| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
University of New Mexico–Health Sciences Center, Aging and Genetic Epidemiology Program, Albuquerque.
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Methods. Fasting serum IGF-1, IGFBP3, testosterone, estrone, and sex hormone binding globulin were analyzed in 48 women on hormone replacement (HRT) (unopposed oral estrogen, HRT+, 
), 135 women not on HRT (HRT-, 
), and 128 healthy men (men, 
). Total lean body mass (LBM) and total fat were measured by dual energy X-ray absorptiometry.
Results. Total LBM decreased with age in all groups (
). LBM was greater, and IGF-1, IGFBP3, and testosterone were lower in HRT+ versus HRT- women (
, 
, , and 
, respectively). LBM in men was positively related to IGF-1 () and testosterone (
), whereas LBM was associated with IGFBP3 () and total fat (
) in female HRT+ and total fat () in HRT– women. IGF-1 decreased with age in men and HRT- women () but did not decrease in HRT+ women. Total fat significantly decreased across age (
). Controlling for age and HRT, the rate of decrease in fat was slower in men versus women (). IGFBP3 decreased in all groups across age (), and the ratio of IGF-1 to IGFBP3 decreased faster in men compared to HRT+ and HRT- women ().
Conclusions. Our data indicate divergent influences of sex steroids, IGF-1, and IGFBP3 on age-related changes in LBM in healthy elderly men and women.
Although the decrease in sex steroids is associated with decreased LBM and increased fat in both sexes, unopposed oral estrogen replacement therapy (HRT) reportedly increases LBM and decreases truncal fat concomitant with increased mean 24-hour GH secretion in the face of decreased serum IGF-1 and IGFBP3 concentrations (7,9). Conversely, although testosterone replacement in older men increases muscle anabolism and GH release, it also increases IGF-1 via increased expression of intramuscular mRNA for IGF (10–12).
The purpose of this study was to further investigate sexually dimorphic and age-related differences in body composition and to determine the associations of endogenous and exogenous sex steroids, IGF-1, and IGFBP3 with changes in body composition in a group of healthy elderly adults.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
), 48 women on unopposed oral estrogen (HRT+, ), and 135 women not on HRT (HRT-, ) who were currently enrolled in the New Mexico Aging Process Study. Body composition data was limited to those subjects for whom complete body composition and hormonal data were available in 1996. Women taking combination hormone therapy were excluded. Participants in this study were nonsmokers and were independently living. Eighty percent of the men were able to perform 7 activities of daily living with no help, compared to 92% HRT+ and 88.5 % HRT- subjects, respectively. Twenty-six participants had diagnosed hypertension (12 men, 4 HRT+ women, and 10 HRT- women). Forty-four participants were on stable doses of thyroid replacement therapy (14 men, 12 HRT+ women, and 18 HRT- women). The University of New Mexico Human Research Review Committee approved the protocol, and written informed consent was obtained from all participants.
Hormone Assays
Fasting blood samples (
50 ml) were obtained by venipuncture between 7:30 and 9:00 AM, aliquoted into Nalgene cryovials, and stored at -70°C. IGF-1 was analyzed using the radioimmunoassay (RIA) acid–ethanol extraction method (Nichols Institute Diagnostics, San Juan Capistrano, CA) and was sensitive to 60 µg/l. Intra- and interassay variability for IGF-1 was 2.4% and 5.2%, respectively. IGFBP-3 was analyzed using RIA (Nichols Institute Diagnostics, San Juan Capistrano, CA), with sensitivity to 0.0625 µg/ml and intra- and interassay variability of 3.8% and 6.3%, respectively. Estrone assays were conducted by Quest Laboratories (San Juan Capistrano, CA) using extraction chromatography and I25I RIA, with a reported sensitivity of 10 pg/ml and interassay precision of 15%. Serum estrone was not measured in the men. Testosterone was assayed using a I25I-coated tube RIA (Incstar Corp., Stillwater, MN), with a sensitivity of 0.059 ng/ml. The precision of this assay ranged from 13.8% at low levels to 4% at concentrations >5 ng/ml. Sex hormone binding globulin (SHBG) was measured using RIA (Radim Group, Wein Laboratories, Succasunna, NJ), with a sensitivity of 5 nmol/l and a precision of 8%.
Body Composition
A Lunar DPX dual energy X-ray absorptiometer (DXA, Lunar Radiation Corp., Madison, WI) was used to measure percentage body fat and total LBM for all participants.
