BACKGROUND. The early onset of puberty may be related to obesity, so there is a need to know the prevalence of early pubertal milestones in nonoverweight children.
OBJECTIVE. We compared attainment of stage 2 breasts, stage 3 (sexual) pubic hair, and menarche in the Third National Health and Nutrition Examination Survey sample of children with normal BMI with those with excessive BMI (≥85th percentile).
DESIGN/METHODS. The ages at which 5%, 50%, and 95% of youth had attained key pubertal stages were estimated by probit models. Logit models were then fit to compare attainment of these milestones in children of excessive and normal BMI.
RESULTS. Pubertal signs occurred before 8.0 years of age in <5% of the normal-BMI general and non-Hispanic white female population. However, pubertal milestones generally appeared earlier in normal-BMI non-Hispanic black and Mexican American girls; thelarche occurred before age 8.0 in 12% to 19% of these groups, and the 5th percentile for menarche was 0.8 years earlier for non-Hispanic black than non-Hispanic white subjects. Pubarche was found in ≤3% of 8.0-year-old girls with normal BMI of all of these ethnic groups but was significantly earlier in minority groups. Pubarche appeared before 10.0 years in <2% of normal-BMI boys. Girls with excessive BMI had a significantly higher prevalence of breast appearance from ages 8.0 through 9.6 years and pubarche from ages 8.0 through 10.2 years than those with normal BMI. Menarche was also significantly more likely to occur in preteen girls with an elevated BMI.
CONCLUSIONS. Prevalence estimates are given for the key pubertal milestones in children with normal BMI. Breast and sexual pubic hair development are premature before 8 years of age in girls with normal BMI in the general population. Adiposity and non-Hispanic black and Mexican American ethnicity are independently associated with earlier pubertal development in girls.
There has been considerable debate about whether pubertal changes between 6 and 8 years of age should be considered normal. Fifteen to 20% of 6- to 8-year-old non-Hispanic black (NHB) girls in pediatric office practice were reported to have breast and pubic hair development, in contrast to <5% of non-Hispanic white (NHW) girls; the average age of menarche was less remarkable.1 The significance of these figures has been argued because of concerns about ascertainment bias in the office practice setting.2 Although analyses of the Third National Health and Nutrition Examination Survey (NHANES III) database supported the outpatient observations,3,4 the inference drawn from these results is also suspect because of potential errors in assessment of the pubertal stage. Hypertrichosis may have been mistaken for early (stage 2) pubic hair development. Furthermore, early breast development in girls was not ascertained by palpation but rather by inspection, which blurs the distinction between breast tissue development and fat deposition (adipomastia). This particularly confounds the interpretation of these data because the US population is experiencing an increasing prevalence of obesity,5 which seems to be a factor in the early onset of puberty,6,7 pubarche,8,9 and menarche.10–13 In addition, NHB ethnicity has been found to be an independent predictor of early menarche in both the NHANES III cross-sectional database14 and the Bogalusa longitudinal database.11,12 Although BMI is an inexact surrogate of adiposity, and other factors may be involved,15 these data suggest that obesity and ethnicity are each associated with early puberty and menarche, and, consequently, there is a need for normative population-based data.16
The aim of this study was to develop estimates of the age of attainment of the key pubertal milestones in the US population of girls with normal BMI. Our premise was that early breast development (stage 2, thelarche) could be ascertained with high accuracy in girls with normal BMI and that, unlike stage 2 pubic hair, sexual pubic hair development (stage 3, pubarche) could be accurately distinguished from hypertrichosis. We hypothesized that pubertal signs would be less prevalent at 8 years of age in girls of normal BMI (10th–84th percentile) than in those with excessive BMI (≥85th percentile) if excessive adiposity were an important determinant of early pubertal development. Therefore, we determined the presence of these pubertal milestones, as well as age of menarche, in girls in the NHANES III database and compared them by BMI category. We similarly determined the prevalence of pubarche in boys for comparison.
MATERIALS AND METHODS
Data were obtained from NHANES III.17–19 NHANES III was a survey of the civilian noninstitutionalized US population conducted from 1988 to 1994 that included NHW subjects with intentional oversampling of NHB and Mexican-American (MA) subjects to increase statistical reliability in these minority groups. In 8- through 18-year-old children, pubertal signs were ascertained by inspection, whereas age of menarche was determined by questionnaire. Height and weight were measured, and BMI was computed. Centers for Disease Control and Prevention BMI percentile tables20,21 were used to convert NHANES BMI values into age- and gender-specific BMI percentile levels.
