| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Growth and Reproduction (G.T., M.B., J.H.P., K.M.M., N.E.S.), Radiology (M.G., K.D.), and Ultrasonography (V.B.), University Hospital of Copenhagen, Rigshospitalet, DK-2100 Copenhagen, Denmark; Department of Biostatistics (J.H.P.), University of Copenhagen, DK-2100 Copenhagen, Denmark; and Department of Environmental Medicine (T.K.J.), University of Southern Denmark, DK-5000 Odense, Denmark
Address all correspondence and requests for reprints to: Grete Teilmann, University Department of Growth and Reproduction GR-5064, Rigshospitalet, DK-2100 Copenhagen, Denmark. E-mail: gteilmann{at}rh.dk.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: The objective was to study clinical, biochemical, and ultrasonographic markers of pituitary-gonadal activation in prepubertal adopted girls and a control group in the same age categories.
Setting: The study took place at University Hospital, Copenhagen, Denmark.
Design and Participants: The study included randomly selected internationally adopted girls [(n = 99; mean age, 6.9 (5.1–8.5) yr] and controls of Danish origin [n = 93; mean age, 6.8 (5.2–8.5) yr] who were studied cross-sectionally.
Methods: Height, weight, and pubertal stage were assessed with serum levels of reproductive hormones. Size and morphology of internal genitals were evaluated by ultrasonography. Bone age was evaluated by x-ray of the left hand.
Results: Serum values of FSH were significantly higher in prepubertal adopted girls compared with controls [median, 1.4 (95% confidence interval, 0.4–3.6) vs. 1.0 (0.4–2.4) IU/liter; P <0.001]. Serum estradiol was above detection limit (>18 pmol/liter) in 46.5% of prepubertal adopted girls and 20.7% of controls (P = 0.001). In prepubertal adopted girls, the proportion of measurable samples increased significantly with age [odds ratio, 2.5 (95% confidence interval, 1.3–5.0; P = 0.009]. In controls, the odds ratio was 1.0 (0.6–1.7) (P = 0.9). Serum SHBG levels were significantly lower in prepubertal adopted girls compared with controls [99.0 (50.4–153.0) vs. 115.0 (53.1–202.1); P < 0.001].
Conclusion: Five- to 8-yr-old adopted girls showed signs of increased pituitary as well as gonadal activity despite prepubertal phenotype in the majority of girls. Our findings suggest that early onset of puberty in adopted girls is centrally driven.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The adopted girls (n = 99) were part of an ongoing, longitudinal cohort study of growth and pubertal maturation in internationally adopted girls. The cohort was recruited through the Danish National Board of Adoption, who randomly selected 635 families with adoptive daughters born from 1991 to 1999 to receive information about the study. Adopted girls (n = 276) aged 4–13 yr were enrolled in the longitudinal study in the autumn of 2003 and were followed over 2 yr with biannual examinations. Data concerning day of birth, accuracy of day of birth, birth weight, as well as age at adoption were obtained from the adoptive parents. In the present analysis, we report cross-sectional data from the 12-month follow-up examination for 99 internationally adopted girls aged 5.05–8.46 yr (mean, 6.92 yr). We selected the 12-month follow-up examination because the age distribution matched the control group (see below), and all ultrasonographic examinations were performed by the same examiner (Birgit Larsen, University Hospital of Copenhagen, Rigshospitalet) within a relatively short timespan (adopted girls, November 2004 to November 2005; control group, January 2006 to May 2006). The adopted girls originated from Asia (n = 83), South America (n = 12), former Eastern Europe (3), and Africa (1). Two eligible girls, who fulfilled the age criteria for the current analysis, were excluded 6–12 months previously because of precocious puberty treated with GnRH analog. Adopted girls were examined by G.T.
Control subjects
Danish-born girls (n = 93, in the following referred to as "controls") from the general population, who took part in a prospective birth cohort study, served as controls. Women of Danish origin were recruited consecutively in the first trimester of pregnancy, during the period of May 1997 to October 2001, from two university hospitals in Copenhagen. Information regarding maternal medical history, parity, and parturition were obtained from hospital records and questionnaires. The children were examined four times before the current examination (at birth and at 3, 18, and 36 months of age), and data concerning reproductive hormones have been published previously (9, 10). From this cohort, 185 randomly selected singleton girls were invited to participate in the present study, and positive response was received from 93 girls. Mean age was 6.81 yr and ranged from 5.15 to 8.53 yr. M.B. examined the control group.
