| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Göteborg Pediatric Growth Research Center, Department of Pediatrics, Institute of Clinical Sciences, The Sahlgrenska Academy at Göteborg University, S-416 85 Göteborg, Sweden
Address all correspondence and requests for reprints to: Kerstin Allvin, M.D., Göteborg Pediatric Growth Research Center, The Sahlgrenska Academy at Göteborg University, The Queen Silvia Childrens Hospital, S-416 85 Göteborg, Sweden. E-mail: kerstin.allvin{at}vgregion.se.
| Abstract |
|---|
|
|
|---|
Objective: We hypothesized that adult human males born small for gestational age (SGA) have an altered sex hormone profile.
Design, Setting, and Patients: Twenty-five adult males born SGA with median birth weight –2.2 SD scores (SDS) and birth length –2.4 SDS were studied. Median age was 23.1 yr and final height –0.5 SDS. They were compared with 44 male controls with median age 20.5 yr and final height 0.4 SDS.
Main Outcome Measure: The primary outcome before the study started was 17β-estradiol (E2) levels in SGA males.
Results: The SGA group showed significantly higher median levels of E2, 17.9 pg/ml (P < 0.001), and dihydrotestosterone (DHT), 0.543 ng/ml (P < 0.05), compared with controls, 12.6 pg/ml and 0.423 ng/ml, respectively. Testosterone (T) levels did not differ between groups. E2 to T ratio correlated negatively to birth weight (r = –0.40, P < 0.01) and birth length (r = –0.44, P < 0.001). DHT to T ratio correlated negatively to birth weight (r = –0.51, P < 0.001) and birth length (r = –0.38, P < 0.01). Males born SGA also had significantly higher median levels of inhibin B, 164 pg/ml (P < 0.05), compared with controls, 137 pg/ml. Inhibin B correlated negatively to birth length (r = –0.34, P < 0.01).
Conclusion: SGA males of normal stature have higher levels of E2, DHT, and inhibin B than controls, indicating a disturbed steroid synthesis or metabolism. Aromatase activity, calculated as E2 to T ratio, and 5
-reductase activity, calculated as DHT to T ratio, is negatively correlated to size at birth.
| Introduction |
|---|
|
|
|---|
Studies on the impact of fetal growth on the hypothalamic-pituitary-adrenal axis have shown an increase in urinary adrenal metabolites in children born light at birth (2). Moreover, it is found that earlier and pronounced adrenarche occurs in children born small for gestational age (SGA) (3, 4). In boys and young males, birth weight is inversely related to cortisol responses to stress, whereas in girls, morning peak cortisol is inversely related to birth weight (5).
Prenatal growth restriction may also affect future fertility in females. Adolescent girls born SGA are at risk of developing a disturbed HPG function, with an ovarian hyporesponsiveness to FSH and a reduced ovulation rate (6, 7). However, little is known about what long-term effects intrauterine growth retardation may have on the HPG function in males. In boys low birth weight is associated with hypospadia, cryptorchidism, and testicular cancer, an entity known as testicular dysgenesis syndrome (8). Cicognani et al. (9) found a reduction of testicular volume in SGA males with low final height relative to target height. The testosterone (T) levels were lower and levels of LH were higher than in normal controls, indicating a different setup of the HPG axis with a tendency to hypogonadism in the SGA subjects. These results, with increased levels of gonadotropins, are similar to those in females, pointing out a peripheral partial insensitivity to gonadotropins (6).
Nilsson et al. (10) have shown that prenatal programming, induced by giving endotoxins to pregnant rats, results in features of the metabolic syndrome, such as obesity, insulin resistance, and high plasma levels of leptin in the male offspring. Interestingly, the male offspring also showed increased plasma levels of 17β-estradiol (E2) and progesterone but normal levels of testosterone (10).
Based on human female data and experimental studies in male rats, we hypothesized that intrauterine growth restriction in human males leads to a disturbed sex hormone synthesis with increased levels of estradiol. The aim of this study was to evaluate sex hormone levels and indirectly the different enzyme activities correlating to birth size in adult males born SGA.
| Patients and Methods |
|---|
|
|
|---|
This study was performed as a case-control study. The study group comprised of 25 adult males of normal stature born SGA (below –2 SDS in birth weight and/or birth length) according to the Swedish reference for newborns (11).
