Central precocious puberty in girls: an evidence-based diagnosis tree to predict central nervous system abnormalities

Central precocious puberty in girls: an evidence-based diagnosis tree to predict central nervous system abnormalities

Martin Chalumeau

In girls, precocious puberty is classically defined as breast development before the age of 8 years. (1-5) Given the secular trend toward earlier puberty, the percentage of girls with precocious puberty rose from 2.5% in 1969 (6) to 10% in the 1990s. (7,8) When precocious puberty is secondary to the activation of the hypothalamic–pituitary–gonadal axis, it is called central precocious puberty (CPP). CPP accounts for 89% to 98% of the cases of precocious puberty in published series. (9-12) CPP can result from central nervous system (CNS) abnormalities, mainly hamartomas and hypothalamic tumors. Brain imaging is thus systematically recommended in girls with CPP. (9,10,13-16)

Clinicians who deal with CPP have 2 main concerns: detection of CNS abnormalities and short final height. The risk for girls with CPP to reach a final height that is too short has been widely debated, and several authors have suggested limiting the use of gonadotropin-releasing hormone analog. (4,5,17) Thus, a means of identifying the group of girls who have CPP and are at high risk of CNS abnormalities is essential. In hospital-based studies, CPP revealed CNS abnormalities in 8% to 33% of the girls. (2,9-11,13,15,16,18) Reported predictors for CNS abnormalities in girls with CPP are young age at the onset of puberty, (13,15,19) presence of pubic hair at the same time, (13) markedly advanced bone age, (15,19) rapid progression of puberty, (15,19) and high plasma gonadotrophins concentrations. (13) Unfortunately, those studies were based on small series with high frequencies of CNS abnormalities suggestive of selection bias (13,15) or did not include appropriate statistical analysis. (19) New recommendations from the Lawson Wilkins Pediatric Endocrine Society (LWPES) (20) and Elders et al (21) were proposed to identify girls who are at high risk (for both CNS abnormalities and short final height), but they were mainly based on a study in which the cause of precocious puberty was not assessed. (7)

The objectives of the present study were to evaluate the accuracy of the recommendations made by the LWPES for our patients; to identify independent predictors (ie, risk factors) for CPP revealing CNS abnormalities in affected girls; and to construct and validate a diagnosis tree, based on independent predictors, aimed at identifying girls who have CPP and require brain imaging.



All girls were seen by one of us (R.B.) between May 1982 and December 2000 at Hopital Necker-Enfants Malades, Paris, France. This tertiary university pediatric hospital is 1 of the 4 referral centers for pediatric endocrinology in the Paris metropolitan area.

CPP was diagnosed on the basis of evidence of breast development before the age of 8 years. Premature thelarche and primary gonadal or adrenal diseases were excluded according to the diagnostic criteria described by Palmert et al. (22) Patients with CPP secondary to a previously diagnosed CNS abnormality were excluded.


Data on 20 clinical, radiological, and biological variables were collected retrospectively from the patients’ records. The age at the onset of puberty was defined as the age when breast development was first noted by the patient or her parents. Other data were the values recorded at the first physical examination. Standing height was measured with a Harpenden stadiometer. Body mass index (BMI) was calculated and expressed as a z score with respect to chronological age. (23) Growth rate was expressed as standard deviation (SD) for chronological age. (24) Breast and pubic hair development were rated according to Tanner stages. (6) For axillary hair, we used the same 3-stage scale as others. (7) Bone age was assessed by 1 senior pediatric endocrinologist (R.B.). (25) Estradiol ([E.sub.2]) was extracted from plasma with ether and measured by radioimmunoassay (Estradiol 2; Sorin Biomedica, Antony, France). The hypothalamic–pituitary–gonadal axis was investigated by measuring basal plasma luteinizing hormone (LH) and follicle-stimulating hormone (FSH) concentrations and peak stimulated (gonadotropin-releasing hormone, 100 [micro]g/[m.sup.2] intravenously) LH and FSH concentrations. The LH/FSH peak ratio and the difference ([DELTA]) between peak and basal LH and FSH values were calculated.

Patients were classified as having idiopathic CPP or CPP revealing a CNS abnormality, depending on the normality of computed tomography and/or magnetic resonance imaging of the brain and pituitary region.

Data Analysis

Statistical analyses were performed using BMDP software (BMDP Statistical Software, Los Angeles, CA) and EpiInfo software (Centers for Disease Control and Prevention, Atlanta, GA). Variables were selected when <10% of the data were missing for girls with idiopathic CPP and no data were missing for those with CPP revealing a CNS abnormality. Continuous variables were dichotomized on the basis of the rounded-off quartiles of the distribution among idiopathic CPP. The quartile with the highest frequency of CNS abnormalities was compared with the rest of the population. Tanner stages were dichotomized around the first stage of the distribution to ensure easy external reproducibility. Univariate analysis was performed using the 2-tailed Fisher exact test to evaluate the association between potential predictors and the detection of CNS abnormalities. Multivariate analysis using logistic regression (backward stepwise procedure) was conducted by entering variables identified by univariate analysis as being associated with CNS abnormalities with a degree of significance <0.20. P < .05 was considered statistically significant.

Diagnosis-Tree Construction

All of the above analyses were performed on a pilot set of data corresponding to patients included between 1982 and 1998. A diagnosis tree with an a priori objective of 100% sensitivity was constructed. The variables that were independently associated with CNS abnormalities were selected for the construction of the diagnosis tree. The order of appearance of the variables in the tree followed the order of data collection in clinical practice (general data followed by clinical signs then biological values). Cutoffs were modified to obtain 100% sensitivity for the diagnosis of CNS abnormalities in the pilot population and to coincide with routine clinical practice. The stability of the crude relationship with the newly modified cutoff was verified. Once 100% sensitivity was obtained (and its corollary, 100% negative-predictive value), the specificity and positive-predictive value of the diagnosis tree were calculated for the pilot population.

Diagnosis-Tree Validation

The diagnosis tree was then applied to the validation set of patients, included between 1999 and 2000. The investigator who constructed the tree (M.C.) was not aware of the values of the variables in the validation population. The diagnosis tree was not modified after it was first applied to the validation data set. The sensitivity, specificity, and positive and negative predictive values of the diagnosis tree were calculated for the validation population.


Pilot Population

During the period corresponding to the pilot group, 197 girls met the inclusion criteria. Among them, 11 (6%) had CPP that revealed a CNS abnormality: hamartoma (n = 6), glioma (n = 3), angioma cavernosum (n = 1), and suprasellar arachnoid cyst (n = 1). Age distribution according to causative diagnosis is presented in Fig 1.


CPP revealed a CNS abnormality (1 glioma, 1 hamartoma) in 2 white girls (18%) who would not have been considered as requiring brain imaging according to the LWPES’s recommendations. (20) Indeed, their puberty began after 7 years of age and was not unusually rapid: their growth rates were -1.8 and 0.5 SD, and breast development was Tanner stage 3 for both. Their bone age advances were 150 cm (161 and 159 cm), and the difference between their predicted and genetic target heights were shorter than -10 cm (-2 and -1 cm). Neither had headaches, seizures, focal neurologic deficits, or adversely affected emotional states. The patient with glioma also had a growth hormone deficiency that explained her slow growth rate and minor bone age advance.

Univariate Analyses

According to univariate analyses (Table 1), the main associations with CNS abnormalities were age at onset of puberty 110 pmol/L, peak FSH >20 U/L, and [DELTA]FSH >15 U/L. No significant associations were found with growth rate >3 SD, BMI 1 U/L, basal FSH >5 U/L, peak LH >15 U/L, LH/FSH peak ratio >1, and [DELTA]LH >15 U/L.

Multivariate Analyses

After adjustment by logistic regression performed with the variables of 188 patients (including all patients with CPP revealing CNS abnormalities), only 3 variables remained independently associated with CNS abnormalities: age at onset of puberty 110 pmol/L (AOR: 4.1; 95% CI: 1.0-17; P = .04).

Diagnosis-Tree Construction

The sensitivities and the specificities of the 3 independent predictors for CNS abnormalities used alone are reported in Table 2. None alone or in combination (data not shown) reached the required 100% sensitivity. Thus, we modified the initial cutoffs. The 6-year age cutoff was retained because it did not seem clinically realistic to propose avoiding evaluation under this age. The cutoff around Tanner stage 1 was kept to ensure easy reproducibility. It was necessary to lower the plasma [E.sub.2] threshold to <54 pmol/L to obtain the required 100% sensitivity in combination with age. It was lowered to 45 pmol/L in light of the intermeasurement variability (SD = 3.9 pmol/L). This value represented the 45th percentile of girls with idiopathic CPP. With this

threshold, the association of [E.sub.2] with CNS abnormalities was still significant (OR: 8.3; 95% CI: 1.1-179; P = .03). The lack of pubic hair was no longer helpful and was not used in the final diagnosis tree.

In the high-risk part of the first branch of the final diagnosis tree (ie, patients <6 years), 19% (8 of 42) of CPP revealed CNS abnormalities (Fig 2). We thought we had identified a very high-risk population for whom brain imaging should systematically be performed, and thus we stopped the segmentation of this branch. In the low-risk branch, a second segmentation was applied using the 45-pmol/L [E.sub.2] cutoff. The final decision tree had 38% specificity and 9% positive predictive value (Table 2). By using it to select patients for brain imaging, no CNS abnormality would have been missed and 67 (38%) noncontributive imagings would have been avoided.


Diagnosis-Tree Validation

Among the 42 girls with CPP in the validation population, CPP revealed a CNS abnormality (hamartoma) in 3 (7%). The new sensitivities and specificities of each of the 3 predictors alone were calculated for these girls (Table 2). The diagnosis tree had 100% sensitivity, 100% negative predictive value, 56% specificity, and 15% positive predictive value (Fig 3, Table 2). By applying it to select patients for brain imaging, no CNS abnormality would have been missed and 22 (56%) noncontributive imagings would have been avoided.



We confirmed that the principal predictor of a CNS abnormality in girls with CPP was young age at the onset of puberty, as was suspected by others. (13,15,19) However, this predictor cannot be used alone: extremely early CPP can be idiopathic, and CPP between 7 and 8 years of age can be the only element revealing CNS abnormality, as shown in our patients. We also confirmed that girls with CNS abnormalities had more biologically advanced puberty with significantly higher [E.sub.2], peak and [DELTA]FSH concentrations and tendencies toward significantly higher basal LH and FSH and peak and [DELTA]LH concentrations. (13)

A new simple clinical predictor of CNS abnormalities in girls with CPP was identified in our study: the lack of pubic hair at the time of diagnosis. The relationship between the lack of pubic hair and a CNS abnormality remained strong after taking into account potential confounders, such as age or biological findings. Cacciari et al (13) found an opposite result based on a very small population (n = 15). Our finding is in agreement with a former description of girls with CPP as a result of hamartoma without evidence of adrenarche. (26) LH-releasing hormone-containing fibers have been observed within hamartoma tissues in patients with CPP. (27,28) However, the relationship between the lack of pubic hair at diagnosis and CNS abnormalities other than hamartoma was also strong and significant (OR: 20.7; 95% CI: 2.1-500; P = .002). Thus, the direct secretion of LH-releasing hormone by CNS abnormalities may not be the only explanation for the dissociation between gonadarche and adrenarche in patients with CPP caused by a brain tumor or malformation.