Ethical approval
The study was performed according to the Helsinki II declaration and was approved by the local ethics committee (journal nos. KF01-081/03 and KF01-276 357) and The Danish Data Protection Agency (journal nos. 2003-41-3019 and 2005-41-5545).
Clinical examinations
All participants were examined using the same clinical and laboratory methods. Height (in centimeters) was measured to the nearest 0.1 cm using a stadiometer (Force Institute, Vejen, Denmark), and weight (kilograms) was measured on a digital weight scale with a precision of 0.1 kg (Biro). SD scores (SDS) for height and weight were calculated as SDS = ((observed value – age-specific mean value)/age-specific SD) based on Danish reference charts (11, 12). Body mass index (BMI) was calculated as kilograms per square meter. BMI Z-scores, based on Danish reference values, were calculated as z = (BMI/M)L – 1/(L x S), where M represents the median, S is the coefficient of variation (CV), and L is a measure of skewness, calculated for each age group (12). Pubertal maturation was assigned using the method of Tanner (13). Presence/absence of breast tissue was confirmed by palpation.
Interobserver analysis in 20 randomly selected participants showed 100% agreement in Tanner staging (breast development and pubic hair) between two observers (M.B. and G.T.).
Bone age determination
Bone age was evaluated in adopted girls, based on an x-ray examination on of the left hand taken at the 6-month follow-up examination. We used a Siemens (Munich, Germany) polymath 50 conventional x-ray system and an Agfa (Mortsel, Belgium) ADC system (analog to digital converting system). Bone age was evaluated using the method of Greulich and Pyle (14) in two independent readings. Examinations were primarily evaluated consecutively by two trained pediatric radiologist (K.D. or M.G.). A second independent reading was performed 12–18 months later by M.G. The radiologists were blinded to pubertal stage of the participant in both readings, as well as to chronological age at the second reading. The mean bone age of the two readings was used in the analysis.
Ultrasonographic examinations
Transabdominal ultrasonography of internal genitals was performed by the same specially trained nurse (Birgit Larsen) supervised by V.B. on an ACUSON Elegra scanner (Siemens) with a 4 MHz convex transducer. The ultrasound examinations were performed when the participant had a full bladder, obtained by voluntary urine retention after oral administration of fluids. The investigator was blinded to the current pubertal stage. Length, anteroposterior, and transverse diameter of the uterus and ovaries were measured. The number of follicles greater than or equal to 4 mm and the size of the largest follicle were registered. Thickness of the midline endometrium was measured if visible. Assuming the forms of an ellipsoid, volumes of the uterus and ovaries were calculated using the simplified formula of an ellipsoid: volume = d1x d2x d3 x 0.52, where d1, d2, and d3 are the three maximal longitudinal, anteroposterior, and transverse diameter. The mean volume of the two ovaries was calculated. Flow in the uterine artery recorded by Doppler was categorized as prepubertal (narrow systolic flow, without diastolic signal), early pubertal (continuous systolic and diastolic flow), or late pubertal (broad systolic flow waves and an uninterrupted flow signal during diastole) (15).
Laboratory analysis
Blood samples were drawn during daytime [morning: 0900 to 1100 h, n = 4 (n = 4 adopted girls, n = 0 controls); noon: 2400 to 1400 h, n = 95 (n = 47 adopted girls, n = 48 controls); afternoon: 1500 to 1700 h, n = 56 (n = 25 adopted girls, n = 31 controls); missing time information: 0900 to 1700 h, n = 4] from the antecubital vein after peripheral anesthesia for 1.5 h [EMLA (Astra Zeneca, London, UK), 25 mg lidocaine or 25 mg prilocaine]. A blood sample was obtained from 76 adopted girls (70 in breast stage B1, four in stage B2, and one in B3) and 83 Danish girls (82 in stage B1 and one in stage B2), and blood sampling was refused by the participant or failed in the remaining cases.
Serum FSH, LH, and SHBG were measured by time-resolved immunofluorometric assays (Delfia; PerkinElmer, Boston, MA) with detection limits of 0.06 and 0.05 IU/liter for FSH and LH, respectively, and 0.23 nmol/liter for SHBG. Intraassay and interassay CVs were less than 5% in both gonadotropin assays and less than 6% in the SHBG assay.
