Glucose ingestion acutely lowers pulsatile LH and basal testosterone secretion in men

Abstract

Chronic hyperglycemia inhibits the male gonadal axis. The present analyses test the hypothesis that acute glucose ingestion also suppresses LH and testosterone (T) secretion and blunts the LH-T dose-response function. The design comprised a prospectively randomized crossover comparison of LH and T secretion after glucose vs. water ingestion in a Clinical Translational Research Center. The participants were healthy men (n = 57) aged 19–78 yr with body mass index (BMI) of 20–39 kg/m². The main outcome measurements were deconvolution and LH-T dose-response analyses of 10-min data. LH-T responses were regressed on glucose, insulin, leptin, adiponectin, age, BMI, and CT-estimated abdominal visceral fat. During the first 120 min after glucose ingestion, for each unit decrease in LH concentrations, T concentrations decreased by 86 (27–144) ng/dl (r = 0.853, P < 0.001). Based upon deconvolution analysis, glucose compared with water ingestion reduced 1) basal (nonpulsatile; P < 0.001) and total (P < 0.001) T secretion without affecting pulsatile T output and 2) pulsatile (P = 0.043) but not basal LH secretion. By multivariate analysis, pulsatile LH secretion positively predicted basal T secretion after glucose ingestion (r = 0.374, P = 0.0042). In addition, the glucose-induced fall in pulsatile LH secretion was exacerbated by higher fasting insulin concentrations (P = 0.054) and attenuated by higher adiponectin levels (P = 0.0037). There were no detectable changes in the analytically estimated LH-T dose-response curves (P > 0.30). In conclusion, glucose ingestion suppresses pulsatile LH and basal T secretion acutely in healthy men. Suppression is influenced by age, glucose, adiponectin, and insulin concentrations.

high body mass index (BMI), advancing adult age, chronic renal failure, metabolic syndrome X, prolonged fasting, weight loss, sustained hyperglycemia, acute hypoglycemia, and diabetes mellitus are associated with decreased testosterone (T) concentrations and variable suppression of LH secretion (3, 7, 10, 14, 16, 26, 29, 38, 40, 43, 44). Intra-abdominal adiposity, hyperinsulinism, metabolic syndrome, and/or glucose intolerance are possible common substrates in the pathophysiology of these conditions. For example, excessive abdominal visceral fat (AVF) is accompanied by reductions in LH pulse amplitude, total T, and sex hormone-binding globulin (SHBG) concentrations (1, 31, 36, 38). Insulin resistance and impaired glucose tolerance are concomitants of visceral adiposity. However, their individual and joint roles in hypoandrogenemia are difficult to quantify precisely. Although in vitro studies indicate that IGF-I and insulin can potentiate human chorionic gonadotropin-stimulated T secretion (2, 6, 8, 35), and adipocytokines like leptin and adiponectin can suppress T secretion (11, 39), clinical data are contradictory and inconsistent. In one study, insulin resistance correlated inversely with the maximal serum T concentration after human chorionic gonadotropin injection (34), whereas in another study a 6-h euglycemic hyperinsulinemic clamp did not alter LH or free T concentrations in men (29). In earlier investigations, an insulin clamp increased T levels in obese men (33), insulin sensitivity correlated positively with free T concentrations in young men (22, 46), oral glucose, rosiglitazone, metformin, and diazoxide reduced T and either did not affect or elevated LH concentrations (19, 30, 32, 45), and weight loss in obese men increased both T and LH levels (26).

Beyond inconsistency of inference, laboratory and clinical data have other major limitations. First, available studies have not controlled for intersubject variability in age, BMI, AVF, fasting plasma glucose, insulin, leptin, or adiponectin concentrations. Second, earlier investigations did not directly compare the effects of glucose loading with those of calorie-free liquid ingestion at the same time of day in the same individuals. And third, the dynamic bases for LH and T changes have not been quantified by frequent blood sampling and modern analytical tools (42), leaving the mechanism(s) of any acute oral glucose effects open to question. The present investigation addresses these limitations in a cohort of 57 healthy men aged 19–78 yr by intraindividual comparison of water vs. oral glucose ingestion at the same time of day, 10-min blood sampling for 6.5 h, deconvolution analysis of LH and T secretion, and analytical dose-response estimation. This design allowed assessment of the interactive effects of glucose and age on LH/T secretion.