Fifty males born SGA, with catch-up growth [defined as final height above –2 SD scores (SDS) according to the Swedish growth reference values (12)], from a previously described cohort (13) were asked to participate in the present study. Twenty-two of these men accepted participation. They all had a testicular volume of 20 ml or greater, and none had hypospadia or cryptorchidism. Three men were recruited from a population-based cohort, randomly chosen according to their living addresses (14). All males in the study group were born either light (n = 5), short (n = 7), or both short and light (n = 13). Five boys were born prematurely (range 33–36 wk of gestation) and 20 boys at term. Auxological data were available from birth for all 25 men and at 2 yr of age for 21 men.
The control group comprised 44 adult healthy men of normal stature, recruited from the population-based cohort mentioned above (14). Thirty-nine of the controls were born appropriate for gestational age (± 2 SDS in birth weight and birth length) and five born large for gestational age (birth weight > 2 SDS), according to the Swedish reference (11). Two boys were born prematurely (33 and 36 wk of gestation). Their birth weights and lengths were collected from the Swedish Birth Register. Individual heights were measured by an Ulmer stadiometer attached to the wall and values were converted into SDS. None of the men from either group had evidence of serious complications in the neonatal period, syndromes, serious malformations, or a condition requiring chronic medical treatment. Body mass index (BMI) was calculated as recommended by the World Health Organization: weight (kilograms)/length (meters)2.
The participants were examined at the outpatient clinic between 0800 and 1000 h after a 12-h fasting period. The blood samples in both groups studied were collected during 3 months. Serum was frozen and stored at –80 C until hormone measurements were done.
Methods
Auxological data and age
The changes in height SDS from birth were calculated for each participant. We used the terms
height(2yr-birth) SDS and
height(adult-birth) SDS for the difference in height SDS between birth and 2 yr of age and adult height, respectively.
The changes in weight SDS for each individual were named
weight(2 yr-birth) SDS and
weight(adult-birth) SDS and calculated similarly.
Hormone assays All serum hormones were determined in duplicate. Intra- and interassay coefficients of variation (CVs) are given for measured range. Serum E2 concentrations were determined using a modified RIA (Spectria estradiol; Orion Diagnostica, Espoo, Finland) as previously been described (15). The detection limit was 1.6 pg/ml. The intraassay CV was 17% for 2.7 pg/ml and less than 8% for levels above 12 pg/ml. The interassay CV was 29% for 3.0 pg/ml, 16% for 11 pg/ml, and 10% for 31 pg/ml.
Serum testosterone concentrations were determined using a modified RIA (Spectria testosterone; Orion Diagnostica) as previously been described (16). The detection limit was 0.01 ng/ml. The intraassay CV was less than 6%. The interassay CV was less than 8%.
Free testosterone was calculated from testosterone, SHBG, and albumin using the method described by Vermeulen et al. (17) and Dunn et al. (18).
Serum dihydrotestosterone (DHT) concentrations were determined using an extraction RIA (active dihydrotestosterone RIA; Diagnostic Systems Laboratories, Webster, TX). The detection limit was 0.003 ng/ml. The intraassay CV was less than 6%. The interassay CV was less than 9%.
Serum dehydroepiandrosterone sulfate (DHEAS) concentrations were determined using a RIA (Coat-A-Count DHEA-SO4; Diagnostic Products Corp., Los Angeles, CA). The detection limit was 11 ng/ml. The intraassay CV was less than 8% for values over 369 ng/ml. The interassay CV was 7% for 1476 ng/ml and 14% for 4428 ng/ml.
Serum inhibin B concentrations were determined by ELISA (Diagnostic Systems Laboratories). The detection limit was 7 pg/ml. The intraassay CV was less than 12%. The interassay CV was 7% for 61 pg/ml and 6% for 121 pg/ml.
