANCO SALOMON, C. MARK WILES, RICHARD HESP, AND PETER H. WNKSEN. Growth hormone treatment in growth hormone-deficient adults. I. Effects on muscle mass and strength. J. Appl. Physiol. 70(Z): 688-694,1991.-
total thigh musclearea in normal and GH-deficient adults both before and after rhGH treatment. Strength of hip flexors (+1.25 t 0.27 vs. +0.25 t 0.12 z-scores; P = 0.004) and limb METHODS girdle musclesincreased (P = 0.02) in the rhGH group. We Patient selection and characteristics. Patients are the conclude that 1) rhGH increaseslean tissue and skeletal mus- same as those from a previous study (23). Eligibility critecle massin adults with human GH deficiency, 2) this suggestsa ria were as follows: 1) age 18-55 yr; 2) GH deficiency for role for GH in the regulation of body composition of adult humans, 3) the increasein strength of limb girdle musclesafter at least 12 mo; and 3) if pituitary hormone replacement rhGH treatment suggeststhat adults with GH deficiency may was required, the treatment was stable for 12 mo before have a proximal myopathy, and 4) the failure to demonstrate entry into the study. Deficiency of GH was defined as a an increasein strength in other musclegroups may require the maximal GH of 5% below the mean. Coefficients of variation for quadriceps force measured over 6 mo in two normal subjects were 2.9 and 3.3% at mean forces of 788 and 689 N, respectively. Standard deviation (SD) scores, or z-scores, were calculated from reference values obtained in 30 healthy males (ages 17-66 yr) and 30 females (ages 19-66 yr) using identical techniques (27) as follows: z-score = (observed force - mean reference force) /SD of reference population. z-Scores allow averaging of several muscle groups and easy comparison with normal data. For example, a z-score of -2.0 represents a value 2 SD below the predicted mean, at the limit of the acceptable normal range. For hip abduction, where normal data were only available for tests performed at the knee, values were
AND
STRENGTH
689
converted using data obtained at both the knee and the ankle in a subset of the patient population. CT of thigh. With the patient in the supine position, a single scan was obtained midway between the greater trochanter and the joint space of the knee, as assessed by a scout film. Slice width was 8 mm, using 125-kV and 21O-mA settings on a Somatom DRH scanner (Seimens, FRG). Scan time was 4 s. The image was recorded on magnetic tape until read. Scans were performed with a lead shield over the pelvis. The gonadal dose was measured in one male (320 and 105 @v from the topogram and tomogram, respectively), and in one female the gonadal dose without the shield was 235 and 45 @v under the shield. Because beam-hardening artifacts in a horizontal plane between the femurs were apparent, preliminary measurements in normal and GH-deficient patients established that muscle and fat in the whole thigh could best be defined using Hounsfield numbers 20-100 and -150 to 0, respectively. In patients without acromegaly, minimal skin area was recorded at these settings. Total cross-sectional areas of muscle were measured in the dominant thigh. By use of a magnetic pencil to outline the muscle planes, quadriceps area was also recorded, with care taken to exclude the neurovascular bundle, the sartorius, and the head of the biceps femoris. Errors due to differences in alignment of the limb with the scanning beam were minimized by correcting the areas of muscle and fat in scans performed at 3 and 6 mo for any change in bone area that differed from that measured at the initial scan, assuming no change in bone area over 6 mo. Quadriceps density and intramuscular fat were estimated in as large an area of quadriceps as possible, anterior to the beam-hardening artifact, with care to avoid obvious fascial fat deposits. Coefficients of variation for measurements based on paired observations on up to 48 scans for total muscle area, quadriceps area, fat area, bone area (Hounsfield nos. 200-2,000), quadriceps density, and percentage of intramuscular fat were 0.4, 1.7, 0.1, 0, 0.8, and 94.7%, respectively. Other measurements. Total body potassium was measured as the naturally occurring radioisotope of potassium (40K) and converted to lean body mass in kilograms as previously reported (23). Activity was assessed as daily distance walked, averaged over 1 wk using a mechanical pedometer (Aurora, Japan). Assays for serum GH and plasma insulin-like growth factor I (IGF-I) and GH absorption profiles were performed as previously reported (23). Statistics. Results are reported as means t SE. Comparisons between groups at baseline used unpaired t tests or x2 tests, and treatment responses were compared using analysis of covariance on the 6-mo data (with baseline data as the covariate) or Mann-Whitney U test on the increment between baseline and 6 mo if nonnormality of the data was evident. Significance was recognized at the 5% level. Relationships between single variables before treatment were explored with simple linear regressions and changes in variables after treatment with multiple regressions, with treatment as a stratifying or binary variable. For changes with more than one significant association, multiple linear regression was also