Models were fit to estimate the prevalence of attainment of the key pubertal stage variables by age in the major ethnic groups, as well as in the general population.22,23 Fieller's method was applied to the output of these models to estimate ages, with 95% confidence intervals (CIs), at which a given percentage of children had reached each landmark.24 A Wald test based on the Taylor expansion of the difference (delta method) was used to test differences in estimated ages of attainment between subgroups.25–27 Logit models were fit to compare attainment of these early pubertal milestones for children with excessive and normal BMI from ages 8 through 14 years, during which 98% of children reached these landmarks.23 Children in this age range comprised 70% to 80% of the full sample. Logit models were adjusted for race and ethnicity; all of the models included the age-by-BMI-level interaction. Logit rather than probit models were used for this purpose because they yield more interpretable results in the form of odds ratios (ORs). The analysis excluded 59 children (32 boys and 27 girls) with severe physical or mental impairment that could influence body habitus or pubertal development. All of the analyses accounted for the survey sampling design through incorporation of sampling strata and clusters, as well as weights that adjusted for differential probability of selection. SEs were computed using the Taylor linearization method.28 Statistical analyses were performed by using Stata 10.0 (Stata Corp, LP, College Station, TX).
Ethnic Comparisons in Normal-Weight Girls
The earliest stage of breast development (stage 2) was observed in 3.2% of 8.0-year-old girls with normal BMI in the general population (Table 1). Stage 2 breast development reached 5% prevalence in NHW girls at 8.6 years of age. However, significantly more NHB girls with normal BMI achieved breast development by age 8.0 years than did NHW girls, 12.1% vs 1.3%, respectively. Breast development was similarly early in MA girls with normal BMI. The median age for onset of breast development (50th percentile) was also significantly earlier in NHB than in NHW girls, by 8 months.
Pubarche, as assessed by sexual pubic hair development (stage 3), was infrequent before 8.0 years of age in girls with normal BMI of any ethnic group, in contradistinction to breast development (Table 1). The 5th percentile of attainment for the normal-BMI population as a whole was 9.3 years; it occurred significantly earlier in NHB girls (at 8.3 years) and MA girls (9.0 years) than in NHW girls (10.0 years). The median age for onset of stage 3 pubic hair in girls with normal BMI as a whole was 11.6 years, with the median age for normal-weight NHB girls being significantly earlier than for both NHW and MA girls. Although the 5th percentile for MA girls with normal BMI was significantly earlier than for NHW girls, their distribution was wider: 95% attainment of pubarche was achieved significantly later, lagging behind NHW and NHB girls by 1.0 to 1.3 years, respectively.
Menarche occurred before 9.0 years of age in <1% of girls with normal BMI of all of the ethnic groups (Table 2). The 5th percentile for achieving menarche occurred 0.8 years earlier in NHB girls and 0.6 years earlier in MA girls than in NHW girls with normal BMI (10.5 and 10.7 vs 11.3 years, respectively; P < .05). The median age of menarche was within 6 months among the 3 ethnic groups; the difference between NHW and NHB girls with normal BMI reached statistical significance.
Comparisons With Normal-Weight Boys
Pubarche occurred later in boys with normal BMI than in girls, for example, overall in 5% by age 10.7 years compared with age 9.3 in girls, with <2% of boys reaching this landmark by 10.0 years of age in any ethnic group (Table 3). The age at which 5% of boys with normal BMI attained pubarche was significantly earlier for NHB boys than NHW boys by 6 months, and this milestone was significantly earlier in both than in MA boys. The median age of pubarche was achieved at a similar age by both NHW and NHB boys, but MA boys still lagged significantly in attainment of this milestone.
Comparisons of Children With Excess Weight Versus Normal Weight
Models including adiposity showed that excess weight was associated with earlier breast development and pubarche in younger girls and with earlier menarche during the preteen years (Fig 1 and Table 1). Specifically, girls with excessive BMI were significantly more likely to exhibit stage 2 breast development at ages 8.0 (OR: 3.86) through 9.6 (OR: 2.02) years than their peers with normal BMI. Pubarche was also significantly more prevalent in girls with excessive weight from ages 8.0 (OR: 4.50) through 10.2 (OR: 1.87) years. In contrast, there was no evidence of an association between adiposity and pubarche in boys. Menarche was significantly more prevalent in heavier girls from ages 10.6 (OR: 3.93) through 12.9 (OR: 1.92) years. An estimated 5% of girls with excessive BMI attained menarche by age 10.2 years, which was 10 months earlier than when 5% of girls with normal BMI achieved this milestone, at age 11.0 years (Table 2). The median age differed by almost half this much (5.4 months), which remained statistically significant. These associations of early pubertal landmarks with adiposity were independent of race or ethnicity.
This report provides normative estimates for the ages at which thelarche, pubarche, and menarche occur in US children with a normal BMI. These results indicate that, in the general population of children of normal weight, breast and sexual pubic hair development, both ascertained by inspection, are unusual before 8 years of age in girls, and pubarche is unusual before 10 years of age in boys, occurring in <2%. However, adiposity and ethnicity are independently associated with earlier pubertal development in girls. An elevated BMI was associated with significantly greater odds of breasts appearing among younger girls, although the extent to which this is real (true breast tissue) or artifactual (adipomastia) is unknown. Excessive weight was clearly associated with significantly greater odds of earlier pubarche and menarche.
It is noteworthy that our analysis showed that an elevated BMI (≥85th percentile) was associated with significantly greater odds of attainment of thelarche and pubarche up to age 10 years. The differences in prevalence were most pronounced among the earliest developers and waned as prevalence increased so that they were no longer significant by the median age of attainment. This suggests that a subset of the population is sensitive to the onset of puberty being accelerated by an increase in adiposity that is too small to affect the majority. This may explain why previous estimates from NHANES III that did not take BMI into account differ from ours at the most by 0.2 years (for the median age of breast budding for NHB girls).3,4 The differences in prevalence of menarche between girls with elevated and normal BMI were similarly more pronounced at younger ages, with the odds of girls with elevated BMI experiencing menarche being significantly greater throughout the preteen years, when the majority of girls attained this landmark. Our analysis does not exclude the possibility of a dose-response effect of adiposity, such that frank obesity may advance the average age of attainment of all of these milestones; the NHANES III database does not have sufficient subjects with BMI >95th percentile for a robust test of this possibility. On the other hand, it is well known that a high BMI does not necessarily indicate increased adiposity,15 so that girls with high BMI who undergo puberty later may actually have a greater proportion of muscle tissue.
Our results add to the evidence that the ongoing epidemic of childhood obesity seems to be a major determinant of an earlier age of attainment of pubertal milestones in the United States over the past several decades. Our data are consistent with the evidence that overweight is approximately 1.5-fold more prevalent and obesity approximately twofold more prevalent in girls who mature at an earlier-than-average age7 and that an overweight girl is ≥2 times as likely to achieve menarche as an average-weight girl of the same age.13
Notably, ethnic differences existed independent of excess adiposity: NHB and MA girls with normal BMI achieved all of the pubertal landmarks earlier than NHW girls with normal BMI. Our analysis indicates that thelarche is normal in NHB and MA nonobese girls before 8 years of age, although pubarche is not. Although the 5th percentile for thelarche clearly occurs before 8.0 years of age in NHB and MA girls of normal weight, the lack of data on the normal population of 6- to 8-year-old girls in these ethnic groups precludes making a reliable estimate of when this milestone is achieved.
The practical implication of our analysis is that BMI should be a major consideration in the evaluation of pubertal development that occurs before 8 years of age. In part, this is because excess adiposity confounds the distinction between breast tissue and adipomastia. However, excess adiposity also advances the onset of puberty modestly in a subset of girls. Although the appearance of breasts before 8 years of age can be considered premature in NHW girls with normal BMI, occurring in <5% of them, it seems to be normal in 7-year-old NHB and MA girls with normal BMI. In contradistinction, sexual pubic hair development before 8 years (10 years in boys) is premature in children with normal BMI of all of these ethnic groups.
This research was supported in part by the Eunice Kennedy Shriver National Institute of Child Health and Human Development/National Institutes of Health through cooperative agreement U54041859 as part of the Specialized Cooperative Centers Program in Reproduction and Infertility Research (to Dr Rosenfield), USPHS grant RR-00055 (to Dr Rosenfield), and ROI-DK44752 (to Dr Lipton).
- Accepted March 26, 2008.
- Address correspondence to Robert L. Rosenfield, MD, University of Chicago Comer Children's Hospital, Section of Pediatric Endocrinology, 5841 S Maryland Ave, MC-5053, Chicago, IL 60637. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
The early onset of puberty may be related to obesity, so there is a need to know the prevalence of early pubertal milestones in nonoverweight children.
What This Study Adds
Prevalence estimates indicate that breast and sexual pubic hair development are premature before 8 years of age in the general population of girls with normal BMI and adiposity and NHB and MA ethnicity are independently associated with earlier pubertal development.