Serum SHBG concentrations were determined using an immunoradiometric assay (Spectria SHBG immunoradiometric assay; Orion Diagnostica). The detection limit was 0.15 µg/dl. The intraassay CV was less than 5%. The interassay CV was less than 6%.
Serum LH concentrations were determined using time-resolved fluoroimmunoassay (AutoDELFIA hLH Spec; Wallac, Turku, Finland). The detection limit was 0.05 U/liter. The intraassay CV was less than 9%. The interassay CV was less than 3%.
Serum FSH concentrations were determined using time-resolved fluoroimmunoassay (AutoDELFIA FSH, Wallac, Turku, Finland). The detection limit was 0.05 U/liter. The intraassay CV was less than 1%. The interassay CV was less than 3%.
Serum adiponectin concentrations were determined by ELISA (Mediagnost, Reutlingen, Germany). The detection limit was 0.5 µg/ml. The intra- and interassay CVs were less than 10%.
Serum leptin concentrations were determined by RIA (Linco Research, Inc., St. Charles, MO). The detection limit was 0.3 µg/liter. The intraassay CV was less than 7%. The interassay CV was less than 10%.
Hormone ratios E2 to testosterone ratio was calculated as serum E2 (picograms per milliliter) divided by serum testosterone (nanograms per milliliter). DHT to testosterone ratio was calculated as serum DHT (nanograms per milliliter) divided by serum T (nanograms per milliliter). LH to testosterone ratio was calculated as serum LH (units per liter) divided by serum T (nanograms per milliliter). FSH to inhibin B ratio was calculated as serum FSH (units per liter) divided by serum inhibin B (picograms per milliliter). Leptin to adiponectin ratio was calculated as serum leptin (nanograms per milliliter) divided by serum adiponectin (micrograms per milliliter).
Statistical analyses
Data are presented as median and interquartile range. All correlation analyses were made with Spearman nonparametric rank correlation. Mann-Whitney U test was used for comparison of data between groups.
When comparing the concentrations of E2 and DHT between SGA and controls, we adjusted for the differences in adult age and height between groups using analysis of covariance with adult age and height as covariates.
For statistical analyses, we used the computer program Statistical Package for Social Science (SPSS, Inc., Chicago, IL). All tests were conducted at a 5% level of significance.
Ethical consideration
The study was approved by the Ethics Committee of the Medical Faculty of the University of Göteborg. Informed consent was obtained from the participants.
| Results |
|---|
|
|
|---|
As expected, there was a difference in adult height between the SGA and the control group, with SGA being shorter (–0.5 vs. 0.4 SDS, P < 0.01) (Table 1
). SGA males were older than controls (23.1 vs. 20.5 yr, P < 0.05). Adult weight or BMI did not differ between groups (Table 1
).
|
E2 levels were higher in the SGA group (17.9 vs. 12.6 pg/ml, P < 0.001) (Table 2
). After adjustment for age and height, there was still a significant mean difference between SGA and controls for E2 of 5.1 pg/ml (95% confidence interval 2.0–8.2 pg/ml, P < 0.01). Moreover, E2 levels correlated negatively with birth length SDS (r = –0.40, P < 0.01) and birth weight SDS (r = –0.30, P < 0.05). There was no correlation between E2 and adult weight SDS or BMI.
|
height(adult-birth) SDS (r = 0.37, P < 0.01) and
weight(adult-birth) SDS (r = 0.45, P < 0.001) (Fig. 2
|
|
DHT levels were higher in the SGA group (0.543 vs. 0.423 ng/ml, P < 0.05) (Table 2
). After adjustment for age and height, there was still a significant mean difference between SGA and controls for DHT of 0.091 ng/ml (95% confidence interval 0.014–0.168 ng/ml, P < 0.05). DHT correlated negatively to birth length SDS (r = –0.27, P < 0.05) and birth weight SDS (r = –0.27, P < 0.05).