METHODS

Subjects.

Fifty-seven healthy men were recruited to participate after providing voluntary written informed consent approved by the local institutional review board. The admissible age range was 18–80 yr, with BMI 20–40 kg/m². Exclusion diagnoses were congestive heart failure, acute or chronic liver or renal disease, anemia, hypothalamo-pituitary disease, neuropsychiatric drug exposure, glucocorticoid use, systemic inflammatory process, malignancy, substance abuse, intracranial disease, sleep apnea, or diabetes mellitus. Inclusion criteria were community-dwelling, independently living, consenting adults with stable diurnal work habits, body weight (within 2 kg in 3 mo), and recreational exercise patterns.

Protocol.

Subjects undertook two 10-min sampling sessions after overnight fasting in the morning beginning at 0800. At 0830, glucose (75 g) or the same volume of water (10 oz) was administered orally. Blood sampling continued thereafter for an additional 6 h (until 1430). In three subjects, there was a delay (not exceeding 1 h) in starting the protocol. Thus, total sampling was 6.5 h.

AVF.

An abdominal CT scan was performed at the L4–L5 interspace to estimate AVF cross-sectional area as described (21a).

Assays.

Serum concentrations LH and T were assayed in each sample (108 samples/subject) by Immulite 2000 (Siemens Healthcare Diagnostics, Flanders, NJ). The assay for LH has a detection range of 0.05–200 IU/l with intra- and interassay coefficients of variation of 3.4 and 6.6%, respectively. The T assay has a detection range of 15–600 ng/dl, with intra- and interassay coefficients of variation of 8.4 and 10%, respectively. Serum samples were used to measure glucose and insulin concentrations at 0800 and then every 10 min via Synchron (Beckman, Chaska, MN) and Siemens (Deerfield, IL) Dimension Vista autoanlyzers.

Adiponection and leptin were assayed in 0800 serum by radioimmunoassay, using reagents from Millipore (Billerica, MA). The assay for adiponectin has intra- and interassay coefficients of variation of 1.78–6.21 and 6.25–9.25%, respectively, at circulating concentrations of 1.5–7.5 mg/l. Leptin assay has intra- and interassay coefficients of variation of 3.4–8.3 and 3.0–6.2%, respectively, at circulating concentrations of 4.9–25.6 μg/l.

Analyses.

LH and T concentration-time series (all 6.5 h) were subjected to automated deconvolution analysis using a validated maximum-likelihood estimation methodology (27). The two-component LH elimination model comprised a fast half-life of 6.9 min and an estimable (37% of total) slow half-life. Corresponding values for T were 1.4 and 27 min (63% slow component) (41). Outcome variables were basal (nonpulsatile), pulsatile, and total (sum of basal plus pulsatile) secretion and the mass (concentration units), number (per 6.5 h), and shape (mode) of LH and T secretory bursts.

Dose-response estimates of LH-T feedforward drive were performed on the 6.5-h profiles, as described recently (25). The program regresses deconvolved T secretion rates on reconvolved LH concentrations via a four-parameter logistic dose-response model.

HOMA-IR (homeostatic model assessment-insulin resistance) was calculated from the fasting (0800) insulin and glucose values, as described (28).

Statistics.

A paired Student's t-test was utilized to evaluate the effect of oral glucose compared with water ingestion on ln-transformed deconvolution and dose-response parameters. Uni- and multivariate linear regression analyses were employed to assess effects of (0800 sample values of) adiponectin, leptin, glucose, and insulin as well as subject age, BMI, and AVF on primary outcomes (17). Matlab 7.8.0 (The MathWorks, Natick, MA) was the software platform. HOMA-IR was assessed univariately. Data are the geometric means ± SE and the median (range).