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 7, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
690
GROWTH
TABLE
1. Patient characteristics rhGH
Age,
HORMONE
yr
Males /Females Ht, cm wt, kg Initial diagnosis Cushing’s disease Prolactinoma Chromophobe adenoma Craniopharyngioma Idiopathic Radiotherapy (medulloblastoma) Pituitary replacement hormones Corticosteroids Thyroxine Sex steroids dDAVP Fludrocortisone
39t3 B/4 170.0t2.7 82.5t3.9
Placebo
38t3 B/4 169.0k3.2 79.3t6.8
AND STRENGTH
only a limited number of patients were available, showed a trend toward increased strength in the rhGH group compared with the placebo group. Mean limb girdle muscle z-scores (shoulder abduction, hip flexion, and hip abduction) increased in the rhGH group (+0.94 t 0.25 vs. +0.28 t 0.16; P = 0.019). Quadriceps force per quadri180
6 0 3 3 0 0 10 11
10 3 1
3 2 3 0 3 1
10 10
170
10 10
8 9
rhGH Placebo
P< 0.02
9 10
8 2 2 I
I
Age, height, and weight are means k SE; n = lZ/group. No significant differences existed between groups before treatment.
10 10
a
9
?hGH
11 11
Placebo
used, with treatment code as a stratifying variable. Treatment-variable interaction was explored and excluded in each case. Coefficients of variation were calculated using the SD of the differences between paired observations divided by the mean of all the observations.
P< 0.03
RESULTS
Patient characteristics are shown in Table 1. Randomization resulted in evenly matched groups with respect to age, height, weight, sex, pituitary pathology, and incidence of pituitary hormone replacement. Lean body mass and CT data. As previously reported (23), the lean body mass increased over 6 mo by 5.6 t 1.1 kg (10.8% higher than baseline) in the rhGH group and decreased 0.1 t 0.9 kg in the placebo group (P < 0.001 for the differences between groups). Similarly, the total cross-sectional area of midthigh muscle in the rhGH group increased 8.1% (137.8 t 13.6 to 149.0 t 14.4 cm2) compared with a decrease of 0.4% (126.0 t 12.3 to 125.4 t 11.4 cm2) in the placebo group (P = 0.015; Fig. 1). Quadriceps area increased in the rhGH group by 5.7% (72.4 t 7.7 to 76.4 t 7.5 cm2) compared with an increase of 0.6% (65.3 t 3.2 to 65.7 t 5.7 cm2) in the placebo group (P = 0.031). Bone area on CT did not change significantly. The reduction in fat area was not statistically significant, in keeping with the skinfold thickness data, where the major reduction occurred in the abdominal region (23). No change in quadriceps density was noted between the two groups (Table 2). Although there was a tendency for the area of quadriceps on CT with a Hounsfield number consistent with fat to decrease in the rhGH group, this failed to reach statistical significance. This conclusion was not altered when those with some detectable intramuscular fat on the initial scan were examined. Muscle strength. A significant increase in the force of hip flexion was demonstrated following rhGH treatment (Table 3). All other muscle groups except knee extension and neck flexion, for which, in the latter case, data on
150
G
FRANC0 SALOMON, C. MARK AND PETER H. SijNKSEN
WILES,
Divisions of Medicine and Neurology, United Medical and Dental Schools of Guy’s and St. Thomas’ Hospitals, St. Thomas’ Hospital, London SE1 7EH; and Division of Radio-Isotopes, Medical Research Council, Northwick Park, Harrow, Middlesex HA1 3UJ, United Kingdom
latter group reported feeling “stronger,” but objective testing showed no consistent change (22). Short-term treatment with hGH in trained healthy adults has shown an increase in fat-free mass and a decrease in fat mass The effect of recombinant DNA human growth hormone (7), but the effect on muscle strength has not been re(rhGH) treatment in adults with growth hormone (GH) defi- ported. Jorgensen et al. (15) recently reported a small but ciency wasstudiedin 24 patients in a double-blind placebo-consignificant increase in quadriceps force trolled trial. The dosewas 0.07 U/kg body wt daily. After 6 mo not statistically after 4 mo of recombinant DNA hGH (rhGH) treatment of treatment, significant increases were noted in the rhGH adults. In adults with GH defigroup for total cross-sectionalarea of thigh muscle(+11.2 k 3.1 in young GH-deficient vs. -0.5 rt 3.0 cm2;P = 0.015vs. placebo) and quadricepsmus- ciency, we have reported major increases in lean body cle (+4.1 t 0.8 vs. +0.4 t, 1.2 cm2; P = 0.031) measuredby mass after 6 mo of treatment with rhGH (23). Here, we computerized tomography. Strong correlations were noted be- describe the effect on skeletal muscle mass and strength tween lean body mass(measuredas total body potassium)and following rhGH treatment in adults with GH deficiency.