Estradiol was measured by RIA (Pantex, Santa Monica, CA). The detection limit was 18 pmol/liter, the intraassay CV was less than 8%, and the interassay CV was less than 13%. Serum inhibin B was measured in double antibody enzyme immunometric assays using monoclonal antibodies raised against either the ßA- or ßB-subunit, respectively, in combination with a labeled antibody raised against the inhibin 
-subunit (Oxford Bio-innovation, Oxford, UK). The detection limit was 20 pg/ml. Intraassay and interassay CVs were less than 15% and less than 18%, respectively.
Serum IGF-I was measured by RIA as described previously (16) with a detection limit of 21 ng/ml and intraassay and interassay CVs of less than 11%. Serum IGF binding protein-3 was also measured by RIA (Mediagnost, Tübingen, Germany) with a detection limit of 300 ng/ml, an intraassay variation less than 6% and an interassay variation of less than 9%. Data concerning IGFBP-3 are not shown.
Statistics
Height, weight, and BMI distributions are summarized using mean ± SD stratified in 1-year age groups. Differences between adopted girls and controls were tested with unpaired t test. The skewed distributions of the reproductive hormones as well as for dimensions of ovaries and uterus are summarized using medians and 2.5–97.5 percentiles. Serum hormone concentrations, as well as uterus measurement, in the groups were compared using Mann-Whitney tests. To assess a possible age dependency, a general linear model was used for the log-transformed concentrations. For LH, estradiol, and inhibin B, a proportion of samples were below the detection limit (i.e. LH, <0.05 IU/liter; estradiol, <18 pmol/liter; and inhibin B, <20 pg/ml). In these cases, the value was set to the detection limit. For these hormones, a logistic regression model for the prevalence of measurable vs. unmeasurable samples was used to assess the age dependence. The result of these analyses are odds ratios (OR) describing the change per year in the prevalence of measurable samples. Age dependency of the number of follicles in the two groups was described using Goodmann’s 
, a correlation coefficient for ordinal data.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
). Two adopted girls aged 7 yr were in stage B2, and in the group of 8-year-old adopted girls three were in stage B2 and one was in B3. One 8-yr-old girl from the control group was in stage B2. The proportion of girls with palpable breast tissue did not differ significantly between the groups (6 of 99 adopted girls and 1 of 93 controls; P = 0.07). However, two adopted girls were excluded 6–12 months earlier because of GnRH-treated precocious puberty. Thus, the true proportion of adopted girls younger than 8.5 yr old with breast development was 8 of 101 (7.9%).
|
.
|
Serum values of LH were not significantly different in prepubertal adopted girls compared with controls (P = 0.5). A high proportion of LH samples was below the detection limit (0.05 IU/liter) in both groups (adopted girls, 45.7%; controls, 54.3%; P = 0.24). In prepubertal adopted girls, the proportion of measurable samples increased significantly with age [OR, 2.5 (95% CI, 1.3–5.0); P = 0.009]. In controls, the OR was 1.0 (0.6–1.7) (P = 0.9). In adopted girls, LH levels were significantly higher in girls in stage B2+ compared with prepubertal subjects (P = 0.004).
Estradiol levels were higher in prepubertal adopted girls [mean, 22.1; median, 18.0 (2.5–97.5 percentiles, 18.0–37.3) pmol/liter] compared with controls [mean, 19.3; median, 18.0 (18.0–30.8) pmol/liter; P = 0.001]. Serum estradiol was above the detection limit (>18 pmol/liter) in 46.5% of blood samples from adopted girls in B1 and in 20.7% of samples in the control group (P = 0.001). The proportion of measurable samples did not change statistically significantly with age in adopted girls [OR, 1.4 (95% CI, 0.9–2.4); P = 0.17] or in controls [OR, 1.1 (0.6–1.9); P = 0.7]. Serum estradiol tended to be higher in adopted girls in B2 compared with adopted girls in B1 (P = 0.06).
Inhibin B was measurable (>20 pg/ml) in 26.3 and 25.3% of blood samples from adopted girls and controls, respectively (P = 0.8). The proportion of measurable samples did not change with age in any of the groups, and inhibin B levels did not differ between girls in B1 and B2.
Serum SHBG levels were significantly lower in prepubertal adopted girls compared with the control group [99.0 (50.4–153.0) vs. 115.0 (53.1–202.1); P < 0.001]. SHBG tended to be lower in adopted girls in B2+ compared with adopted girls in B1 (P = 0.06).