5
-Reductase activity, calculated as DHT to T ratio, was higher in SGA males (0.11 vs. 0.09, P < 0.01) (Table 2
). DHT to T ratio also correlated negatively with both birth length SDS (r = –0.38, P < 0.01) and birth weight SDS (r = –0.51, P < 0.001) (Fig. 1
). DHT to T ratio correlated to
height(adult-birth) SDS (r = 0.30, P < 0.05) and
weight(adult-birth) SDS (r = 0.39, P < 0.01) (Fig. 2
). For the SGA subjects only, we found negative correlations between DHT to testosterone ratio and
height(2yr-birth) SDS (r = –0.61, P < 0.01) (Fig. 3
).
|
The SGA group had higher inhibin B levels (164 vs. 137 pg/ml, P < 0.05) (Table 2
). Inhibin B levels correlated negatively with birth length SDS (r = –0.34, P < 0.01) (Fig. 4
) and birth weight SDS (r = –0.26, P < 0.05).
|
Adiponectin levels were lower in SGA males (5.40 vs. 6.35 µg/ml, P < 0.05) (Table 2
). Adiponectin correlated negatively with BMI (r = –0.43, P < 0.001). Leptin to adiponectin ratio was higher in the SGA group (0.66 vs. 0.38, P < 0.05) (Table 2
) and correlated with E2 to testosterone ratio (r = 0.41, P < 0.001) but not E2 levels. Leptin to adiponectin ratio also correlated with
weight(adult-birth) SDS (r = 0.47, P < 0.001).
Other findings
The levels of testosterone, free testosterone, DHEAS, SHBG, gonadotropins, LH to testosterone ratio, FSH to inhibin B ratio, and leptin did not differ between groups (Table 2
). As expected, leptin correlated with
weight(adult-birth) SDS (r = 0.35, P < 0.01) and BMI (r = 0.42, P < 0.001).
| Discussion |
|---|
|
|
|---|
Testosterone is the precursor of E2 via the enzymatic activity of aromatase. We therefore claim that elevated E2 to T ratio seen in males born SGA might be due to increased aromatase activity. The major source of E2 in adult men is aromatization of androgens in peripheral tissues (muscles and adipocytes), and only 15–25% is synthesized in testes (19, 20). Our finding of increased E2 levels in SGA males could be due to the subsequent weight gain and increased visceral fat mass after birth, although the investigated SGA subjects had normal weight and BMI, compared with controls. Indeed, E2 correlated positively to both gain in weight and height from birth to adulthood but not to adult weight or BMI. We lack data on body composition and fat mass. Serum levels of leptin may serve as an indirect measurement of fat mass because previous publications have shown positive correlations between body fat and serum levels of leptin (21). We found no difference in leptin levels between SGA males and controls, but leptin reflects only total fat mass and gives no information on fat distribution.
Interestingly, our results are in accordance to the findings in male rats exposed to a stressful intrauterine environment, showing features of the metabolic syndrome and increased plasma levels of E2 but normal levels of testosterone (10).
SGA males had increased DHT levels, a possible explanation to the clinical observation of early pubic hair in boys born SGA. DHT and DHT to T ratio correlated negatively to size at birth and interestingly also to early catch-up in height in the SGA group.
The increased levels of E2 and DHT in the SGA group may be caused by decreased capacity of kidney and liver to metabolize E2 and DHT because both E2 and DHT are excreted into urine and bile. This in turn might be due to impaired fetal organ size development after intrauterine growth restriction and does not have to reflect increased activity of aromatase and 5
-reductase.
We have also shown elevated levels of inhibin B in the SGA group. Previous studies have also described increased inhibin B levels in infant SGA boys (22) and SGA males with normal-sized testicles (9). Inhibin B is secreted from Sertoli cells in testes and has been suggested as a marker of spermatogenesis (23). However, as shown by both Klingmuller and Haidl (23), and Kumanov (24), inhibin B correlates more strongly to testicular size than sperm concentration and motility. We lack data on semen quality and therefore do not know whether the high inhibin B levels seen in SGA males reflect a high spermatogenic potential.
Increased levels of inhibin B are also found in nonobese females with polycystic ovary syndrome (25). These women have decreased insulin sensitivity and are more often born SGA (26).