A new approach to identify girls who have CPP and require brain imaging has been described herein. This selection is based on 2 simple and reproducible criteria: age and plasma [E.sub.2] concentrations. Our diagnosis tree was constructed with the aim of obtaining 100% sensitivity. As the classical approach to the diagnosis of CPP includes systematic brain imaging, (9,10,13-16) it did not seem ethical to us to propose a diagnosis tree with sensitivity below 100%. Given this imposed sensitivity and the lack of highly discriminate criteria, the specificity is low (approximately 40%-55%), but these levels of sensitivity and specificity offer the opportunity to avoid safely one third to one half of brain imagings.

We did not use automatic strategies to create our diagnosis tree by segmentation analysis, such as those provided by software (eg, Answer Tree software [SPSS Inc, Chicago, ILl). (29) We excluded variables that were not independently associated with CNS abnormalities to avoid the use of nonrobust predictors. The order of appearance of variables in the diagnosis tree was modeled on the order of data collection in clinical practice. The cutoff values were imposed to fit with clinical concerns (age at onset of puberty), to offer good clinical reproducibility (Tanner stage for pubic hair), or to ensure 100% sensitivity ([E.sub.2]).

Because of the retrospective design of our study, some variables (ethnicity, age at menarche of patient’s mother, weight gain, etc) had too many missing data and were excluded from the analysis. This exclusion did not modify the univariate analysis results, but multiple logistic regression might have suffered from incomplete adjustment.

The frequency of CNS abnormalities (6%) and the histologic distribution of lesions found in our study are close to that (8%) found in the large study by the Italian collaborative group. (9) This frequency is much lower than those reported in older smaller studies, which probably suffered from recruitment bias. (2,10,11,13,15,16,18) However, because our population consisted of girls with CPP rather than girls who had breast development, were younger than 8 years, and were seen in office-based settings, our results cannot be generalized to this latter population.

A prognosis or diagnosis model should not be applied in clinical practice before a well-defined validation process has been completed. (30) We conducted a first external validation using a small sample of patients from the same center as the pilot population. The next step will be to test these results on a large sample of patients seen outside our hospital. The need for external validation is particularly true for models involving puberty for which marked racial variations exists. (7,8) The validation should start with the evaluation of the stability of the predictors established in our patients. For example, an attempt was made to validate our findings using a population of girls with CPP in Salvador-Bahia, Brazil. It was unsuccessful because their age at onset of puberty was younger, their pubic hair was more prevalent, and their plasma [E.sub.2] concentrations were higher than in our patients. However, the advantage of our diagnosis tree is its simplicity. Clinicians can modify thresholds according to the distributions in their populations of girls with CPP. However, each change must be validated before being routinely applied in clinical practice.

There is increasing concern about the quality of practice guidelines developed by specialty societies. (31) The LWPES’s recommendations were formulated to identify girls who have CPP and need evaluation for risk of both CNS abnormalities and short final height. (20) Their conclusions were derived mainly from a study in which the cause of precocious puberty was not assessed. (7) Those new recommendations raised some concerns. (31-33) Given that they were designed to apply to American girls, we found that 2 of our 11 French patients with CPP revealing CNS abnormalities (including 1 glioma) would not have been identified as requiring brain imaging. Among 163 Italian girls who had CPP and were between the ages of 7.0 and 7.9 years, reported by Cisternino et al, (9) 2 hamartomas and I tumor of the fourth ventricle were revealed by CPP. The real impact of the LWPES’s recommendations should now be tested on American girls.

We agree with the LWPES that the age threshold of 8 years for breast development is based on outdated studies, (20) as clearly demonstrated in the United States by Herman-Giddens et al (7) and in China by Huen et al. (8) It is probably true in many other countries. We also agree that a systematic approach using invasive dynamic biological tests and brain imaging in all cases of breast development before 8 years of age is not cost-effective. Thus, there is an urgent need for population-based studies including biological work-up and brain imaging to exclude premature thelarche, primary gonadal precocious puberty, and CPP revealing CNS abnormalities. Predictors for short final height and/or CNS abnormalities, identified by rigorous statistical analysis, need to be established.

TABLE 1. Univariate Analysis: Relationships Between Potential Clinical

or Radiological Predictors and CNS Abnormalities in Girls with CPP


Variable Idiopathic Abnor- OR 95% CI P Value

CPP mali-

(n = 186) ties *

(n = 11)

n % n %

Age at onset of

puberty (y)

<6 34 18.3 8 72.7 11.9 2.7-61 <[10.

[greater than or 152 81.7 3 27.3 1 sup.-3]

equal to] 6

Growth rate (SD)

>3 41 24.4 4 36.4 1.8 0.4-7.3 .50

[less than or 127 75.6 7 63.6 1

equal to] 3

BMI (z score)

<0.5 43 23.1 5 45.5 2.8 0.7-11.0 .14

[greater than or 143 76.9 6 54.5 1

equal to] 0.5


(Tanner stage)

2 88 47.6 7 63.6 1.9 0.5-8.3 .30

>2 97 52.4 4 36.4 1

Pubic hair

(Tanner stage)

1 30 16.2 7 63.6 9.0 2.2-40 <[10.

>1 155 83.8 4 36.4 1 sup.-3]

Axillary hair



1 127 68.6 9 81.8 2.1 0.4-14 .51

>1 58 31.4 2 18.2 1

Advanced bone

age (y)

[less than or 39 21.7 1 9.1 2.8 0.3-60 .50

equal to] 0.5

>0.5 141 78.3 10 90.9 1

[E.sub.2] (pmol/L)

>110 39 22.0 6 54.5 4.3 1.1-17 .02

[less than or 138 78.0 5 45.5 1

equal to] 110

Basal LH (U/L)

>1 54 31.8 7 63.6 3.8 0.9-16 .05

[less than or 116 68.2 4 36.4 1

equal to] 1

Basal FSH (U/L)

>5 55 32.2 7 63.6 3.7 0.9-16 .05

[less than or 116 67.8 4 36.4 1

equal to] 5

Peak LH (U/L)

>15 47 25.3 6 54.5 3.6 0.9-14 .07

[less than or 139 74.7 5 45.5 1

equal to] 15

Peak FSH (U/L)

>20 32 17.2 6 54.5 5.8 1.4-24 .008

[less than or 154 82.8 5 45.5 1

equal to] 20

LH/FSH peak ratio

>1 50 26.9 5 45.5 2.3 0.6-8.1 .18

[less than or 136 73.1 6 54.5 1

equal to] 1


>15 42 24.7 6 54.5 3.7 0.9-15 .07

[less than or 128 75.3 5 45.5 1

equal to] 15


>15 33 19.3 6 54.5 5.0 1.2-21 .01

[less than or 138 80.7 5 45.5 1

equal to] 15

* CPP revealing a CNS abnormality.

([dagger]) According to Hermann-Giddens et al. (7)

TABLE 2. Prediction of CNS Abnormalities in Girls With CPP, in the

Pilot (n = 197) and Validation (n = 42) Populations, Using Either

Independent Predictors Alone or the Diagnosis Tree

Model Characteristic (%) Age <6 Lack of

(Years) Pubic Hair

P * V P V


Sensitivity 73 67 64 100

Specificity 82 85 84 76

Positive predictive value 19 25 19 25

Negative predictive value 98 97 98 100

Model Characteristic (%) [E.sub.2] (pmol/L)

>110 >45


Sensitivity 55 67 91 67

Specificity 78 ([double 87 45 ([double 69

dagger]) dagger])

Positive predictive value 13 29 9 14

Negative predictive value 97 97 99 96


Model Characteristic (%) Tree


Sensitivity 100 100

Specificity 38 56

Positive predictive value 9 15

Negative predictive value 100 100

* Pilot population.

([dagger]) Validation population.

([double dagger]) n = 188, [E.sub.2] was unknown for 9 girls (5%) with

idiopathic CPP.


This study was supported by the Groupe de Recherches Epidemiologiques en Pediatrie. M.C. was financially supported by a grant from the Societe Francaise de Pediatrie. Funding was received from the Association des Juniors en Pediatrie and Laboratoire Gallia to translate the manuscript.


(1.) Kaplan SL, Grumbach MM. Pathophysiology and treatment of sexual precocity. J Clin Endocrinol Metab. 1990;71:785-789

(2.) Lebrethon MC, Bourguignon JP. Management of central isosexual precocity: diagnosis, treatment, outcome. Curr Opin Pediatr. 2000;12: 394-399

(3.) Merke DP, Cutler GB. Evaluation and management of precocious puberty. Arch Dis Child. 1996;75:269-271

(4.) Klein KO. Precocious puberty: who has it? Who should be treated? J Clin Endocrinol Metab. 1999;84:411-414

(5.) Rosenfield RL. Selection of children with precocious puberty for treatment with gonadotropin releasing hormone analogs. J Pediatr. 1994;124: 989-991

(6.) Marshall WA, Tanner JM. Variations in pattern of pubertal changes in girls. Arch Dis Child. 1969;44:291-303

(7.) Herman-Giddens ME, Slora EJ, Wasserman RC, et al. Secondary sexual characteristics and menses in young girls seen in office practice: a study from the Pediatric Research in Office Settings Network. Pediatrics. 1997; 99:505-512

(8.) Huen KF, Leung SS, Lau JT, Cheung AY, Leung NK, Chiu MC. Secular trend in the sexual maturation of southern Chinese girls. Acta Paediatr. 1997;86:1121-1124

(9.) Cisternino M, Arrigo T, Pasquino AM, et al. Etiology and age incidence of precocious puberty in girls: a multicentric study. J Pediatr Endocrinol Metab. 2000;13:695-701

(10.) Garagorri JM, Chaussain JL, Job JC. Precocious puberty in girls. Study of 98 cases. Arch Fr Pediatr. 1982;39:605-611

(11.) Pescovitz OH, Comite F, Hench K, et al. The NIH experience with precocious puberty: diagnostic subgroups and response to short-term luteinizing hormone releasing hormone analogue therapy. J Pediatr. 1986;108:47-54

(12.) Bridges NA, Christopher JA, Hindmarsh PC, Brook CG. Sexual precocity: sex incidence and aetiology. Arch Dis Child. 1994;70:116-118

(13.) Cacciari E, Frejaville E, Cicognani A, et al. How many cases of true precocious puberty in girls are idiopathic? J Pediatr. 1983;102:357-360

(14.) Grosvalet A, Ernest C, Diebler C, Sauvegrain J. Computed tomography in precocious puberty of central origin. Ann Radiol (Paris). 1981;24:32-38

(15.) Kornreich L, Horev G, Blaser S, Daneman D, Kauli R, Grunebaum M. Central precocious puberty: evaluation by neuroimaging. Pediatr Radiol. 1995;25:7-11

(16.) Rieth KG, Comite F, Dwyer AJ, et al. CT of cerebral abnormalities in precocious puberty. AJR Am J Roentgenol. 1987;148:1231-1238

(17.) Leger J, Reynaud R, Czernichow P. Do all girls with apparent idiopathic precocious puberty require gonadotropin-releasing hormone agonist treatment? J Pediatr. 2000;137:819-825

(18.) Lyon AJ, De Bruyn R, Grant DB. Isosexual precocious puberty in girls. Acta Paediatr Scand. 1985;74:950-955

(19.) Cassio A, Cacciari E, Zucchini S, Balsamo A, Diegoli M, Orsini F. Central precocious puberty: clinical and imaging aspects. J Pediatr Endocrinol Metab. 2000;13(suppl 1):703-708