There was no difference in serum levels of IGF-I between adopted girls and controls (P = 0.29). Adopted girls in B2+ had significantly higher IGF-I levels compared with prepubertal adopted girls [268.0 (215.0–389.0) vs. 183.0 (77.6–314.0); P = 0.001].
There was no difference in the distribution of time of sampling between adopted girls and controls (P = 0.1), and hormonal values were independent of time of blood sampling.
Birth weight was known in 55 adopted girls [mean, 2.7 (minimum-maximum, 1.3–4.1) kg]. Birth weight was less than 2.5 kg in 47% of cases. Mean age at adoption was 1.3 ± 0.85 yr. A total of 80.2% were adopted before 2 yr of age. There was no difference in serum levels of any hormones in any age group between girls depending on birth weight (> or <2.5 kg) or age at adoption (> or <2 yr of age) (P 
0.1). Day of birth was considered certain in 66.7% of cases and uncertain in 19.2%, and 14.1% did not know. Differences in FSH and estradiol levels remained significant when including only girls with certain day of birth in the analysis.
Height and weight were significantly lower in adopted girls compared with Danish girls in all age groups (P 
0.005), except for weight in 8-yr-old girls (Table 1
). Mean ± SD height SDS was –0.80 ± 1.3 and –0.36 ± 1.1 in the groups of 5- and 8-yr-old adopted girls, respectively. Weight SDS was significantly higher in older adopted girls (P = 0.012). There was no significant difference in BMI between the two groups, but the proportion of obese children (BMI SDS > 2) was higher among controls compared with adopted girls (controls, 6 of 93; adopted girls, 1 of 99; P = 0.045).
Bone age was normal within every age group for adopted girls (Table 1). Bone age was delayed more than 1 yr in 12 adopted girls, within ±1 yr from chronological age in 62 cases, and advanced more than 1 yr in 23 girls. Median bone advancement was 0.2 (–2.2 to 2.4) yr and 1.6 (–1.4 to 2.2) yr in girls in B1 and B2+, respectively (P = 0.2). Serum levels of IGF-I were significantly higher in girls with advanced bone age compared with girls with delayed bone age [227.5 (134–358) vs. 153.0 (78.0–296.0); P = 0.01]. SHBG was significantly lower in girls with advanced bone age compared with girls with delayed bone age [117.9 (83.0–140.0) vs. 81.5 (44.0–142.0); P = 0.003]. For all other hormones, we found no significant effects of bone age maturation.
Uterine length, anteroposterior, and transverse diameters were identified in all but two cases. All three dimensions as well as uterine volumes are shown in relation to age, pubertal stage, and adoption status in Fig. 2. Median uterine volume was 1.14 ml (2.5–97.5 percentiles, 0.32–5.25) in adopted girls and 0.69 ml (0.27–3.19) in controls. Uterine length, transverse diameter, and volume were significantly larger in prepubertal adopted girls compared with controls (P < 0.001), but there was no difference in the anteroposterior diameter. All three uterine dimensions as well as uterine volume were significantly larger in adopted girls in stage B2+ compared with prepubertal adopted girls (P < 0.02).
|
).
Both ovaries were visualized in 96% of cases. The means of right and left ovary volumes [mean ovarian volume (MOV)] are shown in relation to age, pubertal stage, and adoption status in Fig. 2. Median MOV was 0.74 ml (0.30–3.00) in adopted girls and 0.67 ml (0.29–2.70) in controls (P = 0.28). MOV changed significantly with age in both groups (P < 0.05).
The number of follicles greater than 4 mm in each ovary varied from 0 to 3. There was a weak association between age and group and the total number of follicles in prepubertal adopted girls ( of 0.4; P = 0.07) but not in controls ( of 0.2; P = 0.32). There was no difference in the age dependence between the two groups (P = 0.25).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Pulsatile gonadotropin secretion, particularly during night, has long been recognized in relation to the onset of puberty, and circadian rhythms of FSH and LH secretion have been demonstrated in prepubertal girls aged 5–6 yr (17). We found no effect of time of blood sampling, possibly because 97.5% of samples were collected after 1200 h. It was shown recently that LH pulses were followed by estradiol secretion with a median lag time of 5 h in prepubertal girls who entered puberty within 1 yr (17). Single values of FSH and LH obtained during daytime, as in our study, may therefore underestimate 24-h levels of gonadotropins, whereas estradiol may be a more robust indicator for early puberty. This might explain why we found higher levels of estradiol in adopted girls but similar levels of LH between the groups. Inhibin B is considered a marker of ovarian granulosa cell activity, stimulated by FSH (18). We found no significant differences in serum levels of inhibin B between the groups, but a high proportion of the samples were below the detection limit of 20 pg/ml. Biologically important differences may be concealed in the unmeasurable range of the assay.