- ↵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 Pediatr Research in Office Settings network. Pediatrics.1997;99 (4):505– 512
- ↵Rosenfield RL, Bachrach LK, Chernausek SD, et al. Current age of onset of puberty. Pediatrics.2000;106 (3):622– 623
- ↵Sun SS, Schubert CM, Chumlea WC, et al. National estimates of the timing of sexual maturation and racial differences among US children. Pediatrics.2002;110 (5):911– 919
- ↵Chumlea WC, Schubert CM, Roche AF, et al. Age at menarche and racial comparisons in US girls. Pediatrics.2003;111 (1):110– 113
- ↵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 (2):347– 353
- ↵Wang Y. Is obesity associated with early sexual maturation? A comparison of the association in American boys versus girls. Pediatrics.2002;110 (5):903– 910
- ↵Guercio G, Rivarola MA, Chaler E, Maceiras M, Belgorosky A. Relationship between the growth hormone/insulin-like growth factor-I axis, insulin sensitivity, and adrenal androgens in normal prepubertal and pubertal girls. J Clin Endocrinol Metab.2003;88 (3):1389– 1393
- ↵Freedman DS, Khan LK, Serdula MK, Dietz WH, Srinivasan SR, Berenson GS. Relation of age at menarche to race, time period, and anthropometric dimensions: the Bogalusa Heart Study. Pediatrics.2002;110 (4). Available at: www.pediatrics.org/cgi/content/full/110/4/e43
- ↵Frontini MG, Srinivasan SR, Berenson GS. Longitudinal changes in risk variables underlying metabolic Syndrome X from childhood to young adulthood in female subjects with a history of early menarche: the Bogalusa Heart Study. Int J Obes Relat Metab Disord.2003;27 (11):1398– 1404
- ↵Wu T, Mendola P, Buck GM. Ethnic differences in the presence of secondary sex characteristics and menarche among US girls: the Third National Health and Nutrition Examination Survey, 1988–1994. Pediatrics.2002;110 (4):752– 757
- ↵Herman-Giddens ME, Kaplowitz PB, Wasserman R. Navigating the recent articles on girls’ puberty in Pediatrics: what do we know and where do we go from here? Pediatrics.2004;113 (4):911– 917
- ↵National Centers for Health Statistics. Third National Health and Nutrition Examination Survey (1988–94). Physician examiners training manual. NHANES III reference manuals and reports (CD-ROM). Available at: www.cdc.gov/nchs/data/nhanes/nhanes3/cdrom/nchs/manuals/phys.pdf. Accessed January 11, 2008
- National Centers for Health Statistics. Third National Health and Nutrition Examination Survey (1988–94). Analytic and reporting guidelines: NHANES III reference manuals and reports (CD-ROM). Available at: www.cdc.gov/nchs/data/nhanes/nhanes3/nh3gui.pdf. Accessed January 11, 2008
- ↵National Centers for Health Statistics. Third National Health and Nutrition Examination Survey (1988–94). NHANES III data files, documentation, and SAS code. NHANES III reference manuals and reports (CD-ROM). Hyattsville, MD: US Department of Health and Human Services, Centers for Disease Control and Prevention; 1997. Available at: www.cdc.gov/nchs/about/major/nhanes/nh3data.htm. Accessed January 11, 2008
- ↵National Centers for Health Statistics. National Health and Nutrition Examination Survey. 2000 CDC growth charts: United States. Available at: www.cdc.gov/nchs/about/major/nhanes/growthcharts/datafiles.htm. Accessed June 7, 2004
- ↵National Centers for Health Statistics. National Health and Nutrition Examination Survey. 2000 CDC growth charts: United States. BMIAGE. Available at: www.cdc.gov/nchs/data/nhanes/growthcharts/zscore/bmiagerev.xls. Accessed June 7, 2004
- ↵Finney DJ. Probit Analysis. London, United Kingdom: Cambridge University Press; 1971
- ↵McCullagh P, Nelder JA. Generalized Linear Models. London, United Kingdom: Chapman and Hall; 1989
- ↵Hubert JJ, Bohidar NR, Peace KE. Assessment of pharmacological activity. In: Peace KE, ed. Biopharmaceutical Statistics for Drug Development. New York, NY: Marcel Dekker; 1988
- ↵Gould WW. crc43: Wald test of nonlinear hypotheses after model estimation. Stata Tech Bull.1996;29 :2– 4
- ↵Philips PCB, Park JY. On the formulation of Wald tests of nonlinear restrictions. Econometica.1988;56 (5):1065– 1083
- Copyright © 2009 by the American Academy of Pediatrics