CUNEO, Ross C., FRANCO SALOMON, C. MARK WILES, RICHARD HESP, AND PETER H. WNKSEN. Growth hormone treatment in growth hormone-deficient adults. I. Effects on muscle mass and strength. J. Appl. Physiol. 70(Z): 688-694,1991.-
total thigh musclearea in normal and GH-deficient adults both before and after rhGH treatment. Strength of hip flexors (+1.25 t 0.27 vs. +0.25 t 0.12 z-scores; P = 0.004) and limb METHODS girdle musclesincreased (P = 0.02) in the rhGH group. We Patient selection and characteristics. Patients are the conclude that 1) rhGH increaseslean tissue and skeletal mus- same as those from a previous study (23). Eligibility critecle massin adults with human GH deficiency, 2) this suggestsa ria were as follows: 1) age 18-55 yr; 2) GH deficiency for role for GH in the regulation of body composition of adult humans, 3) the increasein strength of limb girdle musclesafter at least 12 mo; and 3) if pituitary hormone replacement rhGH treatment suggeststhat adults with GH deficiency may was required, the treatment was stable for 12 mo before have a proximal myopathy, and 4) the failure to demonstrate entry into the study. Deficiency of GH was defined as a an increasein strength in other musclegroups may require the maximal GH of 5% below the mean. Coefficients of variation for quadriceps force measured over 6 mo in two normal subjects were 2.9 and 3.3% at mean forces of 788 and 689 N, respectively. Standard deviation (SD) scores, or z-scores, were calculated from reference values obtained in 30 healthy males (ages 17-66 yr) and 30 females (ages 19-66 yr) using identical techniques (27) as follows: z-score = (observed force - mean reference force) /SD of reference population. z-Scores allow averaging of several muscle groups and easy comparison with normal data. For example, a z-score of -2.0 represents a value 2 SD below the predicted mean, at the limit of the acceptable normal range. For hip abduction, where normal data were only available for tests performed at the knee, values were
AND
STRENGTH
689
converted using data obtained at both the knee and the ankle in a subset of the patient population. CT of thigh. With the patient in the supine position, a single scan was obtained midway between the greater trochanter and the joint space of the knee, as assessed by a scout film. Slice width was 8 mm, using 125-kV and 21O-mA settings on a Somatom DRH scanner (Seimens, FRG). Scan time was 4 s. The image was recorded on magnetic tape until read. Scans were performed with a lead shield over the pelvis. The gonadal dose was measured in one male (320 and 105 @v from the topogram and tomogram, respectively), and in one female the gonadal dose without the shield was 235 and 45 @v under the shield. Because beam-hardening artifacts in a horizontal plane between the femurs were apparent, preliminary measurements in normal and GH-deficient patients established that muscle and fat in the whole thigh could best be defined using Hounsfield numbers 20-100 and -150 to 0, respectively. In patients without acromegaly, minimal skin area was recorded at these settings. Total cross-sectional areas of muscle were measured in the dominant thigh. By use of a magnetic pencil to outline the muscle planes, quadriceps area was also recorded, with care taken to exclude the neurovascular bundle, the sartorius, and the head of the biceps femoris. Errors due to differences in alignment of the limb with the scanning beam were minimized by correcting the areas of muscle and fat in scans performed at 3 and 6 mo for any change in bone area that differed from that measured at the initial scan, assuming no change in bone area over 6 mo. Quadriceps density and intramuscular fat were estimated in as large an area of quadriceps as possible, anterior to the beam-hardening artifact, with care to avoid obvious fascial fat deposits. Coefficients of variation for measurements based on paired observations on up to 48 scans for total muscle area, quadriceps area, fat area, bone area (Hounsfield nos. 200-2,000), quadriceps density, and percentage of intramuscular fat were 0.4, 1.7, 0.1, 0, 0.8, and 94.7%, respectively. Other measurements. Total body potassium