RESULTS

Subject characteristics are summarized in Table 1. Age and BMI ranges were 19–78 yr and 18–39 kg/m², respectively. There were three subjects with impaired fasting glycemia, viz. fasting glucose concentrations of 119, 112, and 108 mg/dl (BMI 35, 26, and 34 kg/m²; insulin 43, 78, and 23 μU/ml). None had glucose ≥126 mg/dl.

Table 1.

Subject characteristics

End point	Means ± SE	Median (Range)
Age, yr	39 ± 2.3	47 (19–78)
BMI, kg/m2	26 ± 0.52	26 (18–39)
AVF, cm2	90 ± 14	105 (14–410)
Insulin, μU/ml	4.8 ± 0.62	4.7 (2–22)
Adiponectin, mg/l	9.3 ± 0.97	11 (3.1–33)
Plasma glucose, mg/dl	93 ± 1.2	95 (73–123)
Leptin, μg/l	4.8 ± 0.66	5.4 (1.0–16)
SHBG, nM	32 ± 1.9	34 (13–78)
Cortisol, μg/dl	10 ± 0.31	10 (6.8–17)
LH, IU/l	4.0 ± 0.32	4.1 (1.5–10)
T, ng/dl	416 ± 20	399 (123–801)
HOMA-IR	1.46 ± 0.16	1.0 (0.4–6.0)

Data are geometric means ± SE and median (range) for n = 57 healthy men. HOMA-IR, homeostatic model assessment-insulin resistance; BMI, body mass index; AVF, abodminal visceral fat; SHBG, sex hormone-binding globulin; T, testosterone.

Figure 1 presents median (group of n = 57) concentrations of LH (dashed line) and T (solid line) (top) and analogously for glucose and insulin (bottom) over time (400 min) beginning at 0800 on the day of water ingestion (A) and oral glucose tolerance testing (OGTT; B). On the glucose day, there was a sharp fall in both LH and T concentrations after the third blood sample when glucose was ingested. To visualize intraindividual responses, Fig. 2 depicts median δ (glucose − water decrement) LH and T concentrations measured every 10 min for 6.5 h. Nadirs in LH and T concentrations occurred 139 and 113 min after glucose ingestion. The slope of LH concentrations regressed on time over the first 120 min of these intervals, although they were small in absolute value, and decreased consistently, becoming 10-fold more negative (steeper) after glucose than after water ingestion (P < 10⁻⁴; Table 2). The slope of total T concentrations over the same 120 min also decreased consistently, becoming 5.5-fold steeper (more negative) after glucose than after water exposure (P < 0.001). Univariate regression of decremental T on decremental LH concentrations (algebraic differences between glucose and water values) showed that the drop in LH strongly predicted the drop in T concentrations due to glucose ingestion [viz.: r = 0.853, P < 0.001, slope 86 (27–144) (ng/dl)·(IU/l)⁻¹]. Thus, LH and T decrease more over time after glucose than after water ingestion, and decreases in the two hormones are significantly correlated.

Fig. 1.

Median 10-min concentrations of LH and testosterone (T) (top) and insulin and glucose (bottom) on the day of water ingestion (fasting; A) and glucose ingestion [oral glucose tolerance test (OGTT); B] in 57 men. The 1st sample (0 time) was obtained at 0800 (methods). Water/glucose was ingested after the 3rd blood sample (0800).

Fig. 2.

Mean δ-serum (glucose − water) LH and T concentrations sampled every 10 min for 6.5 h after water vs. glucose ingestion in 57 men. Water/glucose ingestion occurred after the 3rd blood sample.

Table 2.

