Are Pubertal Changes in Girls Before Age 8 Benign?

(1)

Are Pubertal Changes in Girls Before Age 8 Benign?

L. Kurt Midyett, MD; Wayne V. Moore, MD, PhD; and Jill D. Jacobson, MD

ABSTRACT. Objective. The Lawson Wilkins Pediat-ric Endocrine Society recently issued new recommenda-tions for the age at which puberty should be considered precocious, lowering the prevailing standards from 8 years to 7 years for white girls and to 6 years for black girls. The new recommendations were based on a single epidemiologic study that focused on the conditions of premature thelarche and premature adrenarche (both characterized by a single sign of puberty). Although the data were available, the authors did not comment on the low incidence of true precocious puberty (characterized by breast and pubic hair development) in their popula-tion. The hypothesis for the present study is that the new recommendations lead to underdiagnosis of endocrine pathology.

Methods. Using 29 International Classification of Diseases, Ninth Revisioncodes for diagnoses known to be associated with precocious puberty, we identified 1570 patient visits to our outpatient pediatric endocrinol-ogy clinic of white girls aged 7 to 8 and black girls aged 6 to 8 during a 5-year period.

Results. Of the 1570 patient visits, 223 unique pa-tients were identified as having been referred for the sole finding of precocious pubertal development. These 223 patients carried no other endocrine diagnoses. Eleven patients (4.9%) were found to have no true breast buds and no terminally differentiated pubic hair. A total of 105 (47.1%) of 223 patients were found to have 2 signs of puberty, consistent with true precocious puberty accord-ing to the conventional guidelines of precocity of 8 years in girls. Overall, 12.3% of patients also had diagnoses of other endocrine conditions that included congenital ad-renal hyperplasia, McCune-Albright syndrome, growth hormone deficiency, hypothyroidism, hyperinsulinism, pituitary adenoma, and neurofibromatosis. A total of 35.2% of girls with true precocious puberty exhibited bone ages>3 standard deviations above the mean, indi-cating markedly diminished growth potential.

Conclusions. We conclude that signs of puberty in 6-to 8-year-old girls should not be considered normal or benign. Implementation of the new guidelines for the evaluation of puberty will result in failure to identify conditions that respond to early intervention.Pediatrics

2003;111:47–51; precocious puberty, premature thelarche, premature adrenarche, body mass index, polycystic ovary syndrome, hyperinsulinism.

ABBREVIATIONS. PROS, Pediatric Research in Office Settings; BMI, body mass index; SD, standard deviation.

L

ong-held views on the definition of precocious puberty were challenged by a 1997 study pub-lished by the Pediatric Research in Office Set-tings (PROS) in which 17 077 girls, 3 to 12 years of age, were evaluated for the age of onset of secondary sexual development. The authors of that study con-cluded that girls “across the United States are devel-oping pubertal characteristics at a younger age than currently used norms.”1_{On the basis of this study,}

the Drug and Therapeutics Committee of the Lawson Wilkins’ Pediatric Endocrine Society issued a state-ment lowering the normal age of puberty to 7 years in white girls and 6 years in black girls.2 _{The joint}

committee stated, “In most cases, evaluation of girls with early breast and/or pubic hair development to look for a pathologic etiology of precocious puberty need not be performed for white girls older than 7 years and African-American girls older than 6 year of age.”2_{This recommendation seemed to us somewhat}

premature, considering the acknowledged limita-tions of the study and lack of data on endocrine evaluations performed on girls with early pubertal changes.1 _{In fact, the revision has led to extensive}

controversy with a number of pediatric endocrinol-ogists issuing statements contesting the new guide-lines. We share the concern expressed by other en-docrinologists that liberalizing the definition of “normal” carries with it the risk of overlooking pa-thology.3,4_{In fact, the PROS study included no data}

as to what percentages of patients younger than 8 years were subsequently referred for complete endo-crinologic evaluations and what pathology was found. According to medical standards of care at the time, 22 of their “normal” patients should have been referred for diagnostic evaluations.5

The hypothesis of the current study is that the new guidelines, if followed, will lead to significant under-diagnosis of endocrine conditions. Therefore, it was our primary objective to ascertain the extent of en-docrine pathology in girls who are referred to our center for the ages in question: ages 6 to 8 in black girls and 7 to 8 in white girls. A second objective was to identify girls with true precocious puberty (char-acterized by 2 signs of puberty) in our referral pop-ulation and to distinguish this entity from the con-ditions of isolated premature thelarche and isolated premature adrenarche.