In normal puberty, SHBG declines as puberty progresses, regulated by sex steroids as well as insulin (19, 20). We found significantly lower levels of SHBG in adopted girls than in controls. Our finding might be a component of early sexual maturation, and it can be speculated that the estrogen/androgen balance might be different between adopted girls and controls. In addition, nutritional factors, not reflected in the simple calculation of the BMI, such as central adiposity, low lean body mass, or hyperinsulinaemia, may contribute to low levels of SHBG among adopted girls.
We found no difference in IGF-I levels between the groups. Theoretically, higher IGF-I levels could have been expected in adopted girls as a part of early maturation, because IGF-I levels normally rises in puberty (16). The effect of early maturation might, however, be counteracted by low levels of circulating IGF-I attributable to short stature and low growth velocity, which was seen in a high proportion of adopted girls.
We found significant differences in uterus volume between controls and adopted girls, mainly attributable to differences in uterine length. Determination of uterine length in the evaluation of puberty has been questioned (15), because the organ is very small and the cervix and the uterus have the same length and are difficult to distinguish (21). The observed difference in uterine length might therefore be an artifact. However, it may also indicate a physiological effect of the higher estradiol levels. Previous ultrasonographic studies have suggested that a flow pattern displaying systolic as well as diastolic flow can confirm the onset of puberty (15, 22). We could not reproduce this observation. In contrast, control girls showed this flow pattern more frequently than adopted girls. We speculate that controls had larger body sizes (higher weights and heights), which could affect the flow in the uterine arteries.
Higher FSH levels in the adopted girls are most likely a component of early physiological maturation. It was proposed previously that early puberty could be caused by exposure to persistent environmental pollutants such as endocrine disrupters (1, 23, 24). In our study, all girls with clinically palpable breast development showed a central activation of the pituitary-gonadal axis, and uterine as well as ovarian volumes were higher than in prepubertal girls. Thus, a direct stimulatory effect of exogenous hormones is not likely to have caused early breast development in our population, although it cannot be excluded that preadoption exposure to endocrine disrupters can trigger early pituitary maturation (1).
American studies have indicated a recent trend toward earlier onset of puberty in girls (25). It has been suggested that increased body fat might contribute to advance the onset of puberty (26), possibly mediated by leptin or other adipocyte-derived hormones. In the present study, we found no differences in BMI SDS between adopted girls and controls. This suggests that other factors than body fat are likely to explain earlier sexual maturation in adopted girls. Weight SDS was significantly higher in adopted girls of 8 yr of age compared with the group of 5-yr-old adopted girls. It can be speculated that rapid weight gain during childhood might be a more important trigger of puberty than the absolute content of body fat (27). However, no firm conclusion can be drawn based on these data because they are cross-sectional.
We found a high frequency of low birth weight (47% <2.5 kg) among adopted girls. Our clinical examinations revealed no signs of premature adrenarche. Future studies including analysis of adrenal androgens will show whether alterations in the pituitary-adrenal axis contribute to early pubertal maturation as has been observed in other populations (28).
In conclusion, we found that adopted girls showed signs of increased pituitary and gonadal activity despite prepubertal phenotype. We hypothesize that these early markers precede an early onset of clinical puberty. The reason for early pubertal maturation in adopted girls remains to be clarified.
|
Footnotes |
|---|
First Published Online May 1, 2007
Abbreviations: BMI, Body mass index; CV, coefficient of variation; MOV, mean ovarian volume; OR, odds ratio; SDS, SD scores.
Received September 25, 2006.
Accepted April 19, 2007.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J.-C. Carel, E. A. Eugster, A. Rogol, L. Ghizzoni, M. R. Palmert, and on behalf of the members of the ESPE-LWPES GnRH An Consensus Statement on the Use of Gonadotropin-Releasing Hormone Analogs in Children Pediatrics, April 1, 2009; 123(4): e752 - e762. [Abstract] [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 |
).