was measured as the naturally occurring radioisotope of potassium (40K) and converted to lean body mass in kilograms as previously reported (23). Activity was assessed as daily distance walked, averaged over 1 wk using a mechanical pedometer (Aurora, Japan). Assays for serum GH and plasma insulin-like growth factor I (IGF-I) and GH absorption profiles were performed as previously reported (23). Statistics. Results are reported as means t SE. Comparisons between groups at baseline used unpaired t tests or x2 tests, and treatment responses were compared using analysis of covariance on the 6-mo data (with baseline data as the covariate) or Mann-Whitney U test on the increment between baseline and 6 mo if nonnormality of the data was evident. Significance was recognized at the 5% level. Relationships between single variables before treatment were explored with simple linear regressions and changes in variables after treatment with multiple regressions, with treatment as a stratifying or binary variable. For changes with more than one significant association, multiple linear regression was also
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 7, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
690
GROWTH
TABLE
1. Patient characteristics rhGH
Age,
HORMONE
yr
Males /Females Ht, cm wt, kg Initial diagnosis Cushing’s disease Prolactinoma Chromophobe adenoma Craniopharyngioma Idiopathic Radiotherapy (medulloblastoma) Pituitary replacement hormones Corticosteroids Thyroxine Sex steroids dDAVP Fludrocortisone
39t3 B/4 170.0t2.7 82.5t3.9
Placebo
38t3 B/4 169.0k3.2 79.3t6.8
AND STRENGTH
only a limited number of patients were available, showed a trend toward increased strength in the rhGH group compared with the placebo group. Mean limb girdle muscle z-scores (shoulder abduction, hip flexion, and hip abduction) increased in the rhGH group (+0.94 t 0.25 vs. +0.28 t 0.16; P = 0.019). Quadriceps force per quadri180
6 0 3 3 0 0 10 11
10 3 1
3 2 3 0 3 1
10 10
170
10 10
8 9
rhGH Placebo
P< 0.02
9 10
8 2 2 I
I
Age, height, and weight are means k SE; n = lZ/group. No significant differences existed between groups before treatment.
10 10
a
9
?hGH
11 11
Placebo
used, with treatment code as a stratifying variable. Treatment-variable interaction was explored and excluded in each case. Coefficients of variation were calculated using the SD of the differences between paired observations divided by the mean of all the observations.
P< 0.03
RESULTS
Patient characteristics are shown in Table 1. Randomization resulted in evenly matched groups with respect to age, height, weight, sex, pituitary pathology, and incidence of pituitary hormone replacement. Lean body mass and CT data. As previously reported (23), the lean body mass increased over 6 mo by 5.6 t 1.1 kg (10.8% higher than baseline) in the rhGH group and decreased 0.1 t 0.9 kg in the placebo group (P < 0.001 for the differences between groups). Similarly, the total cross-sectional area of midthigh muscle in the rhGH group increased 8.1% (137.8 t 13.6 to 149.0 t 14.4 cm2) compared with a decrease of 0.4% (126.0 t 12.3 to 125.4 t 11.4 cm2) in the placebo group (P = 0.015; Fig. 1). Quadriceps area increased in the rhGH group by 5.7% (72.4 t 7.7 to 76.4 t 7.5 cm2) compared with an increase of 0.6% (65.3 t 3.2 to 65.7 t 5.7 cm2) in the placebo group (P = 0.031). Bone area on CT did not change significantly. The reduction in fat area was not statistically significant, in keeping with the skinfold thickness data, where the major reduction occurred in the abdominal region (23). No change in quadriceps density was noted between the two groups (Table 2). Although there was a tendency for the area of quadriceps on CT with a Hounsfield number consistent with fat to decrease in the rhGH group, this failed to reach statistical significance. This conclusion was not altered when those with some detectable intramuscular fat on the initial scan were examined. Muscle strength. A significant increase in the force of hip flexion was demonstrated following rhGH treatment (Table 3). All other muscle groups except knee extension and neck flexion, for which, in the latter case, data on
150
G