A recent paper by Kaplowitz et al6_{suggested that}

obesity was linked to the earlier timing of puberty From the Section of Endocrinology, Children’s Mercy Hospital, Department

of Pediatrics, University of Missouri-Kansas City School of Medicine, Kan-sas City, Missouri.

Received for publication Jan 2, 2002; accepted May 24, 2002.

Reprint requests to (J.D.J.) Section of Endocrinology, Children’s Mercy Hospital, 2401 Gillham Rd, Kansas City, MO 64108. E-mail: jjacobson@ cmh.edu

(2)

and that patients with obesity were less likely to display pathologic causes for their precocious pu-berty. Thus, a third objective was to determine whether obese patients were less likely to display additional endocrine pathology.

METHODS Study Population

This study was reviewed by the Institutional Review Board at Children’s Mercy Hospital and received an exempt status. A ret-rospective chart review was undertaken. Using 29 International Classification of Diseases, Ninth Revisioncodes for diagnoses known to be associated with precocious puberty, we identified 1570 in-dividual outpatient visits to our pediatric endocrinology clinic by white girls aged 7 to 8 and by black girls aged 6 to 8 from 1996 to 2001. From these visits, 663 unique patients were identified in this age group. A review of these charts identified 223 patients whose only reason for referral was precocious pubertal changes.

These 223 girls were seen in our pediatric endocrine clinic, which is based in a large metropolitan children’s hospital, by any of 6 pediatric endocrinologists. Pediatricians or family practitio-ners referred most patients, but some were self-referred. All of the patients were new to our clinic and had no previously diagnosed endocrine conditions. The diagnosis of true precocious puberty was defined as the presence of both breast and pubic hair devel-opment before the age of 8 with a concomitant growth spurt. Tanner staging was assessed on all patients by both visual inspec-tion and palpainspec-tion. All radiographs for bone age assessment were read by pediatric radiologists at our institution. We requested films on all patients with abnormal bone ages (⬎2 standard devi-ations [SD]) to assess accuracy of the bone age reports. Fifty-three of the 83 abnormal bone ages were available and reviewed by a single pediatric endocrinologist: 43 of 53 bone age readings were concordant between the radiologists’ and the endocrinologist’s reviews. Of the 11 that were discordant, 8 readings differed by 6 months or less. No bone age reading was discrepant by more than 1 year. Girls were classified as obese when the body mass index (BMI) was greater than the 95th percentile determined from the Centers for Disease Control and Prevention growth charts for girls in the United States as revised in May 2000.

Statistical Analysis

Fisher exact test was used to compare the differences between girls with 2 signs of puberty versus 1 and to compare the differ-ences in pathology between lean girls and obese girls. Fisher exact test also was used to analyze the PROS data for differences be-tween white and black girls with 2 signs of puberty.Zscores for height and weight were derived using computer software based on the revised Centers for Disease Control and Prevention growth charts for girls in the United States. Studentttest was used for the comparisons of height and weight, BMI, ⌬ bone ages, and ⌬ predicted adult height between patients with 2 signs of puberty and those with 1 sign of puberty.

RESULTS

We compared differences between groups with 1 sign of early sexual development (breast or pubic hair alone) or 2 signs of early sexual development (breast and pubic hair). Of the total 223 patients referred to us, 105 (47%) had evidence of both breast and pubic hair development, 83 (37%) had evidence of pubic hair development alone, and 24 (11%) had only breast development. Eleven girls (5%) were re-ferred for signs of puberty but were found to have no true signs of early sexual development.

Clinical findings seen in our population included increased BMI, advanced bone ages, and diminished growth potential. Of the 212 patients found to have signs of puberty, 95 (45%) were obese, with a greater percentage of girls with 2 signs of pubertal develop-ment being overweight compared with those with 1

sign of puberty (52 [50%] of 105 vs 36 [34%] of 107; P ⫽ .024; Table 1). The majority of girls received a diagnosis of idiopathic pubertal development. However, 1 girl (4%) with early breast development, 13 girls (16%) with pubic hair development, and 12 girls (11%) with both breast and pubic hair develop-ment were found to have additional endocrine diag-noses. These diagnoses included neurofibromatosis, McCune-Albright syndrome, growth hormone defi-ciency, hypothyroidism, congenital adrenal hyper-plasia, pituitary microadenoma, and hyperinsulin-ism (Table 2).

Patients with 2 signs of pubertal development ex-hibited the following: 1) significantly greater weight zscores (1.65 ⫾ 0.09) compared with patients with either premature thelarche (1.00 ⫾ 0.23) or prema-ture adrenarche (1.22 ⫾ 0.14; P ⫽ .006 and 0.008, respectively; Fig 1), 2) significantly greater heightz scores (1.27 ⫾ 0.11) compared with patients with either premature thelarche (0.54 ⫾ 0.22) or prema-ture adrenarche (0.80 ⫾ 0.12; P ⫽ .006 and 0.007, respectively; Fig 2), 3) significantly higher BMI (20.67⫾ 0.35) compared with patients with prema-ture thelarche (18.77⫾0.66;P⫽ .02; Fig 3), 4) bone ages that were significantly more advanced (1.99⫾ 0.14 years) compared with those of patients with either premature thelarche (1.19 ⫾ 0.32 years) or premature adrenarche (0.80 ⫾ 0.18 years; P ⫽ .02 and⬍.001, respectively; Fig 4), and 5) a greater loss of final predicted adult heightzscores (⫺1.03⫾0.31) compared with girls with premature adrenarche alone (⫺0.77⫾ 0.10;P⫽ .000001; Fig 5)

Bone ages were advanced up to 5 years above chronological age in girls with 2 signs of puberty. Of girls with 2 signs of puberty, 37 (35%) of 105 girls exhibited bone ages ⬎3 SD above the mean. Eight patients with 2 signs of puberty and 1 patient with premature thelarche exhibited predicted height be-low 150 cm (4 ft 11 in). No girl with premature adrenarche exhibited a predicted height⬍150 cm.

With regard to endocrine pathology in obese ver-sus nonobese patients, the results revealed that al-though 45% of patients overall were obese, 81% of girls who had a diagnosis of additional endocrine

TABLE 1. Clinical Findings in 212 Girls, Between the Ages of 6 and 8, Who Were Referred to Pediatric Endocrine Subspecialists for Signs of Early Puberty

Breast Development

(n⫽24)

Pubic Hair Development

(n⫽83)

Breast and Pubic Hair (n⫽105)

Age of presentation (mean⫾SD)

7.44⫾0.61 7.45⫾0.49 7.42⫾0.55

Ethnicity

White 18 (75%) 51 (61.45%) 56 (53.34%)

Black 6 (25%) 31 (37.35%) 47 (44.76%)

Hispanic 0 1 (1.20%) 2 (1.90%)

Obesity 6 (25%) 30 (36%) 52 (49.5%)

Bone age⬎2 SD and⬍3 SD

3 (12.5%) 7 (8.4%) 25 (23.8%)

Bone age⬎3 SD 3 (12.5%) 8 (9.6%) 37 (35.2%) Mean loss of height

(PAH⫺current height) (zscore)

⫺0.90 (17) ⫺0.77 (67) ⫺1.55 (94)

(3)

conditions were obese. Twenty-one of 95 patients who were obese exhibited additional endocrine pa-thology, whereas only 5 of 117 patients who were not obese displayed additional pathology (P ⫽ .0001; odds ratio: 6.3).

We compared black and white girls with the same pubertal changes for differences in BMI, weight z scores, heightzscores,⌬bone age, and potential loss in predicted adult height. Black girls in the 6- to 7-year range were excluded, as there were no white girls in that age group. The only difference between the 7- to 8-year-old black and white girls was in the premature adrenarche group when comparing po-tential loss of adult height. Black girls with prema-ture adrenarche displayed significantly lower ⌬ heightzscores between current height and predicted

height based on bone age (zscore:⫺1.06 vs⫺0.62). The significance of this difference became marginal when adjusted for multiple comparisons.

DISCUSSION

Our data indicate that pathologic endocrine con-ditions were present in a subset of patients who were referred to us for precocious pubertal changes occur-ring at an age defined as normal by the PROS study. Overall, 12.3% of the patients who were referred to our endocrine clinic between the ages of 6 and 8 manifested pathologic explanations for their sexual precocity. These endocrine diagnoses included insu-lin resistance/hyperinsuinsu-linism, congenital adrenal hyperplasia, neurofibromatosis, growth hormone de-ficiency, hypothyroidism, and McCune-Albright

TABLE 2. Diagnoses in 212 Girls, Between the Ages of 6 and 8, Who Were Referred to Pediatric Endocrine Subspecialists for Signs of Early Puberty

Breast Development

(n⫽24)

Pubic Hair Development

(n⫽83)

Breast and Pubic Hair (n⫽105)

Idiopathic 23 (95.80%) 70 (84.40%) 93 (88.55%)

Acanthosis

Nigricans/hyperinsulinism

1 (4.20%) 7 (8.40%) 7 (6.70%)

Hypothyroidism 0 3 (3.60%) 1 (0.95%)

Neurofibromatosis type 1 0 0 1 (0.95%)

Growth hormone deficiency 0 1 (1.20%) 1 (0.95%)

Pituitary adenoma 0 0 1 (0.95%)

McCune-Albright syndrome 0 1 (1.20%) 1 (0.95%)

Congenital adrenal hyperplasia

0 1 (1.20%) 0

Fig 1. Weightzscores for girls with premature thelarche, prema-ture adrenarche, and true precocious puberty. *Girls with 2 signs of puberty exhibited significantly higher weightzscores than girls with breast development alone (P⫽.006) or pubic hair alone (P⫽

.008).

Fig 2. Heightzscores for girls with premature thelarche, prema-ture adrenarche, and true precocious puberty. *Girls with 2 signs of puberty exhibited significantly higher heightzscores than girls with breast development alone (P⫽.006) or pubic hair alone (P⫽

.007).

Fig 3. BMI in girls with premature thelarche, premature adren-arche, and true precocious puberty. *Girls with 2 signs of puberty exhibited significantly higher BMI than girls with breast develop-ment alone (P⫽.02).

Fig 4. ⌬bone age (bone age⫺chronological age) in girls with premature thelarche, premature adrenarche, and true precocious puberty. *Girls with both signs of puberty exhibited significantly higher⌬bone ages than girls with breast development alone (P⫽

(4)

syndrome. It is important to note that the physical findings can be subtle in patients with nonclassical congenital adrenal hyperplasia, subclinical hypothy-roidism, forme frustes of McCune-Albright syn-drome, or neurofibromatosis. Thus, it is not surpris-ing that these conditions were not previously diagnosed by the referring physicians. That preco-cious puberty may be the first and/or only recog-nized manifestation raises the concern that the diag-nosis of endocrine disorders might be overlooked if the new guidelines are followed.

Perhaps of greater concern than the pathologic diagnoses was the additional finding of advanced bone ages, indicating diminished growth potential in a larger subset of girls. More than one third of girls with 2 signs of puberty exhibited bone ages⬎3 SD above the mean. A subset of girls who presented with breast development alone also exhibited ad-vanced bone ages. In fact, girls with breast develop-ment alone were just as likely to display potential losses in adult height as girls with 2 signs of puberty. This is not surprising, as estrogen is a major factor in progression of skeletal maturation. The finding of loss of adult height potential in girls with isolated premature thelarche underscores the need for indi-vidual evaluation by pediatric subspecialists to de-termine who is at risk for compromise of adult height. If these girls had not been evaluated between the ages and 6 and 8, then it is doubtful that they would have been evaluated later, unless premature menarche occurred, at which time growth potential would have been limited.

Our study was based on a referral population and so does not address prevalence of pathology in the general population. One might argue that our refer-ral population has changed since the new guidelines have been in place, with pediatricians referring fewer and more complicated patients. Our analysis does not support this contention. We found that the num-ber of girls who were referred to our clinic for eval-uation of early pubertal development did not de-crease after publication of the new guidelines.

Pediatric endocrinologists have long made the im-portant distinction between the potentially benign conditions of isolated premature thelarche (charac-terized by breast development alone), premature

adrenarche (characterized by pubic hair develop-ment alone), and true precocious puberty (character-ized by breast and pubic hair development), which carries with it a higher prevalence of significant pathologic conditions to explain the precocious de-velopment.7_{Although it is important to distinguish}

true precocious puberty from the partial forms of puberty, our data indicate that almost 16% of girls who presented with only early pubic hair develop-ment were found to have additional endocrine diag-noses. Importantly, hyperinsulinism was the most frequent additional endocrine diagnosis in girls who presented with early pubic hair or both. Recent stud-ies have emphasized that premature adrenarche may be the first recognizable sign of the symptom com-plex identified as polycystic ovary syndrome or the metabolic syndrome (hyperinsulinism, obesity, im-paired glucose tolerance, dyslipidemia, cardiovascu-lar disease).8 –11_{In fact, the number of girls who are}

at risk for this problem is likely higher in our popu-lation than the data suggest, as insulin levels and oral glucose tolerance testing were not routinely done on girls with signs of premature adrenarche in the initial years of our study period. Early recognition and dietary intervention may be beneficial in this sub-group of patients, again underscoring the need for early identification and referral to subspecialists.

An editorial by Herman-Giddens et al12 _stated,

“We question the recommendation that all girls with the sole factor of breast development or pubic hair before 8 and 9 years of age, respectively, have a diagnostic evaluation.” In practice, it has not been the standard of care for pediatricians or pediatric endocrinologists to perform extensive evaluations in girls with 1 sign of puberty before the age of 8 or 9. Before the revision in guidelines, standard prac-tice would be to take an appropriate history, perform a thorough physical examination, observe linear growth, and possibly perform simple bone age as-sessment in girls with single signs of puberty. More extensive studies would be reserved for girls with 2 signs of puberty. Our findings suggest that all girls with any secondary sexual development before 8 years of age deserve at the minimum a bone age assessment and close longitudinal follow-up, espe-cially when one considers that breast and pubic hair development can be asynchronous in the early stages of precocious puberty.

A close inspection of the graphs in the original study from the PROS group reveals that only a small percentage of patients exhibited findings of true pre-cocious puberty. It seems that only 2 black girls (1.6%) and 6 white girls (0.4%) aged 6 years to 7 years displayed 2 signs of puberty. Eight black girls (5.9%) and 12 white girls (1.1%) aged 7 to 8 years displayed 2 signs of puberty.1_{These calculations are important}

for 2 reasons. First, assuming a normal distribution of onset of puberty, it calls into question whether a major epidemiologic change in the timing of puberty has occurred at all. This low incidence of true preco-cious puberty is in line with the PROS finding that the age of menarche has not changed more than a few months since the early 1960s.1,13 _{Second, our}

analysis of the PROS data published reveals that

(5)

there is no statistically significant difference in the incidence of true precocious puberty between white and black girls in the 6- to 7-year age group.1_Thus,

the PROS data do not provide a scientific basis for using separate guidelines for black and white girls between the ages of 6 and 7 years

Additional analysis of the data in the original study from the PROS group reveals that of the total of 17 077 girls—55 black girls between the ages of 6 and 8 and 75 white girls between the ages of 7 and 8 — exhibited 1 sign of puberty.1 _{Accordingly, only}

130 girls (0.8%) would have required additional eval-uation under the prevailing standards of care at the time. The additional evaluation might have been as simple as a bone age assessment or fasting serum insulin and glucose levels. Only 22 girls out of the entire population of 17 077 girls (0.1%) of the ages and races in question would have met the definition of true precocious puberty under the definitions of precocity at the time. Subspecialty referral of 100% of girls with 2 signs of puberty would hardly over-whelm the current health care system, even with the current shortage of pediatric endocrinologists.

A recent study by Kaplowitz et al6_{suggested that}

obesity should be considered a causative factor in early puberty and that weight should be taken into account as a mitigating factor when deciding whether to refer patients to search for pathologic causes. Our data do not support this recommenda-tion. Whereas 45% of patients overall were obese, 81% of the girls with additional endocrine diagnoses were obese. This finding suggests that equal atten-tion should be given to obese girls and nonobese girls who present with signs of early pubertal devel-opment.

The extensive list and complexity of the endocrine conditions associated with precocious sexual devel-opment underscores the need for complete evalua-tion and individualized attenevalua-tion by subspecialists. On the basis of the results of our study, we will continue to recommend referral of any girl who is younger than 8 years and has 2 signs of puberty. We will also continue to recommend (at a minimum) a bone age evaluation on all girls who are younger than 8 years and have even 1 sign of sexual devel-opment. The subsequent evaluation and/or referral should be tailored to identify endocrine conditions expected in each subtype of early sexual develop-ment.

Additional studies in other endocrinology prac-tices will help to determine whether a similar degree of pathology exists in other referral centers. Our study underscores the hazard in revising population guidelines based on a single epidemiologic study. Strict adherence to the new guidelines and exclusion from endocrine evaluations of girls who are aged 6 to 8 and have secondary sexual development will lead to underdiagnosis and undertreatment in this age group.

ACKNOWLEDGMENT

We thank Steve Simon, PhD, for expertise and assistance in the statistical analysis for this article.

REFERENCES

1. 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

2. 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 Com-mittees of the Lawson Wilkins Pediatric Endocrine Society.Pediatrics. 1999;104:936 –941

3. Pathomvanich A, Merke DP, Chrousos GP. Early puberty: a cautionary tale.Pediatrics. 2000;105:115–116

4. Rosenfield RL, Bachrach LK, Chernausek SD, et al. Current age of onset of puberty.Pediatrics. 2000;106:622– 623

5. Rogol A, Blizzard RM.Wilkins’ The Diagnosis and Treatment of Endocrine Disorders in Childhood and Adolescence. Springfield, IL: Charles C Thomas; 1994

6. Kaplowitz PB, Slora EJ, Wasserman RC, Pedlow SE, Herman-Giddens ME. Earlier onset of puberty in girls: relation to increased body mass index and race.Pediatrics. 2001;108:347–353

7. Lee PA.Principles and Practice of Endocrinology and Metabolism. Philadel-phia, PA: JB Lippincott; 1990

8. Dimartino-Nardi J. Premature adrenarche: findings in prepubertal Af-rican-American and Caribbean-Hispanic girls.Acta Paediatr Suppl. 1999; 88:67–72

9. Ibanez L, Potau N, Dunger D, de Zegher F. Precocious pubarche in girls and the development of androgen excess.J Pediatr Endocrinol Metab. 2000;13(suppl 5):1261–1263

10. Ibanez L, Dimartino-Nardi J, Potau N, Saenger P. Premature adrenarche: normal variant or forerunner of adult disease?Endocr Rev. 2000;21:671– 696

11. Silfen ME, Manibo AM, McMahon DJ, Levine LS, Murphy AR, Ober-field SE. Comparison of simple measures of insulin sensitivity in young girls with premature adrenarche: the fasting glucose to insulin ratio may be a simple and useful measure.J Clin Endocrinol Metab. 2001;86: 2863–2868

12. Herman-Giddens ME, Bourdony C, Slora E, Wasserman R. Early puberty: a cautionary tale.Pediatrics. 2001;107:609 – 610

(6)

DOI: 10.1542/peds.111.1.47

2003;111;47

Pediatrics

L. Kurt Midyett, Wayne V. Moore and Jill D. Jacobson