Initial slopes of LH and T concentrations on time (1st 120 min after water or glucose ingestion)

	Slope	95% CI	r2	P Value
LH
Water	−0.00060	−0.0022 to 0.0010	0.048	0.43
Glucose*	−0.0058	−0.0077 to −0.0038	0.759	<10−4
T
Water	−0.13	−0.20 to −0.06	0.577	0.001
Glucose*	−0.71	−0.87 to −0.55	0.879	<10−4

Data are mean linear-regression slopes with 95% confidence interval (CI) and P values vs. zero slope.

^*P < 0.001 for glucose slope vs. water slope.

Deconvolution analysis was applied to assess the mechanisms mediating changes in mean hormone levels. Estimates of LH secretion on the water and glucose days were strongly correlated, viz. for basal, pulsatile, and total LH secretion (all P < 10⁻⁴ with r = 0.736 to 0.961) and for corresponding estimates of T secretion (all P < 0.001, r = 0.595 to 0.942). Paired comparisons disclosed that pulsatile LH secretion fell after glucose vs. water ingestion (P = 0.043; Table 3). Total LH secretion (IU·l⁻¹·6.5 h⁻¹) tended to change in a similar direction (53 ± 5.3 water and 48 ± 4.3 glucose, P = 0.055, n = 57). Basal LH secretion was not affected. In contrast, basal (rather than pulsatile) T secretion declined after glucose vs. water administration (P < 0.0001; Table 4). Concomitantly, total T secretion fell (P < 0.0001) due to the fall in basal T secretion.

Table 3.

Deconvolution analysis of pulsatile LH secretion after water (control) or glucose ingestion

Measurement	Control (Water)	Oral Glucose	P Value
No. of pulses/6.5 h	4.2 ± 0.14 (4.0)	3.9 ± 0.15 (4.0)	0.16
LH t1/2, min	43 ± 2.0 (40)	48 ± 2.5 (42)	0.085
Mode of burst, min	11 ± 0.55 (13)	12 ± 0.47 (14)	0.17
Basal LH secretion, IU·l−1·6.5 h−1	32 ± 4.4 (330)	28 ± 3.8 (28)	0.20
Pulsatile LH secretion, IU·l−1·6.5 h−1	20 ± 1.55 (20)	18 ± 1.5 (18)	0.043
Total LH secretion, IU·l−1·6.5 h−1	53 ± 5.3 (54)	48 ± 4.3 (46)	0.055
Mass/burst, IU/l	4.8 ± 0.43 (5.5)	4.5 ± 0.48 (4.6)	0.41

Data are the geometric means ± SE (median in parentheses); n = 57. P values are the results of paired 2-tailed t-tests on log data. Boldface denotes P < 0.05.

Table 4.

Deconvolution analysis of pulsatile T secretion after water (control) or glucose ingestion

Measure	Control (Water)	Oral Glucose	P Value
No. of pulses/6.5 h	4.9 ± 0.20 (5.0)	4.8 ± 0.19 (5.0)	0.66
Mode of burst, min	11 ± 0.77 (13)	10 ± 0.67 (13)	0.48
Basal T secretion, ng·dl−1·6.5 h−1	5,256 ± 276 (5,269)	4,608 ± 219 (4,913)	<0.0001
Pulsatile T secretion, ng·dl−1·6.5 h−1	685 ± 65 (742)	748 ± 73 (737)	0.30
Total T secretion, ng·dl−1·6.5 h−1	6,065 ± 294 (6,165)	5,474 ± 248 (5,598)	<0.001
Mass/burst, ng/dl	140 ± 14 (148)	156 ± 13 (147)	0.18

Data are means ± SE (median in parentheses) and presented as defined in Table 1. Boldface denotes P < 0.001.

Univariate linear regression analysis revealed that higher fasting glucose concentrations predict lesser (glucose-water) decrements in T concentrations (r = 0.324, P = 0.014) and in total T secretion (r = 0.332, P = 0.011) (Fig. 3; note that negative values of the x-axis denote a greater glucose − water fall in the y-parameter). Older compared with young age was associated with a lesser (glucose-water) fall in LH half-life (r = 0.332, P = 0.012; Table 5). This indicates that LH half-life is more stable toward OGTT in older men. Higher insulin predicted a larger (r = −0.256, P = 0.054) whereas higher adiponectin predicted a smaller (r = 0.382, P = 0.0037) glucose-associated drop in pulsatile LH secretion (Fig. 4, top and bottom). During OGTT, pulsatile LH secretion correlated positively with basal T secretion (r = 0.374, P = 0.0042; Fig. 5).

Fig. 3.

Linear regression of T concentration (Conc) decrements (glucose − water values) on fasting plasma glucose concentrations in 57 men. P and r² are noted, since r² values estimate the fraction of variance explained by the independent variable. Negative abscissa values denote a fall in the y-parameter value after glucose compared with water ingestion. Zero denotes no change, and positive values denote an increase after glucose.

Table 5.

Significant univariate regression of deconvolution parameters on metabolic measurements

	r Value	P Value	Independent Variable
Decrement of LH (glucose-water)
Half-life	0.332	0.012	Age (+)
Basal secretion	0.370	0.0050	Adiponectin (−)
Pulsatile secretion	0.382	0.0037	Adiponectin (+)
Pulsatile secretion	0.257	0.054	Insulin (−)
Decrement of T (glucose-water)
Total T secretion	0.324	0.014	Glucose (+)
Basal T secretion	0.275	0.038	HOMA-IR (−)

Correlations were made between parameter decrements (glucose − water ingestion) and age or metabolic measurements carried out on 0800 serum samples (see methods). A negative correlation coefficient (r value) means that the parameter value falls more after glucose when the independent variable increases and vice versa. For example, higher adiponectin concentrations blunt the fall in pulsatile LH secretion, but the converse is true for insulin concentrations.

Fig. 4.

Negative relationship between fasting insulin (top) and positive relationship between fasting adiponectin concentrations (bottom) and the decrement (glucose − water values) in pulsatile LH secretion in 57 men. See format in Fig. 3.

Fig. 5.

Positive association between basal T secretion after glucose and pulsatile LH secretion after glucose ingestion in 57 men. See format in Fig. 3.

Multivariate regression was used to evaluate joint associations of LH (or T) with AVF, BMI, age, glucose, insulin, adiponectin, and leptin (Table 6). On the water ingestion day, increased age (after correction for BMI, AVF, insulin, glucose, adiponectin, and leptin) was associated with decreased nadir (P = 0.0049, r² = 0.327) and decreased peak (P = 0.0019, r² = 0.357) T concentrations. On the glucose ingestion day, greater age and higher fasting (0800) insulin levels together forecast lower nadir T concentrations (P = 0.033, multivariate r² = 0.256), whereas greater age and higher fasting glucose together forecast lower peak T concentrations (P = 0.0018, r² = 0.357; joint correlation).

Table 6.

Multivariate correlates of nadir and peak T concentrations

Hormone Measurement	Water (Baseline)	Glucose Ingestion
Nadir T concentration	Log age*	Log age/log insulin†
Peak T concentration	Log age‡	Log age/log glucose§

Multivariate analysis comprised adjustment for all of BMI, AVF, insulin, glucose, age, adiponectin, and leptin. All correlations were negative.

^*P = 0.0049, r² = 0.327;

^†P = 0.033, r² = 0.256;

^‡P = 0.0019, r² = 0.357;

^§P = 0.0018, r² = 0.359.

Endogenous LH-T dose-response parameters were estimated separately after glucose and water ingestion, as summarized in Table 7. Allowance for possible T downregulation during an LH pulse yielded both onset (initial) and offset (downregulated) LH potency estimates. Potency is defined algebraically as the estimated concentration of LH driving one-half maximal T secretion (EC₅₀ value). Onset LH EC₅₀ values were comparable in the two study sessions, viz. 4.6 ± 0.62 (water) and 4.8 ± 0.29 (glucose) IU/l. Downregulated (offset) LH EC₅₀ values were also similar at 9.9 ± 2.4 (water) and 9.8 ± 0.62 (glucose) IU/l. T secretion sensitivity (slope term), LH efficacy (maximal asymptotic T secretion rate), and model error terms also did not differ for the two interventions (all P > 0.30).

Table 7.

LH-T dose-response analysis after water (control) vs. glucose ingestion

Measurement	Water	Oral Glucose	P Value
Onset potency (EC50), IU/l	4.6 ± 0.62 (7.4)	4.8 ± 0.29 (11)	0.34
Recovery potency (EC50), IU/l	9.9 ± 2.4 (30)	9.8 ± 0.62 (39)	0.44
Sensitivity (slope)	3.2 ± 0.37 (2.6)	3.3 ± 0.33 (2.9)	0.98
Efficacy, ng·dl−1·min−1	56 ± 14 (63)	74 ± 17 (65)	0.75
Time delay to downregulation, min	23 ± 1.3 (25)	24 ± 1.7 (28)	0.98
Model SD	6.3 ± 0.41 (6.0)	6.2 ± 0.39 (5.4)	0.95
Random effects (×103)	3.5 ± 3.0 (1.6)	1.9 ± 1.7 (2.0)	0.89

Data are geometric means ± SE (median in parentheses); n = 57. Dose-response P values reflect paired 2-tailed t-tests of log data.

DISCUSSION

In 57 men aged 19–78 yr with BMI of 20–39 kg/m², glucose ingestion at 0830 significantly lowered mean total T concentrations, basal T secretion rates, and total (basal + pulsatile) T secretion rates compared with water ingestion at the same time of day. In the case of LH, pulsatile rather than basal secretion fell with OGTT. T and LH concentrations, as well as basal T and pulsatile LH secretion, declined proportionately. In particular, the initial 2-h decrement in mean T concentrations was 86 (27–144) ng/dl (P = 0.0062) per unit decrease in LH concentrations (IU/l). T secretory burst size (mass, ng/dl), mode (min), and number (per 6.5 h) did not change, indicating that glucose selectively suppresses nonpulsatile (basal) T secretion. Analytical LH-T dose-response estimates yielded no detectable changes in LH sensitivity, potency (EC₅₀), or efficacy values after glucose ingestion. Accordingly, barring a selective increase in the metabolic clearance rate of T, the data indicate that oral glucose administration in men acutely suppresses pulsatile LH and basal T secretion after an overnight fast. The basis for this novel finding is not yet clear, since more prolonged fasting inhibits LH and T secretion in men (3, 7, 10).

Linear regression analysis revealed significant relationships between glucose-induced decrements in LH and T secretion parameters and age, glucose, insulin, and adiponectin. First, the glucose-associated δ in LH half-life was related to age such that as age increased LH half-life changed less. The reason for this interaction between age and glucose effect is not known but may reflect longer LH half-lives in older men at baseline. Second, glucose-associated decrements in pulsatile LH secretion correlated positively with fasting (0800) adiponectin concentrations and negatively with fasting (0800) insulin concentrations. Higher pulsatile LH secretion in turn predicted higher basal and higher total T secretion acutely. These outcomes suggest the hypotheses that adiponectin may protect against, whereas insulin may exacerbate, the capability of oral glucose to suppress LH and T secretion. Third, higher fasting (0800) glucose predicted a lesser drop in post-OGTT total T concentrations and total T secretion. This may relate to the tendency of the metabolic syndrome itself to lower total T levels (30, 32, 33, 36, 38).

A multivariate assessment with adjustment for age, leptin, adiponectin, glucose, insulin, BMI, and AVF disclosed significant negative effects of age and insulin on nadir T concentrations and of age and glucose on peak T concentrations post-OGTT. The correlations explained about 26 and 36% of the variance in nadir T and peak T values, respectively. The effect of insulin on T secretion tends to be stimulatory when studied in vitro (2, 8, 9, 9, 22), albeit not necessarily chronically or in vivo (29, 33). A plausible postulate is that insulin serves as a marker of metabolic alterations associated with obesity and risk of type 2 diabetes, which result in lower T levels (31). Whether testis and brain GLUT1 and GLUT8 transporters or brain glucagon-like peptide-1 (GLP-1) receptors contribute to direct effects of glucose on T secretion is not known (4, 13, 21, 22, 37). This seems unlikely because the estimated LH-T dose-response curve did not change during OGTT.

A simple hypothesis for glucose-induced lowering of LH and T concentrations acutely would be via hypothalamic effects, since both LH and T fell commensurately as in the hyperglycemic rat (16). This could be mediated by glucose taken up in the central nervous system (13), in intestinal glucose receptors directing vagal feedback to the brain (5), or indirectly via inhibitory effects of glucose-stimulated GLP-1 release. GLP-1 is an insulin cosecretagogue with glucose. The capability of two other insulin sensitizers, rosiglitazone and metformin, to also reduce T secretion in men (30, 45) would be consistent with a postulated negative effect of insulin or GLP. In addition, hypothalamo-testicular neural pathways exist, which inhibit T secretion via brainstem-spinal cord-testis connections after central interleukin-1β administration (37). We did not measure this cytokine, but like other inflammatory peptides, interleukin levels correlate positively with AVF, age, insulin, and metabolic syndrome X (18, 20, 36). In the present study, insulin and adiponectin correlated oppositely with glucose-induced suppression of pulsatile LH (and thereby basal T) secretion, with insulin exacerbating and adiponectin preventing LH-T suppression. Although both leptin and adiponectin can inhibit Leydig cells in vitro (11, 12), leptin was not a significant correlate, and adiponectin was a protective predictor of LH measurements here after glucose ingestion. The positive association between adiponectin concentrations and LH secretion may reflect the broader positive metabolic association between adiponectin and effectual insulin action in the brain (24). The anti-inflammatory effects of adiponectin provide another plausible testable pathway for mediating this dissociation since brain inflammation reduces LH secretion (23).

Caveats include the relatively smaller number of men studied here (n = 57), albeit requiring 6,156 measurements of LH and the same number for T. Thus, regression inferences should be confirmed in larger cohorts of individuals. In addition, whether diabetes or prediabetes alters the T-suppressive effect of oral glucose cannot be determined by these data. As an indirect measure of fasting insulin resistance, HOMA-IR correlated inversely with post-OGTT basal T secretion. Analytical estimates of LH-T dose responsiveness should be confirmed ultimately by exogenous LH stimulation protocols. The possibility that glucose ingestion rapidly increases T clearance is unlikely, since SHBG concentrations that determine T clearance change by <5% during OGTT (15, 19).

In summary, oral glucose administration acutely lowers LH and total T concentrations by suppressing pulsatile LH secretion and basal T secretion commensurately, with no significant change in the calculated LH-T dose-response function. Regression analyses_suggest that adiponectin and insulin may have protective and exacerbating effects, respectively, on the acute LH/T fall after oral glucose administration in men.

GRANTS

This work is supported in part via Center for Translational Science Activities Grant no. 1 UL-1-RR-024150 from the National Center for Research Resources (Rockville, MD) and AG-031763, AG-023133, AG-02377, and DK-050456 (Metabolic Studies Core of the Minnesota Obesity Center) from the National Institutes of Health (Bethesda, MD). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Aging or the National Institutes of Health. Matlab versions of the deconvolution methodology are available from J. D. Veldhuis at veldhuis.johannes@mayo.edu.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.I. and J.D.V. did the conception and design of the research; A.I. and J.D.V. interpreted the results of the experiments; A.I. and J.D.V. edited and revised the manuscript; A.I., D.L., and J.D.V. approved the final version of the manuscript; D.L. performed the experiments; J.D.V. analyzed the data; J.D.V. prepared the figures; J.D.V. drafted the manuscript.