Statistical Analyses
Statistical analyses were performed using the SAS statistical package (Version 3.2; SAS Institute, Inc., Cary, NC). The distributions for estrone and SHBG were positively skewed within each sex and transformed using natural logarithms to approximate normal distribution. Correlations among variables were examined using Pearson correlation matrices within each sex and/or HRT group. Differences between the sexes in group mean values and between HRT+ and HRT- were tested using one-way analysis of variance. Analyses to determine the independent association of each variable with changes in body composition within each sex included linear regression, stepwise variable selection, and Student's t test. Regression models for women included either estrone or HRT to examine the effects of estrogen.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
; Figure 1). LBM was significantly greater in the men than in both the HRT+ and HRT- women () and was greater in the HRT+ versus the HRT- women ().
|
|
) in men than in HRT+ and HRT- women (p =.02; Figure 2). Total body fat significantly decreased with age in women regardless of HRT status (p <.05). Controlling for age and HRT, the rate of decrease in total body fat in the men was slower than in the women (p =.02).
|
; Table 1) but did not decrease significantly with age in either HRT group. Estrone was significantly greater in HRT+ versus HRT- women (
; Table 1).
IGF-1 decreased with age in men and HRT- women (p <.01) but did not significantly decrease in HRT+ women (Figure 3). Controlling for LBM, the rate of change was significantly greater in men compared to women regardless of HRT (). IGF-1 was lower in HRT+ versus HRT- women (; Table 1).
|
; Table 1) and significantly lower in HRT+ versus HRT- women (p <.05). Binding protein 3 decreased with age in all groups (p =.005; Figure 4), but there was no difference between the groups for rates of change in IGFBP3.
|
; Table 1).
Controlling for IGFBP3 and SHBG, regression analysis revealed that LBM was positively related to IGF-1 () and testosterone () in the men. There was a significant age–sex interaction for the IGF1/IGFBP3 ratio, with men demonstrating a greater decrease in the ratio across age when compared to the women. Controlling for HRT, IGFBP3, and SHBG in the women, LBM was positively related to IGFBP3 (), and total fat was negatively related to IGF-1 ( and ) and positively related to testosterone ().
Correlation Analysis
Women HRT-.--
LBM was significantly correlated to total fat (
, ) and negatively correlated to age (
, ). Total fat was negatively correlated to age (
, ) and positively correlated to estrone (
, 
) and to IGF-1 and IGFBP3 (, and 
, , respectively). IGF-1 was negatively correlated with age (
, p =.03) and positively correlated to IGFBP3 (r2 =.76, p <.001).
Women HRT+.-- LBM was positively correlated to IGFBP3 and total fat (r2 =.30, p =.04 and r2 =.58, p <.0001, respectively) and negatively correlated to age (r2 =.32, p =.02). Total fat was negatively correlated to age (r2 =.33, p =.02) and positively correlated with LBM (r2 =.58, p <.001). IGF-1 was positively correlated with IGFBP3 (r2 =.66, p <.001).
Men.-- LBM was correlated to IGF-1 and testosterone (r2 =.25, p <.001, r2 =.21, p =.02, respectively) and negatively correlated to age (r2 =.28, p <.01). Total fat was negatively correlated with testosterone (r2 =.25, p =.01). IGF-1 was negatively correlated to age (r2 =.32, p <.01) and positively correlated to LBM (r2 =.25, p <.01) and BP3 (r2 =.71, p <.001).
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The loss of LBM with aging has been reported by many investigators, and the decrease is generally greater in men than in women (13,14). We confirmed earlier reports that fat mass decreases with age in women but remains relatively stable with age in men (13,14). Age-related declines in IGF-1 concentration in women have been previously reported; however, the presence of HRT may have confounded some of the findings (5,6,8,15). HRT has been prescribed as a lifelong intervention for the prevention of osteoporosis and cardiovascular disease, although the safety concerns regarding increased incidence of breast cancer and cardiovascular events have recently been raised. Notwithstanding these current concerns, the vast majority of studies have reported positive effects of oral HRT on LBM and fat mass in postmenopausal women (4,9,16).