(20.) Kaplowitz PB, Oberfield SE. Reexamination of the age limit for defining when puberty is precocious in girls in the United States: implications for evaluation and treatment. Drug and Therapeutics and Executive Committees of the Lawson Wilkins Pediatric Endocrine Society. Pediatrics. 1999;104:936-941

(21.) Elders MJ, Scott CR, Frindik JP, Kemp SF. Clinical workup for precocious puberty. Lancet. 1997;350:457-458

(22.) Palmert MR, Malin HV, Boepple PA. Unsustained or slowly progressive puberty in young girls: initial presentation and long-term follow-up of 20 untreated patients. J Clin Endocrinol Metab. 1999;84:415-423

(23.) Rolland-Cachera MF, Cole TJ, Sempe M, Tichet J, Rossignol C, Charraud A. Body mass index variations: centiles from birth to 87 years. Eur J Clin Nutr. 1991;45:13-21

(24.) Sempe M, Pedron G, Roy-Pernot MP. Auxologie, Methode et Sequences. Paris, France: Theraplix; 1979

(25.) Greulich WW, Pyle SI. Radiographic Atlas of Skeletal Development of the Hand and Wrist. 2nd ed. Stanford, CA: Stanford University Press; 1959

(26.) Rappaport R, Brauner R. Lack of adrenarche in two children with precocious puberty secondary to hypothalamic hamartoma. Horm Res. 1989;31:226-229

(27.) Judge DM, Kulin HE, Page R, Santen R, Trapukdi S. Hypothalamic hamartoma: a source of luteinizing-hormone-releasing factor in precocious puberty. N Engl J Med. 1977;296:7-10

(28.) Hochman HI, Judge DM, Reichlin S. Precocious puberty and hypo thalamic hamartoma. Pediatrics. 1981;67:236-244

(29.) Klinger G, Chin CN, Beyene J, Perlman M. Predicting the outcome of neonatal bacterial meningitis. Pediatrics. 2000;106:477-482

(30.) Wade A. Derivation versus validation. Arch Dis Child. 2000;83: 459-460

(31.) Grilli R, Magrini N, Penna A, Mura G, Liberati A. Practice guidelines developed by specialty societies: the need for a critical appraisal. Lancet. 2000;355:103-106

(32.) Rosenfield RL, Bachrach LK, Chernausek SD, et al. Current age of onset of puberty. Pediatrics. 2000;106:622-623

(33.) Finlay F, Jones R. Precocious puberty. Pediatrics. 2000;106:162-163

Martin Chalumeau, MD*; Wassim Chemaitilly, MD ([double dagger]); Christine Trivin, PhD ([section]); Luis Adan, MD ([double dagger])([parallel]); Gerard Breart, MD*; and Raja Brauner, MD ([double dagger])

From *INSERM U149, Epidemiological Research Unit on Women’s and Children’s Health, Paris, France; ([double dagger]) Pediatric Endocrinology Department, ([section]) Physiology Laboratory, Universite Rene-Descartes, Hopital Necker-Enfants Malades, Assistance Publique-Hopitaux de Paris, Paris, France; and ([parallel]) SESAB/CEDEBA, Centro de Diabetes e Endocrinologia do Estado da Bahia, Salvador-Bahia, Brazil.

Received for publication Mar 12, 2001; accepted Jun 25, 2001.

Reprint requests to (R.B.) Service d’Endocrinologie et Croissance, Hopital Necker-Enfants Malades, 149, rue de Sevres, 75743 Paris Cedex 15, France. E-mail: raja.brauner@wanadoo.fr

PEDIATRICS (ISSN 0031 4005). Copyright [c] 2002 by the American Academy of Pediatrics.

COPYRIGHT 2002 American Academy of Pediatrics

COPYRIGHT 2002 Gale Group