Fasting Suppresses Pulsatile Luteinizing Hormone (LH) Secretion and ...

8 downloads 61 Views 202KB Size Report
tisol, dehydroepiandrosterone sulfate, androstenedione, testosterone,. FSH, GH, and .... and free testosterone, FSH, estradiol, dehydroepiandrosterone-sulfate.
0021-972X/98/$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1998 by The Endocrine Society

Vol. 83, No. 6 Printed in U.S.A.

Fasting Suppresses Pulsatile Luteinizing Hormone (LH) Secretion and Enhances Orderliness of LH Release in Young but Not Older Men* MATTI BERGENDAHL, JOSEPH A. ALOI, ALI IRANMANESH, THOMAS M. MULLIGAN, AND JOHANNES D. VELDHUIS Departments of Pediatrics and Physiology (M.B.), University of Turku, FIN-20520 Turku, Finland; Division of Endocrinology (J.A.A., J.D.V.), Department of Internal Medicine, National Science Foundation Center for Biological Timing, University of Virginia Health Sciences Center, Charlottesville, Virginia 22908; Endocrine Section Medical Service (A.I.), Salem Veterans Affairs Medical Center, Salem, Virginia 24513; and Geriatrics Medicine (T.M.M.), Hunter Holmes McGuire Veterans Affairs Medical Center, Richmond, Virginia 23249 ABSTRACT Pulsatile gonadotropin secretion and sex-steroid concentrations are suppressed reversibly in young fasted or malnourished human subjects. In this study, we investigated the impact of age on the dynamic neuroendocrine mechanisms underlying this stress response in healthy young (age, 28 6 3 yr, n 5 8) vs. older (age 67 6 2 yr, n 5 8) men with similar body mass indices (mean, 26 6 0.6 vs. 26 6 1.3 kg/m2, respectively). Serum LH concentrations were measured by immunoradiometric assay (IRMA) in blood collected at 10-min intervals over 27 h on a control (fed) day and on the third day of a 3.5-day fast (water only) assigned in randomized order. After 24 h of basal sampling, GnRH (10 mg iv bolus) was administered to test gonadotrope responsiveness. Cortisol, dehydroepiandrosterone sulfate, androstenedione, testosterone, FSH, GH, and PRL were measured in 24-h pooled serum as positive and negative control hormones. Approximate entropy was used to quantitate the orderliness of LH release over 24 h, and a multiple-parameter deconvolution method was applied to quantify pulsatile LH secretion and LH half-life. In the fed state, older men exhibited elevated mean (24-h pooled) serum FSH and cortisol concentrations compared with young controls but equivalent serum LH concentrations and reduced serum GH, free testosterone, androstenedione, and dehydroepiandrosterone sulfate concentrations. Fed older men also manifested a lower frequency and amplitude of 24-h pulsatile LH secretion, and, by approximate entropy calculations, a more disorderly pattern of basal LH release than younger individuals. Three- and one-half days of fasting evoked 40% and 47% increases in mean (24-h) serum cortisol concentrations in young and older men, respectively (P , 0.01 vs. fed, but P 5 not significant for percentage rise in older vs. young men). Concurrently, fasting induced a 2.1-fold fall in the 24-h endogenous LH production rate in young men

Received November 21, 1997. Revision received February 12, 1998. Accepted February 24, 1998. Address all correspondence and requests for reprints to: Johannes D. Veldhuis, Division of Endocrinology and Metabolism, Department of Internal Medicine, Box 202 University of Virginia Health Sciences Center, Charlottesville, Virginia 22908. E-mail: [email protected]. * This work was supported in part by NIH Grant RR-00847 to the Clinical Research Center of the University of Virginia; NIH Reproduction Research Center P30 HD 28934; Research Career Development Award 1-KO4-HD-00634 (to J.D.V.); 1-FO5-TWO5080 from the Fogarty International Center (FIC) of the NIH (to M.B.); Veterans Administration (VA) Merit Review Medical Research Funds (to T.M.); the Baxter Healthcare Corporation, Round Lake, IL (to J.D.V.); the Academy of Finland (to M.B.); the Yrjo¨ Jahnsson Foundation (to M.B.); the Emil Aaltonen Foundation (to M.B.); the University of Virginia Pratt Foundation and Academic Enhancement Program (to J.D.V.); and the National Science Foundation (NSF) Science Center in Biological Timing (to J.D.V.). The contents of this publication are solely the responsibility of the authors and do not necessarily represent the views of the VA, FIC, NIH, or NSF.

(fed 36 6 9.7 vs. fasted 17 6 2.0 IU/L of distribution volume/day, P , 0.01), but did not significantly affect the daily LH secretion rate in older men (fed 27 6 4.5 vs. fasted 21 6 3.4 IU/day). The reduced LH production rate in fasting young men was accounted for by a 1.7-fold decline in the mass of LH secreted per burst (fed 2.5 6 0.45 vs. fasted 1.5 6 0.16 IU/L, P , 0.05), whereas LH burst mass in older men remained unchanged (and low) during fasting. In addition, in young men, during the 3.5-day fast the number of computer-resolved LH secretory bursts per 24 h decreased (fed 15 6 0.7 vs. fasted 11 6 0.7, P , 0.01), and the interburst interval increased (fed 94 6 4.2 vs. fasted 125 6 8.7 min, P , 0.05). In contrast, in older men in the fed state, basal LH peak frequency and serum free testosterone concentrations were reduced compared with corresponding values in young men, and did not decline further with fasting. Whereas the orderliness of LH release patterns increased significantly during fasting in the young men, the approximate entropy measure failed to change significantly in unfed older subjects. By cosinor analysis, young men showed lower 24-h mesor (mean of nyctohemeral rhythm of) serum LH concentrations than older volunteers during fasting. Moreover, young but not older men manifested preserved 24-h variations in LH interpulse intervals when fasting. Exogenously stimulated LH release (mean 3-h serum LH concentration or calculated mass of LH secreted) following a single iv injection of 10 mg GnRH was independent of age and fasting status. We conclude that the metabolic stressor of short-term fasting unmasks specific age-related neuroendocrine contrasts in the stress-responsive control of both the pulsatile and nyctohemeral regulation of the male hypothalamo-pituitary-gonadal-axis. (J Clin Endocrinol Metab 83: 1967–1975, 1998)

I

NADEQUATE nutritional intake for any particular level of whole-body energy utilization alters neuroendocrine activity in the healthy young human, e.g. by suppressing pulsatile activity of the reproductive axis, and stimulating various components of the stress-responsive somatotropic and corticotropic axes (1–3). Thus, short-term nutritional deprivation specifically decreases serum gonadotropin and gonadal sex-steroid hormone concentrations, while amplifying GH, ACTH, and cortisol secretion. Available mechanistic studies in the young healthy human and experimental animals suggest that reduced release of hypothalamic GnRH plays a key role in the acute fasting-associated suppression of reproductive function in both the male and female (1). Clinical investigations in young men using repeated blood sampling, LH immunoradiometric assay (IRMA), and deconvolution analysis have shown that a 3.5-day fast elicits an

1967

1968

BERGENDAHL ET AL.

approximately 50% decrease in mean (24-h) serum LH and free testosterone concentrations, with an attendant reduction in the daily LH secretion rate ascribable to two mechanisms: first, a fall in the apparent number of computer-resolved LH secretory bursts; and second, a decrease in the mass of LH secreted per burst (4). Recently, it has been shown that pulsatile iv GnRH infusions over the last 24 h of calorie withdrawal can completely prevent the fasting-induced decline in LH secretory burst mass and frequency (without altering the increase in serum cortisol concentrations), and restore serum total and free testosterone concentrations to baseline (fed) values (4). The ability of pulsatile GnRH infusions alone to overcome this form of stress-induced (in males) hypogonadism supports the pathophysiological hypothesis that nutrient withdrawal reversibly decreases output of the human hypothalamic GnRH pulse generator. Thus, the metabolic stressor of fasting in young men is a useful experimental model for short-term hypothalamic GnRH deficiency mediating hypogonadotropic hypogonadism. Healthy aging in men also results in changes in the LH pulse signal (5–13), as well as in the steroidogenic responsiveness of the testis to gonadotropin challenges (14). Using deconvolution analysis to calculate underlying gonadotropin secretion rates and LH half-life, we observed that LH secretory burst amplitude and mass decrease progressively with increasing age (15). The clinical relative hypogonadism of aging is accompanied by a decrease in serum total and free testosterone concentrations (16 –20). Some of these features of the aging GnRH-LH-testosterone axis are similar qualitatively to those of stress responses of the young male reproductive axis (above). However, whether age-related changes in the hypothalamic-pituitary-gonadal axis reflect undue susceptibility of older individuals to stress-associated inhibition of the reproductive axis is not known. Moreover, to our knowledge, there are no data available to define the relative impact of a metabolic stressor, such as acute nutrient withdrawal, on gonadotropin secretion in older vs. young men (14). In the present study, we investigated the effects of shortterm fasting on the hypothalamic-pituitary-Leydig cell axis in eight healthy young and eight older men. The metabolic stressor of acute fasting was used as an experimental paradigm to investigate the hypothesis that age alters the ability of the dynamic (pulsatile and nyctohemerally rhythmic) reproductive axis to respond to an inhibitory stressor. Selected other hormones were measured in 24-h serum pools as positive and negative controls. Subjects and Methods Clinical protocol Eight young and eight older healthy men [body mass indices (BMIs) 26 6 0.9 kg/m2, age 28 6 3 yr, range 22– 44 yr in the group of young men; and BMI 26 6 1.3 kg/m2, age 67 6 2 yr, range 55–73 yr in older men] were studied after provision of written informed consent approved by the Human Investigation Committee of the University of Virginia. No volunteer was taking medications, had undertaken transmeridian travel for at least 1 week, or had recent weight loss. Each had a negative detailed clinical history and physical examination with normal adult sexual maturation and testicular size, normal screening biochemical tests of renal, hepatic, metabolic, and hematological function, and unremarkable (age-adjusted) morning serum concentrations of total and

JCE & M • 1998 Vol 83 • No 6

free T4, TSH, GH, PRL, estradiol, free and total testosterone, immunoreactive LH and FSH, and insulin growth factor-I (IGF-1). The pulsatile LH data in six of the eight young controls were reported earlier (4). BMI (but not visceral adiposity) was matched in the two study groups. The volunteers were admitted to the General Clinical Research Center of the University of Virginia the night before blood sampling both in the fed state and again before 3.5 days of fasting. The fed and fasting admissions were assigned in randomized order at least 1 month apart. In both the fed and fasting studies, blood sampling was carried out at 10-min intervals for 27 h beginning at 0800 h at least 1 h after venipuncture. Daytime naps were disallowed. In the 3.5-day fast, the 27-h blood sampling interval occupied hours 56 – 83 (hour zero defined as 2400 h on the evening of admission). After blood sampling for 24 h, a single bolus of 10 mg GnRH was given iv to test pituitary responsiveness as assessed by 3 more h of 10-min blood withdrawal. Blood samples were allowed to clot at room temperature, centrifuged, and the subsequent sera frozen at 2C for later assays. Subjects remained in a bed or chair during sampling except for ambulation to the lavatory as needed. In the fed state, three isocaloric meals were given per day (at 0800, 1200, and 1730 h). During the 3.5-day fast, the volunteers received caffeine- and calorie-free liquids only, slept in the Clinical Research Center, and had urinary ketones monitored twice to four times daily to assess compliance with the fast. All patients had consistently positive urinary ketones throughout the fast. Potassium chloride (40 mmol) and water-soluble vitamins were administered orally daily, as described earlier in other studies (2, 3, 21, 22)

Assays Serum LH concentrations were measured robotically in each sample in duplicate by a two-site IRMA (Nichols Labs., San Juan Capistrano, CA), as described previously (15). This assay correlates well (P , 0.001) with an in vitro Leydig cell LH bioassay over the range of LH concentrations 2–50 IU/L. The median inter- and intraassay coefficients of variation were less than 8.5% for these studies. All 181 samples in each admission were assayed together. The sensitivity of the assay was 0.20 IU/L, using the First International Reference Preparation. Serum total and free testosterone, FSH, estradiol, dehydroepiandrosterone-sulfate (DHEA-S), androstenedione, PRL, GH, IGF-1, IGF binding protein-3 (IGFBP-3), and TSH were assayed by RIA or chemiluminescent or immunoradiometric assays, in a single 24-h pool of serum (50 mL aliquoted from each of 145 samples) (2, 3, 15, 21, 23, 24).

Deconvolution analysis Deconvolution analysis is a mathematical technique applied to a pulsatile serum hormone concentration vs. time series to estimate subject-specific measures of pulsatile hormone secretion and half-life (25– 27). The daily LH secretion rate was computed assuming negligible basal LH secretion as the product of secretory burst frequency and the mean mass of LH released per secretory pulse. Based on recent validation studies in men, deconvolution analysis was carried out at 95% joint statistical confidence intervals for all calculated LH secretory burst amplitudes with the technician blinded to the randomized order of the fed vs. fasted admissions. After deconvolving the entire 27-h time series of serum LH concentrations, statistical analysis was applied separately to the 24-h baseline (spontaneous pulsatile LH release) and the 3-h postGnRH (stimulated) segments.

Nyctohemeral (24-h) rhythmicity Diurnal rhythms of serum LH concentrations as well as computed LH secretory burst characteristics (mass per burst and interpulse interval) were appraised using cosinor analysis, as described previously (21).

Approximate entropy (ApEn) ApEn is a statistic that quantitates relative orderliness or regularity of hormone release profiles. It complements usual pulse analysis, but gives information about (sub)pattern recurrence or repetition within the data. Pattern reproducibility is lost in tumoral hormone secretion and reduced in aging. Loss of feedback control also is expected to elicit more disorderly or irregular release patterns.

FASTING AND LH IN YOUNGER VS. OLDER MEN ApEn comprises a family of model- and concentration-independent statistics for assessing the apparent process randomness or serial irregularity of a time series by quantifying the subpattern reproducibility not necessarily identified by pulse-detection algorithms (28). A particular ApEn statistic is a single finite nonnegative real number assigned as an ensemble measure to a series of hormone concentrations with larger ApEn values corresponding to relatively greater pattern randomness. Specifically, ApEn measures serial data regularity or, technically, the logarithmic likelihood that runs of patterns (of length m) that are similar (within r) remain similar on next incremental comparison. The formal definition of ApEn is given elsewhere (28). Two principal input parameters, namely m and r, are fixed to compute ApEn from vector sequences constructed from the observed hormone concentration profiles, where, m represents the window length of consecutive pattern measurements, and r the tolerance or threshold for testing subpattern regularity. To maintain scale invariance, r is typically fixed as a percentage of the total (between-sample) sd of each hormone time series, e.g. 20%, and m as a value of 1 or 2 denoting consecutive vectors of length 1 or length 2 data points. In the present study, given 145 measurements of LH in each 24-h time series, we calculated ApEn values with r 5 0.2 and m 5 1, which provides the more appropriate statistic for assessing subpattern reproducibility in data series of this size.

Statistical analyses Differences between fed and fasted measures in young and older men were assessed after logarithmic transformation using ANOVA followed by Duncan’s multiple range test. Results are presented as the mean 6 sem. Statistical significance was accepted for a P value , 0.05 or for nonoverlapping group 95% statistical confidence intervals (cosinor analysis).

Results Mean serum hormone concentrations

The 3.5-day fast in young men (n 5 8) resulted in a more than 2-fold decrease in the mean (24-h) serum LH concentration (IU/L, averaged over the 145 samples collected). Older men showed no significant suppression (Table 1). In particular, the mean (6 sem) fed vs. fasting serum LH concentrations were: in young men, 3.5 6 1.0 vs. 1.6 6 0.16 IU/L [fasting value P , 0.05 vs. young fed, and P 5 not significant (NS) vs. older fed]; and, in older men, 4.2 6 0.69 vs. 3.1 6 0.47 IU/L (fasting value P , 0.01 vs. young fasting). Other measurements were made in 24-h serum pools. Pooled serum concentrations of total and free testosterone decreased, respectively, by 40% and 46% in fasting young men (P , 0.01) (Table 1). The corresponding androgen values were not changed by fasting in older men, despite similar statistical variances. Baseline (24-h pooled) serum free testosterone concentrations were lower, and pooled serum FSH and cor-

1969

tisol concentrations higher, in fed older vs. young men (Table 1). Serum pooled concentrations of cortisol increased significantly and similarly during fasting in young and older men (specifically by 40% and 47%, respectively, P , 0.01 vs. fed, Table 1 and P 5 NS young vs. older for the percentage rises). Serum pooled estradiol and IGFBP-3 concentrations were statistically independent of age, whereas GH and DHEA-S levels were lower (both P , 0.01) at baseline (fed) in the older cohort (Table 1). Deconvolution analysis of pulsatile LH secretion

Illustrative 24-h profiles of serum LH concentrations, which were pulsatile in all subjects in both the fed and fasting studies, and deconvolution-resolved LH secretory rates from one young and one older male are depicted in Fig. 1. The specific quantitative changes induced by fasting in various pulsatile attributes of LH secretion, as well as LH half-life, are summarized in Table 2. Statistical analyses revealed that the half-duration in minutes (the duration of the calculated secretory event at half-maximal amplitude) of computed LH secretory bursts decreased, but the calculated LH half-life did not change, significantly in response to fasting in young men. In older men, the LH secretory burst half-duration and LH half-life did not change during fasting, but both of these parameters were greater at baseline and during fasting than the corresponding values in young men. Individual half-life data are given in Fig. 2A. The number of LH secretory pulses per 24 h with statistically non-zero amplitudes (jointly at P , 0.05) fell significantly in fasting young men. In contrast, LH secretory burst frequency was lower at baseline in fed older men, and remained unchanged during fasting (Fig. 2B). Conversely, the mean LH interpulse interval in young men rose in response to fasting. In older men, the LH interburst interval was not changed by fasting, and was greater in both fed and fasted states than corresponding values in fed young men. The mass of LH secreted per burst (area of the calculated LH secretory pulse) in young men decreased significantly during fasting, but this change was not evident in older men. There were no significant alterations in computed LH secretory burst amplitude (maximal rate of calculated LH secretion attained within a release episode) in young or older men during fasting, but the LH secretory pulse amplitude was lower in older men in both dietary states compared with fed young men. The product of the mass of LH secreted per burst and LH

TABLE 1. Mean (24-h) serum hormone concentrations (pools) in fed and fasting young and older men (n 5 8 in each group) Hormone

Young fed

Young fasting

Older fed

Older fasting

FSH (IU/L) Total testosterone (ng/dL) (to convert to nmol/L 3 0.0347) Free testosterone (pg/mL) (to convert to pmol/L 3 3.467) Estradiol (pg/mL) (to convert to pmol/L 3 3.671) Androstenedione (ng/mL) (to convert to nmol/L 3 3.49) Cortisol (mg/dL) (to convert to nmol/L 3 27.6) DHEAS (mg/mL) (to convert to mmol/L 3 2.714) GH (mg/L) IGF-1 (mg/L) IGFBP-3 (mg/L)

2.4 6 0.46 581 6 70 23 6 4.7 32 6 6.0 1.1 6 0.04 8.9 6 0.71 484 6 63 1.7 6 1.19 242 6 38 3249 6 273

1.8 6 0.18 343 6 42a 13 6 1.0a 26 6 4.5 1.1 6 0.06 12.5 6 1.10a 654 6 39 3.8 6 0.76a 221 6 45 3258 6 287

6.1 6 1.40 512 6 63 13 6 1.0a 34 6 3.3 0.8 6 0.12b 11.4 6 0.59b 98 6 17a 0.4 6 0.10a 131 6 22 2778 6 127 a

4.7 6 1.0b 425 6 61 11 6 1.2a 30 6 5.3 1.0 6 0.17 16.8 6 1.04c 130 6 31a 2.3 6 0.38d 97 6 17b 2640 6 128

LH was determined in each of 169 samples in each volunteer, with 24-h means calculated from first 145 samples (pre-GnRH injections). Data are mean 6 SEM, n 5 8 (ANOVA followed by Duncan’s multiple range test after logarithmic transformation). a P , 0.01; b P , 0.05, compared with young fed controls; c P , 0.01; d P , 0.05, compared with older fed controls.

1970

JCE & M • 1998 Vol 83 • No 6

BERGENDAHL ET AL.

FIG. 1. Illustrative 24-h serum LH concentration (IRMA) profiles and calculated LH secretory rates as determined by deconvolution analysis in one healthy young man and one older man studied in fed and fasted states. Blood samples were collected at 10-min intervals for 27 h when volunteers were nutritionally replete (fed) and again during last 27 h of a 3.5-day fast. Fed and fasted sessions were assigned in randomized order at least 4 weeks apart. GnRH (10 mg iv) was administered after first 24 h of baseline sampling. Baseline is shown over time 0 –1440 min (A and B), and post-GNRH responses over time 1440 –1640 min (C and D). Continuous curves through observed serum LH concentrations (A and C) are predicted by deconvolution analysis (see Subjects and Methods). Vertical bars through sample LH concentration values denote dose-dependent intrasample SDS (estimated from all 191 replicated samples in each original 27-h time series). Punctuated episodes of spontaneous LH release over 24 h or following GnRH iv (B and D) are computer-estimated LH secretory bursts (P , 0.05 vs. random sample variance), which give rise to pulsatile serum LH concentration profiles. Calculated mean number, duration, mass, and amplitude of LH secretory bursts and half-life of endogenous LH in eight young and eight older subjects each studied in both fed and fasted states are summarized in Table 2. Table 3 gives LH responses to iv GnRH (10 mg). TABLE 2. Deconvolution analysis of (24-h) pulsatile LH secretion and half-life in fed and fasting states in healthy young and older men (n 5 8 in each group) Parameter

Young fed

Young fasted

Older fed

Older fasted

LH secretory burst (half-duration (min) LH half-life (min) LH secretory burst frequency (events/24 h) Interburst interval (min) LH secretory burst mass (IU/L) LH secretory burst amplitude (IU/L/min) LH production rate/24 h (IU/L distribution volume/day)

7.8 6 1.53 84 6 2.8 15 6 0.71 94 6 4.2 2.5 6 0.45 0.67 6 0.23 36 6 9.7

4.4 6 1.05b 91 6 6.0 11 6 0.73a 125 6 8.7b 1.5 6 0.16b 0.41 6 0.07 17 6 2.0b

12.4 6 1.80b 149 6 9.7a,d 12 6 0.42a 119 6 3.9a 2.3 6 0.31 0.21 6 0.05b 27 6 4.5

10.1 6 1.44b 146 6 8.8a,d 12 6 0.75a 118 6 9.2a 1.81 6 0.23 0.22 6 0.06b 21 6 3.4

Data are mean 6 SEM, n 5 8 (ANOVA followed by Duncan’s multiple range test after logarithmic transformation). a P , 0.01; b P , 0.05, compared with young fed controls; c P , 0.05, compared with older fed controls; d P , 0.01; e P , 0.05, compared with young fasted.

secretory event frequency is the calculated 24-h pulsatile LH secretion rate. In young but not older men, this measure decreased in the fasted state. Calculated LH secretory burst mass and the mean (3-h) serum LH concentration following a single iv bolus injection of 10 mg GnRH were both independent of fasting or age (Table 3).

Nyctohemeral rhythmicity (cosinor analysis)

As shown in Fig. 3, A and B, young but not older men exhibited significant 24-h variations in individual LH secretory pulses mass in the fed state. Fasting reduced the mesor (cosine mean) and amplitude of this rhythmicity in young men but evoked a detectable rhythm in older men. Both

FASTING AND LH IN YOUNGER VS. OLDER MEN

FIG. 2. Individual LH half-life (A) and LH secretory burst frequency (B) values in eight young and eight older men studied in fed vs. fasting state for 3.5 days. Blood was sampled at 10-min intervals for 24 h and assayed for LH content by IRMA. Deconvolution analysis was applied to quantitate various LH secretory measures (Table 2) and half-life. Numerical values are mean 6 SEM.

young and older men exhibited a significant 24-h variation in LH interpulse interval in the fed state, which was abolished in older men by fasting (Table 4). As summarized in Table 5, cosinor analysis of serum LH concentrations per se (rather than the above deconvolutioncalculated parameters) revealed significant age-related contrasts only in the fasting state, in which the mesor (mean) reached a significantly lower value in young vs. older volunteers. ApEn

ApEn averaged 1.21 6 0.055 in fed young males, and fell to 0.823 6 0.089 during fasting (P , 0.01) (Fig. 4). This change indicates significantly greater orderliness or regularity of LH release (lower ApEn value) in the fasting environment. Baseline ApEn tended to be higher in older men at 1.43 6 0.054, denoting greater irregularity of LH release over 24 h, as observed earlier in overnight blood sampling (6). In contrast to young men, older men’s ApEn values remained statistically unchanged during fasting (1.16 6 0.074). Discussion

The present investigation refutes our a priori hypothesis of greater vulnerability of older men to the gonadotropin-sup-

1971

pressing effects of short-term fasting. Indeed, fasting reduced the secretion rate of LH and the serum free testosterone concentration less in healthy older men. Moreover, fasting enhanced the orderliness of LH release patterns (as defined by the ApEn statistic) in young but not older men. In contrast, young and older men had comparable adrenalgland activation during fasting with, respectively, 40% and 47% (young and older) rises in mean (24-h) serum cortisol concentrations. This suggests a similar adrenal stress response (at least fractionally in young and older individuals). As a positive control, both age groups showed severalfold increases in (24-h pooled) serum GH concentrations in response to fasting. Thus, the age distinction in neuroendocrine adaptations to fasting was quite specific to the hypothalamopituitary-gonadal axis. Using deconvolution analysis, we reported earlier that a 3.5-day fast in young men brings about an approximately 50% fall in the calculated 24-h LH secretion rate. This was neuroendocrinologically caused by a decline in both the apparent number of computer-resolved LH secretory bursts and the mass of LH secreted per burst (4). The suppressive effect of fasting was completely reversed by pulsatile iv infusions of GnRH. In the present study, we now observe that in healthy older men short-term fasting fails to alter significantly the 24-h LH production rate, LH pulse frequency, or the mass of LH secreted per burst. Under stringent deconvolution-fitting conditions of 95% joint confidence intervals, baseline LH secretory burst frequency and amplitude (mass) were both reduced at baseline in older men [the latter confirming an evaluation earlier by 2.5-min blood sampling overnight (29)], and remained low during fasting. The lower (baseline) serum free testosterone concentration in older men would tend to drive LH release (assuming normal negative feedback), which might have opposed the tendency of fasting to suppress the (older) axis. Older men also were distinguished by more irregular or disorderly patterns of 24-h LH release at baseline, reflected in higher ApEn values [as also described independently in 6 – 8 h of overnight blood sampling (6)]. Moreover, in older men, fasting failed to elicit significantly more orderly (lower ApEn) LH release profiles, unlike the LH-gonadal-axis responses identified in young men under the same study conditions. The capacity of gonadotroph cells to augment secretion of biologically active LH during blockade of estrogen negative feedback is decreased in older men (30, 31). On the other hand, the secretion of bioactive LH in older men can be amplified by treatment with a nonsteroidal androgen-receptor antagonist (24). The latter finding indicates significantly preserved GnRH/LH secretory capacity in older individuals. Taken together with other data, such observations allow for the possibility (but do not prove) that aging is marked by alterations in sex-hormone feedback control of the GnRHLH-testosterone axis (32–35). In addition, we now describe an apparent resistance to fasting-induced suppression of both pulsatile LH secretion and circulating (free) testosterone concentrations in older men. This is consistent with an ageassociated difference in adaptation of the GnRH-LH-Leydig cell axis to a short-term metabolic stress. Whether this putative resistance to fasting stress is related to the relative

1972

JCE & M • 1998 Vol 83 • No 6

BERGENDAHL ET AL.

TABLE 3. Deconvolution analysis of GnRH-stimulated LH secretion over 3 h immediately following a single iv bolus injection of 10 mg GnRH in fed and fasting states in healthy young and older men LH measure

Young fed

Young fasted

Older fed

Older fasted

LH secretory burst mass (IU/L/burst) 3-h mean serum LH concentration (IU/L) 3-h integrated LH level (IU/L 3 min)

6.5 6 0.73 6.0 6 0.53 1140 6 98

15 6 4.1 8.7 6 1.9 1650 6 360

9.6 6 2.1 9.1 6 1.5 1660 6 282

13 6 2.6 10 6 2.0 1870 6 360

GnRH-stimulated LH release significantly above preinjection baseline (or 24-h mean) in all groups studied. Data are mean 6 SEM, n 5 8/group (ANOVA followed by Duncan’s multiple range test after logarithmic transformation revealed no significant effects of age or dietary state on the measures shown).

inability of older men to increase pulsatile LH secretion during treatment with an opiate-receptor antagonist (specifically used to block opiate-dependent and hence putatively stressmediated inhibition of GnRH secretion) (36) is not known. In the rat, sex steroids significantly modulate the inhibitory actions of both endogenous opiates and the stress of food restriction on pulsatile LH release (37–39). The roles (if any) of other putative hypothalamo-pituitary stress modulators defined in the rat, such as CRH, neuropeptide Y, leptin, etc. (1, 40, 41), in fasting-induced hypogonadotropism in men are unestablished. In addition, whether and how estradiol, testosterone, and/or endogenous opiate pathways modulate fasting’s suppressive effects on GnRH secretion and/or its feedback control in the young or older human are not known. Because the type, duration, intensity, and novelty of various stressors may influence the magnitude and nature of the subsequent stress-adaptive responses, the present findings of age-related contrasts in GnRH/LH secretory adaptations to a particular metabolic-stress paradigm may or may not be applicable in other stress contexts. Fasting did not affect the calculated half-life of endogenous LH in either group of men, but older men had an apparently longer LH half-life at baseline and during fasting. A longer half-life in older men could be because of reduced LH removal, a larger LH distribution volume, altered LH isoforms [e.g. more acidic (14) with reduced metabolic clearance], and/or, on a technical basis, (unrecognized) basal LH secretion or a greatly skewed LH waveform (42)]. We know of no direct experimental data to distinguish among these possibilities. Even so, fasting did not change the apparent half-life of LH. As reported previously using a simplified model of purely pulsatile LH release (15), we found a significant prolongation of the computed LH secretory-burst duration in older men. LH secretory-burst duration also is altered in end-stage renal failure in men (23), and increases in estrogen-treated postmenopausal women (43). The intrapituitary and/or extraglandular mechanisms underlying a prolonged (LH) secretory event duration are not known. In this study, renal function and serum estradiol concentrations were normal and similar in both age groups (Table 1). On the other hand, the significantly lower mean serum free testosterone concentration in the older men could explain in part their calculated longer LH half-life and/or LH secretory burst duration, because short-term androgen deprivation with flutamide will induce both of these changes (44). Older men had more disorderly LH release than young individuals, as quantified by ApEn. This was reported previously based on overnight blood sampling every 2.5 min in another group of older (vs. young) men (6). In this study, via

10-min sampling over 24-h, we found that ApEn values fall in young men during fasting, signifying more orderly or regular LH release at this time. In contrast, older men retained more disorderly (24-h) LH release. On mathematical grounds, the orderliness of hormone release is believed to reflect the strength and/or complexity of key feedback interactions within a neuroendocrine axis (6, 45). Thus, our observations point to reduced feedback organization within the older (male) hypothalamo-pituitary-Leydig cell axis at baseline [present data and (6)], which fails to normalize with fasting. The suppression of LH secretory burst mass (and its 24-h rhythmicity, Fig. 3A) in fasting young but not older men was not attributable to measurable differences in pituitary gonadotrope-cell responsiveness to GnRH injections in the two age groups. Indeed, 10 mg GnRH iv stimulated similar LH release independently of age or fasting. Recent iv GnRH dose-LH secretory response analyses in (fed) young and older men revealed enhanced maximal stimulatory effects of GnRH (greater stimulus efficacy) in older subjects, with similar half-maximal GnRH actions (similar agonist potency of, or pituitary sensitivity to, GnRH) across age (46). Because GnRH dose-LH response curves are not available in fasted young and older individuals, we cannot exclude unequal sensitivity to GnRH in fasting aged vs. young volunteers. Masking of results by greater visceral fat [possibly suppressive of LH pulse amplitude (15)] in older men also cannot be excluded, because volunteers were matched for BMI only. The older men in this study exhibited lower basal (fed) serum GH, androstenedione, and DHEA-S concentrations, and higher serum FSH and cortisol concentrations in 24-h serum pools as expected, indicating representativeness of these cohorts (14). In both older and young men, mean (24-h) serum GH and cortisol concentrations rose significantly with fasting corroborating prior data in young men and documenting compliance with the fast (2, 3). Nyctohemeral (24-h cosinor) rhythms in serum LH concentrations showed reduced mesor (mean) values in fasted young compared with older men, indicating suppressed overall 24-h LH release. Separate cosinor analyses of (deconvolution-computed) LH secretory burst mass disclosed loss of 24-h rhythmic variation in young fasted (but not older fasted) men, thus unmasking another age contrast. Both age groups exhibited day-night variations in (deconvolution-calculated) LH interburst intervals in the fed state, which were abolished in older (but not young) fasted individuals. Although the exact (presumptively neural) mechanisms that generate such 24-h rhythmicities within the human LH axis are not yet known (14), in this study we document age differences both basally and fasting. Our LH data comple-

FASTING AND LH IN YOUNGER VS. OLDER MEN

1973

FIG. 3. Nyctohemeral (24-h) rhythmicity of deconvolution-calculated LH secretory burst mass (A) and LH interpulse intervals (B) for all subjects combined in each of two groups, specifically young (n 5 8) vs. older (n 5 8) men, as assessed by cosinor analysis of deconvolution measures (Subjects and Methods). NS, Not significant (P , 0.05) amplitude of 24-h periodic fit. Fitted amplitudes (and 95% statistically confidence intervals) are noted for significant group rhythms.

1974

JCE & M • 1998 Vol 83 • No 6

BERGENDAHL ET AL.

TABLE 4. Cosinor analysis of deconvolution measures in fed vs. fasting young versus older men

LH secretory burst mass (IU/L) Fed young Fed older Fasting young Fasting older LH interburst interval (min) Fed young Fed older Fasting young Fasting older

Amplitude

Phasea

Mesorb

0.50 (0.02– 0.98)c NS NS 0.49 (0.057– 0.93)c

34 (2186 –261)c NS NS 136 (62–340)c

2.6 (2.3–2.9)c 2.3 (2.0 –2.6)c 1.6 (1.4 –1.9)c 1.9 (1.6 –2.2)c

12 (3.2–21)c 11 (0.77–22)c 21 (1.3– 40)c NS

292 (110 – 474)c 352 (101– 620)c 644 (437– 836)c NS

94 (87–101)c 119 (111–127)c 114 (102–127)c 115 (103–126)c

a

Acrophae, Time in min (before 0800) of the maximum of 24-h rhythm. Mesor or mean of 24-h rhythm. 95% group confidence interval for each parameter. NS, Significant amplitude, and hence no determinable phase. b c

TABLE 5. Serum 24-h LH concentration profiles (not fitted parameters of secretion per se): cosine fits of 24-h rhythmicity Group

Amplitude

Phase

Mesor

Young fed Older fed P Young fasting Older fasting P

0.77 6 0.350 0.49 6 0.056 NS* 0.43 6 0.048 0.37 6 0.054 NS

740 6 172 1030 6 141 NS 780 6 160 510 6 140 NS

2.9 6 0.96 3.9 6 0.66 NS 1.0 6 0.16 2.6 6 0.41 P , 0.01

Means 6 SEM of individual fits, n 5 8 young, and n 5 9 older men (Wilcoxon unpaired two-tailed test: NS, denotes P $ 0.05.) a One-half the difference between zenith and nadir values. b Time in minutes (before 0800 h) of maximum value in 24-h rhythm.

FIG. 4. Individual ApEn values of 24-h serum LH concentration profiles in fed vs. fasted young (n 5 8) vs. older (n 5 8) men. ApEn is a relative measure of disorderliness or irregularity of hormone release, with higher absolute values denoting greater irregularity. Conversely, lower ApEn values as observed in this study in young fasted men indicate more orderly patterns. Data are presented as indicated in legend of Fig. 2.

ment an earlier report of diminished testosterone rhythmicity over 24 h in (fed) older men (47). In summary, fasting in young but not older men reduces 24-h LH secretory burst frequency and mass and (24-h pooled) serum free testosterone concentrations. Fasting in young but not older men enhances the quantifiable orderliness (lower ApEn) of LH release, and alters 24-h LH rhythmicity. Thus, we conclude that the metabolic stress of short-

term fasting unmasks age-related dynamic differences in the pulsatile, 24-h rhythmic, and orderly release of LH. Acknowledgments We thank Patsy Craig for her skillful preparation of the manuscript; Paula Azimi for her data analysis, statistical assistance, and artwork; Brenda Grisso for performance of the immunoassays; and Sandra Jackson and the expert nursing staff at the University of Virginia General Clinical Research Center for conduct of the research protocols.

References 1. Bergendahl M, Veldhuis JD. 1995 Altered pulsatile gonadotropin signaling in nutritional deficiency in the male. Trends Endocrinol Metab. 6:145–159. 2. Bergendahl M, Vance ML, Iranmanesh A, Thorner MO, Veldhuis JD. 1996 Fasting as a metabolic stress paradigm selectively amplifies cortisol secretory burst mass and delays the time of maximal nyctohemeral cortisol concentrations in healthy men. J Clin Endocrinol Metab. 81:692– 699. 3. Hartman ML, Veldhuis JD, Johnson ML, et al. 1992 Augmented growth hormone (GH) secretory burst frequency and amplitude mediate enhanced GH secretion during a two day fast in normal men. J Clin Endocrinol Metab. 74:757–765. 4. Aloi JA, Bergendahl M, Iranmanesh A, Veldhuis JD. 1997 Pulsatile intravenous gonadotropin-releasing hormone administration averts fasting-induced hypogonadotropism and hypoandrogenemia in healthy, normal-weight men. J Clin Endocrinol Metab. 82:1543–1548. 5. Mitchell R, Hollis S, Rothwell C, Robertson WR. 1995 Age-related changes in the pituitary-testicular axis in normal men; lower serum testosterone results from decreased bioactive LH drive. Clin Endocrinol (Oxf). 42:501–507. 6. Pincus SM, Mulligan T, Iranmanesh A, Gheorghiu S, Godschalk M, Veldhuis JD. 1996 Older males secrete luteinizing hormone and testosterone more irregularly, and jointly more asynchronously, than younger males: dual novel facets. Proc Natl Acad Sci USA. 93:14100 –14105. 7. Pincus SM, Veldhuis JD, Mulligan T, Iranmanesh A, Evans WS. 1997 Effects of age on the irregularity of LH and FSH serum concentrations in women and men. Am J Physiol. 273:E989 –E995. 8. Mulligan T, Delemarre-van de Waal HA, Johnson ML, Veldhuis JD. 1994 Validation of deconvolution analysis of LH secretion and half-life. Am J Physiol. 267:R202–R211. 9. Baker HWG, Burger HG, DeKretser DM, et al. 1976 Changes in the pituitarytesticular system with age. Clin Endocrinol (Oxf). 5:349-358. 10. Harman SM, Tsitouras PD, Costa PT, Blackman MR. 1982 Reproductive hormones in aging men. II. Basal pituitary gonadotropins and gonadotropin responses to luteinizing hormone-releasing hormone. J Clin Endocrinol Metab. 54:547–551. 11. Korenman SG, Morley JE, Mooradian AD, et al. 1990 Secondary hypogonadism in older men: its relation to impotence. J Clin Endocrinol Metab. 71:963–969. 12. Vermeulen A, Deslypere JP, De Meirleir K. 1989 A new look at the andropause: altered function of the gonadotrophs. J Steroid Biochem. 32:163–165. 13. Bergendahl M, Evans WS, Veldhuis JD. 1996 Current concepts on ultradian rhythms of luteinizing hormone secretion in the human. Human Reproduction Update. 2:507–518. 14. Veldhuis JD. 1991 The hypothalamic pituitary-testicular axis. In: Yen SSC, Jaffe RB, eds. Reproductive endocrinology. Philadelphia, PA: W.B. Saunders; 409 – 459. 15. Veldhuis JD, Urban RJ, Lizarralde G, Johnson ML, Iranmanesh A. 1992 Attenuation of luteinizing hormone secretory burst amplitude is a proximate

FASTING AND LH IN YOUNGER VS. OLDER MEN

16. 17. 18. 19. 20.

21.

22. 23.

24. 25.

26. 27. 28. 29.

30. 31. 32.

33.

basis for the hypoandrogenism of healthy aging in men. J Clin Endocrinol Metab. 75:52–58. Gray A, Berlin JA, McKinlay JB, Longcope C. 1991 An examination of research design effects on the association of testosterone and male aging: results of a meta-analysis. J Clin Epidemiol. 44:671– 684. Harman SM, Tsitouras PD. 1980 Reproductive hormones in aging men. I. Measurement of sex steroids, basal luteinizing hormone, and Leydig cell response to human chorionic gonadotropin. J Clin Endocrinol Metab. 51:35-43. Nahoul K, Roger M. 1990 Age-related decline of plasma bioavailable testosterone in adult men. J Steroid Biochem. 35:293–299. Tenover JL. 1997 Testosterone in the aging male. J Androl. 18:103–106. Winters SJ, Troen P. 1982 Episodic luteinizing hormone (LH) secretion and the response of LH and follicle-stimulating hormone to LH-releasing hormone in aged men: evidence for coexistent primary testicular insufficiency and an impairment in gonadotropin secretion. J Clin Endocrinol Metab. 55:560 –565. Veldhuis JD, Iranmanesh A, Evans WS, Lizarralde G, Thorner MO, Vance ML. 1993 Amplitude suppression of the pulsatile mode of immunoradiometric LH release in fasting-induced hypoandrogenemia in normal men. J Clin Endocrinol Metab. 76:587–593. Ho KY, Veldhuis JD, Johnson ML, et al. 1988 Fasting enhances growth hormone secretion and amplifies the complex rhythms of growth hormone secretion in man. J Clin Invest. 81:968 –975. Veldhuis JD, Wilkowski MJ, Urban RJ, Lizarralde G, Iranmanesh A, Bolton WK. 1993 Evidence for attenuation of hypothalamic GnRH impulse strength with preservation of gonadotropin-releasing hormone (GnRH) pulse frequency in men with chronic renal failure. J Clin Endocrinol Metab. 76:648 – 654. Veldhuis JD, Urban RJ, Dufau ML. 1994 Differential responses of biologically active LH secretion in older vs. young men to interruption of androgen negative feedback. J Clin Endocrinol Metab. 79:1763–1770. Veldhuis JD, Carlson ML, Johnson ML. 1987 The pituitary gland secretes in bursts: appraising the nature of glandular secretory impulses by simultaneous multiple-parameter deconvolution of plasma hormone concentrations. Proc Natl Acad Sci USA. 84:7686 –7690. Johnson ML, Veldhuis JD. 1995 Evolution of deconvolution analysis as a hormone pulse detection method. Methods Neurosci. 28:1–24. Veldhuis JD, Johnson ML. 1995 Specific methodological approaches to selected contemporary issues in deconvolution analysis of pulsatile neuroendocrine data. Methods Neurosci. 28:25–92. Pincus SM, Keefe DL. 1992 Quantification of hormone pulsatility via an approximate entropy algorithm. Am J Physiol. 262:E741–E754. Mulligan T, Iranmanesh A, Gheorghiu S, Godschalk M, Veldhuis JD. 1995 Amplified nocturnal luteinizing hormone (LH) secretory burst frequency with selective attenuation of pulsatile (but not basal) testosterone secretion in healthy aged men: possible Leydig cell desensitization to endogenous LH signaling–a clinical research center study. J Clin Endocrinol Metab. 80:3025–3031. Urban RJ, Veldhuis JD, Blizzard RM, Dufau ML. 1988 Attenuated release of biologically active luteinizing hormone in healthy aging men. J Clin Invest. 81:1020 –1029. Reyes-Fuentes A, Veldhuis JD. 1993 Neuroendocrine physiology of the normal male gonadal axis. Endocrinol Metab Clin North Am. 22:93–124. Deslypere JP, Kaufman JM, Vermeulen T, Vogelaers D, Vandalem JL, Vermeulen A. 1987 Influence of age on pulsatile luteinizing hormone release and responsiveness of the gonadotrophs to sex hormone feedback in men. J Clin Endocrinol Metab. 64:68 –73. Vermeulen A, Deslypere JP. 1985 Long-term transdermal dihydrotestosterone

34. 35.

36. 37.

38. 39. 40. 41. 42.

43.

44.

45.

46.

47.

1975

therapy: effects on pituitary gonadal axis and plasma lipoproteins. Maturitas. 7:281-285. Mulligan T, Iranmanesh A, Johnson ML, Straume M, Veldhuis JD. 1997 Aging alters feedforward and feedback linkages between LH and testosterone in healthy men. Am J Physiol. 42:R1407–R1413. Veldhuis JD, Iranmanesh A, Samojlik E, Urban RJ. 1997 Differential sexsteroid negative feedback regulation of pulsatile follicle-stimulating hormone secretion in healthy older men: deconvolution analysis and steady state sex steroid hormone infusions in frequently sampled healthy older individuals. J Clin Endocrinol Metab. 82:1248 –1254. Vermeulen A, Deslypere JP, Kaufman JJ. 1989 Influence of antiopioids on luteinizing hormone pulsatility in aging men. J Clin Endocrinol Metab. 68:68 –72. Cagampang FRA, Maeda K-I, Tsukamura H, Ohkura S, Ota K. 1991 Involvement of ovarian steroids and endogenous opioids in the fasting-induced suppression of pulsatile LH release in ovariectomized rats. J Endocrinol 129:321–328. Bergendahl M, Huhtaniemi I. 1994 The time since castration influences the effects of short-term starvation on gonadotrophin secretion in male rats. J Endocrinol. 143:209 –219. Dong Q, Rintala H, Handelsman DJ. 1994 Androgen receptor function during undernutrition. J Neuroendocrinol. 6:397– 402. Allen LG, Kalra PS, Corwley WR, Kalra SP. 1985 Comparison of the effects of neuropeptide Y and adrenergic transmitters on LH release and food intake in male rats. Life Sci. 37:617– 623. Ahima RS, Prabakaran D, Mantzoros C, et al. 1996 Role of leptin in the neuroendocrine response to fasting. Nature. 382:250 –252. Veldhuis JD, Evans WS, Johnson ML. 1995 Complicating effects of highly correlated model variables on nonlinear least-squares estimates of unique parameter values and their statistical confidence intervals: estimating basal secretion and neurohormone half-life by deconvolution analysis. Methods Neurosci. 28:130 –138. Quyyumi SA, Pinkerton JV, Evans WS, Veldhuis JD. 1993 Estradiol amplifies the amount of luteinizing hormone secreted in response to increasing doses of gonadotropin-releasing hormone (GnRH) by specifically augmenting the duration of evoked LH secretory events and hence their mass. J Clin Endocrinol Metab. 76:594 – 600. Veldhuis JD, Zwart AD, Iranmanesh A. 1997 Neuroendocrine mechanisms by which selective Leydig-cell castration unleashes increased pulsatile LH release in the human: an experimental paradigm of short-term ketoconazole-induced hypoandrogenemia and deconvolution-estimated LH secretory enhancement. Am J Physiol. 272:R464 –R474. Veldhuis JD, Metzger DL, Martha Jr PM, et al. 1997 Estrogen and testosterone, but not a non-aromatizable androgen, direct network integration of the hypothalamo-somatotrope (growth hormone)-insulin-like growth factor I axis in the human: evidence from pubertal pathophysiology and sex-steroid hormone replacement. J Clin Endocrinol Metab. 82:3414 –3420. Zwart AD, Urban RJ, Odell WD, Veldhuis JD. 1996 Contrasts in the gonadotropin-releasing dose-response relationships for luteinizing hormone, follicle-stimulating hormone, and alpha-subunit release in young vs. older men: appraisal with high-specificity immunoradiometric assay and deconvolution analysis. Eur J Endocrinol. 135:399 – 406. Tenover JS, Matsumoto AM, Clifton DK, Bremner WJ. 1988 Age-related alterations in the circadian rhythms of pulsatile luteinizing hormone and testosterone secretion in healthy men. J Gerontol. 43:M163-168.