A caveat in many studies investigating the effects of estrogen replacement in postmenopausal women has been the use of higher dose and shorter duration therapies than are typically prescribed in long-term therapy (7). We are aware of only 2 cross-sectional studies that have addressed the question of long-term HRT therapy on body composition and the GH/IGF-1 axis in postmenopausal women (7,17). In women, the effect of estrogen on the GH/IGF-1 axis is reputed to facilitate pulsatile GH secretion and decrease IGF-1 concentrations (7,17,18). Our findings are in agreement with those of Sugimoto and colleagues (5), who reported a decrease in IGF-1 with increasing age in postmenopausal women not on HRT therapy, but our findings are dissimilar to those described in the study conducted by Goodman-Gruen and colleagues (17), who reported a linear decline in IGF-1 with increasing duration of HRT. We reported lower IGF-1 concentrations in HRT+ women, which persisted, but did not decrease, across age. Whether there was a persistent decline over time in HRT+ women could not be discerned from our study design. The initial reduction of IGF-1 with unopposed oral estrogen has been well described, and our data indicates that the initial suppression of IGF-1 production may be muted over time. Such a blunting of the decline in IGF-1 could be via an estradiol-dependent decrease in the negative feedback of IGF-1 on GH secretion and an increase in mean GH secretion, resulting in an increased drive for IGF-1 secretion (19). Although we did not measure GH concentrations in this study, increased GH secretion would be consistent with our findings and with those of others of greater levels of LBM in women on oral HRT (4,7,16). Longitudinal studies, which are designed to follow the changes in IGF-1 and GH with the initiation and continuation of HRT, are necessary to address these questions.
Age-related declines in IGF-1 concentrations have also been reported in men (2,15), although our understanding of the relationship between gonadopause and somatopause is far less complete than it is in the case of pre- or postmenopausal women. Bioavailable testosterone decreases by approximately 1.0–1.5% per year (2,15,19,20), and androgen deficiency has features similar to estrogen deficiency, including decreased muscle mass and increased central adiposity (10,11,20,21).
It has been proposed that somatopause and andropause are "bidirectionally" linked with declining androgen levels, exacerbating GH/IGF-1 decreases and decreasing GH concentrations, enhancing steroidogenic failure in Leydig cells (19). Pharmacological doses of testosterone in older men have been successful in increasing GH secretion and IGF-1 concentrations but had no effect on IGFBP3 levels (6,22). This dual control may provide some basis for our finding of a positive relationship between LBM, IGF-1, and testosterone in men.
We corroborated previous findings that HRT therapy decreased mean concentrations of IGFBP3 in women; however, the age-related decline in IGF-1 was muted. In the men, the decreasing IGFBP3 in conjunction with more rapidly decreasing IGF-1 resulted in a quicker decrease in the IGF-1/IGFBP3 ratio. GH is believed to be the primary regulator of IGFBP3 expression, although it is also produced in many nonhepatic tissues and is regulated by compounds, such as estrogen (23,24). The effects of IGFBP3 can either be IGF-1 related or independent, and the ratio of these hormones is believed to be important for IGF-1 activity (25). In addition, all IGF binding proteins have inhibitory effects on IGF-1 activity. Thus, in men, in whom IGF-1 declined more rapidly over age, IGFBP3 could have exerted an increased inhibitory effect on IGF-1 activity (26).
In women not on HRT, the ratio of IGF-1 to IGFBP3 has been reported to positively regulate fat mass (5). We reported this relationship in our HRT- women and did not find this association in HRT+ women or men. Schoen and colleagues (27) reported no relationship between adipose tissue and IGFBP3 or the ratio of IGF-1 to IGFBP3 in men or women, but they did not control for the use of HRT. Since IGFBP3 can target IGF-1 toward cell growth and away from glucose-consuming tissues, the role of IGFBP3 in the maintenance of LBM in HRT+ women and fat mass regulation in nonsupplemented women is plausible (27).
This study is limited by its cross-sectional design, and cause and effect cannot be determined. The subjects in this investigation were in very good health, had relatively high levels of functioning, and adequate nutritional intake. Thus, they may not be representative of the general aging population. Consequently, assessment of larger numbers of subjects and population studies are necessary to clarify the precise role of the GH, IGF-1, and IGFBP3 axis in the regulation of body composition in aging.
The present study demonstrated age-dependent changes in body composition in healthy elderly men and women, changes that appear to be bidirectionally influenced by sex steroids, IGF-1, and IGFBP3. Studies that are designed to uncover the mechanisms involved could be clinically important by contributing to our knowledge of hormonal influences on body composition changes in aging.
|
Acknowledgments |
|---|
Address correspondence to Debra L. Waters, PhD, University of New Mexico School of Medicine, 2701 Frontier Place NE, Surge Building Room 215, Albuquerque, NM, 87131.
Received August 30, 2002
Accepted November 18, 2002
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