In our study, SGA males had lower adiponectin levels. This in concordance to the findings by Cianfarani et al. (27), who found reduced adiponectin levels in children born SGA with catch-up growth. Adiponectin, an adipocyte-derived hormone, is decreased in states of insulin resistance (28). Furthermore, leptin to adiponectin ratio, an atherosclerotic index (29), was increased in SGA males. Leptin to adiponectin ratio was found to correlate with catch-up in weight and E2 to testosterone ratio.
It is not likely that the found differences are due to intraindividual variations because previous studies have shown the day-to-day variation to be negligible for E2, DHT, inhibin B, and adiponectin (30, 31, 32). In the present study, samples were collected between 0800 and 1000 h to reduce influence of diurnal variations.
A recent study found normal HPG function and E2 levels in adolescent males born SGA (33). Their population consisted of males who were born only slightly SGA (below –1.28 SDS in birth weight) and the definition of SGA was based only on birth weight. The studied males had a median age of 17 yr, and some may still have been in puberty (testicular volume 13.5–20 ml). E2 levels in males continue to increase until late puberty (34, 35). Moreover, the E2 results by Jensen et al. (33) seems to be overestimated because previous results from E2 measurements in late pubertal boys have shown only half the concentration (34, 35). The functional sensitivity in direct immunoassay may be far above the analytical detection limit of an E2 assay (36), giving falsely high E2 concentrations in boys and males. By use of an extraction immunoassay, the study by Jensen et al. may have resulted in a difference in E2 levels between the SGA group and controls.
The clinical relevance of the findings of increased E2 in SGA males with catch-up growth is still to be elucidated. Male transgenic mice with overexpression of aromatase have an increased risk of testicular cancer (37). Furthermore, low birth weight in boys is known to be associated with increased risk of testicular cancer, hypospadia, and cryptorchidism (8). We therefore hypothesize that increased aromatase activity could be part of the explanation to testicular dysgenesis syndrome.
We can only speculate whether the found increase in serum levels of estradiol is harmful or just a confounder coexisting with a pathological metabolic profile. The well-known protective effect of E2 against cardiovascular disease in women has not been found in men and could even be harmful. Interestingly, overexpression of the gene encoding for aromatase has been found to play some part in the development of atherosclerosis (38). In other words, increased aromatase activity in men may lead to increased cardiovascular disease, hypertension and decreased insulin sensitivity, diseases found in obese as well as nonobese males born SGA (39).
In conclusion, the current study is the first to report elevated serum levels of estradiol, dihydrotestosterone and inhibin B in SGA males of normal stature. We speculate that fetal growth restriction can lead to a permanent disturbance in the steroid biosynthesis. The mechanism remains to be clarified, and the clinical importance is uncertain. However, our findings should raise concern on what impact a deranged sex hormone profile, in men born SGA, might have on insulin sensitivity and later on cardiovascular disease.
| Acknowledgments |
|---|
| Footnotes |
|---|
Disclosure Statement: K.A. and C.A.-L. have received travel support from Novo Nordisk and H.F. from Ipsen. J.D. has received lecture fees and travel support from Novo Nordisk, Pfizer, Serono, and Ipsen and declares financial interest in Tercica, Inc.
First Published Online February 5, 2008
Abbreviations: BMI, Body mass index; CV, coefficient of variation; DHEAS, dehydroepiandrosterone sulfate; DHT, dihydrotestosterone; E2, 17β-estradiol;
height(adult-birth) SDS, difference in height SDS between birth and adult height;
height(2yr-birth) SDS, difference in height SDS between birth and 2 yr of age; HPG, hypothalamic-pituitary-gonadal; SDS, SD score; SGA, small for gestational age; T, testosterone;
weight(adult-birth) SDS, difference in weight SDS between birth and adult height.
Received August 3, 2007.
Accepted January 25, 2008.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
N. B. Ojeda, D. Grigore, and B. T. Alexander Developmental Programming of Hypertension: Insight From Animal Models of Nutritional Manipulation Hypertension, July 1, 2008; 52(1): 44 - 50. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |