PEDIATRICS Vol. 108 No. 2 August 2001, pp. 347-353

Earlier Onset of Puberty in Girls: Relation to Increased Body Mass Index and Race

Paul B. Kaplowitz, MD, PhD^*, Eric J. Slora, PhD, Richard C. Wasserman, MD, MPH^{, §}, Steven E. Pedlow, BS, and Marcia E. Herman-Giddens, PA, DrPH^¶

From the ^* Department of Pediatrics, Virginia Commonwealth University School of Medicine, Richmond, Virginia;  Pediatric Research in Office Settings, American Academy of Pediatrics, Elk Grove Village, Illinois; ^§ Department of Pediatrics, University of Vermont College of Medicine, Burlington, Vermont;  Department of Statistics, University of Chicago, Chicago, Illinois; and ^¶ School of Public Health, Department of Maternal and Child Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.

	ABSTRACT

Top Abstract Methods Results Discussion Conclusion References

Objective.  A recent study conducted by the Pediatric Research in Office Settings network provided evidence that girls in the United States, especially black girls, are starting puberty at a younger age than earlier studies had found, but the reasons for this are not known. Because nutritional status is known to affect timing of puberty and there is a clear trend for increasing obesity in US children during the past 25 years, it was hypothesized that the earlier onset of puberty could be attributable to the increasing prevalence of obesity in young girls. Therefore, the objective of this study was to reexamine the Pediatric Research in Office Settings puberty data by comparing the age-normalized body mass index (BMI-ZS; a crude estimate of fatness) of girls who had breast or pubic hair development versus those who were still prepubertal, looking at the effects of age and race.

Results.  For white girls, the BMI-ZS were markedly higher in pubertal versus prepubertal 6- to 9-year-olds; for black girls, a smaller difference was seen, which was significant only for 9-year-olds. Higher BMI-ZS also were found in girls who had pubic hair but no breast development versus girls who had neither pubic hair nor breast development. A multivariate analysis confirms that obesity (as measured by BMI) is significantly associated with early puberty in white girls and is associated with early puberty in black girls as well, but to a lesser extent.

Conclusions.  The results are consistent with obesity's being an important contributing factor to the earlier onset of puberty in girls. Factors other than obesity, however, perhaps genetic and/or environmental ones, are needed to explain the higher prevalence of early puberty in black versus white girls.  Key words:  puberty, obesity, BMI, breast, pubic hair.

In 1997, Herman-Giddens et al¹ reported evidence that young girls in the United States, especially black girls, are developing pubertal characteristics earlier than currently used norms. Trained pediatrician examiners who were part of the Pediatric Research in Office Settings (PROS) network staged breast and pubic hair development in >17 000 girls between the ages of 3 and 12 years. Results showed that the mean age of onset of breast development, which had been close to 11 years in earlier studies,^2-4 is now approximately 10 years in white and just under 9 years in black girls; similar results were reported for development of pubic hair. The mean age of menarche in white girls (12.88 years) was changed little from earlier reports, whereas for black girls, it was slightly earlier (12.16 years). As a result of these findings, new guidelines for the age at which puberty in girls should be considered precocious have been proposed.⁵

A trend for earlier timing of menarche in several countries has been documented for the period from the mid-1800s to the mid-1900s and presumably is related to improved health and nutrition.⁶ It has been established that overweight girls tend to mature earlier than lean girls.⁷ These observations led to the hypothesis that the degree of body fatness may trigger the neuroendocrine events that lead to the onset of menses.⁸ Studies that compared body mass indices (BMIs) in children from national surveys between the early 1960s and the late 1980s clearly show a trend toward increasing obesity over a period of 25 years.⁹ However, there are few data available to test directly the hypothesis that increasing obesity may be a significant cause of earlier pubertal onset in the United States. For example, it is not known whether otherwise normal girls who mature earlier than average are more overweight than age-matched girls who have not started to enter puberty. Longitudinal and cross-sectional studies have found that mid- to late pubertal girls have greater body fat mass and greater BMI than prepubertal girls.^10,11 However, such studies do not resolve the question of whether the hormonal changes of puberty trigger an increase in body fat or whether a preexisting increase in body fat contributes to the onset of puberty at an earlier age.

To define better the relationship between obesity and early puberty in girls, we reexamined the PROS network data collected in the Herman-Giddens et al study by looking at BMI in relation to the pubertal status. BMI has been found to correlate well in both white and black with fat mass (r = 0.94-0.96) and percentage of body fat (r = 0.83).¹² The PROS study has the unique advantages of including height, weight, and Tanner staging for breast and pubic hair development on a large sample of healthy, young, white and black girls collected over a relatively brief time interval (July 1992-September 1993) in a geographically diverse group of pediatric offices. We tested the hypothesis that girls who were in early puberty would be more overweight than age- and race-matched girls who were prepubertal. We also speculated that the trend for black girls to mature earlier than white girls might be explained by their greater degree of obesity.

	METHODS

Top Abstract Methods Results Discussion Conclusion References

The methodology for the collection of data used in this study has been described in detail.¹ In brief, pediatricians in 65 practices in the PROS network participated in training to ensure accuracy of ratings of breast and pubic hair development according to the staging of Tanner, by visual inspection. Girls were eligible to be enrolled in the study if they were 3 to 12 years of age and came to the office for a visit that required a complete physical examination. The clinician recorded height (to the nearest cm or 0.5 in) and weight (to the nearest 0.5 kg or lb) as well as Tanner stage of breast and pubic hair development and whether the girl had reached menarche. Of the 17 077 girls whose data were analyzed for the study, 90.4% were white and 9.6% were black (because of prohibitively small numbers, other racial groups were excluded from analysis). For most of the analyses done in the original study and in this study, patients were grouped by 1-year age intervals, which extended from the day of the child's birthday to the day before the next birthday. For example, 7-year-old girls ranged from 7.000 to 7.997 years of age and had a mean age of approximately 7.5 years.

To compare the degree of fatness for different subgroups of participants, we converted each girl's BMI [weight in kilograms/(height in meters)²] to a standard BMI score for that child's age (referred to as the BMI z score [BMI-ZS]) using the reference tables of Rosner et al.¹³ These tables were developed by pooling data from 9 large epidemiologic studies published from the late 1970s and early 1990s and were based on 101 000 visits from 66 000 different children between the ages of 5 and 17 years (approximately 60% white and 30% black). Because we wanted to compare BMI-ZS in white and black children, we used the pooled mean BMI data for all girls of a given age, rather than race-specific BMI, which was slightly greater for black than white girls.

To examine the independent and combined associations between main variables of interest and the presence of breast development (Tanner stage for breast, 2 or greater), we used logistic regression modeling, which initially involved all of the variables examined in the PROS study, including age, race, BMI-ZS (calculated from height and weight), payment status (Medicaid vs other), type of visit (well-child vs other), presence of chronic disease, and use of long-term medications. Preliminary analysis showed that there was a small but significant interaction between BMI and use of steroids and other chronic disease. However, because the percentages involved in steroid use and other chronic disease were prohibitively small, the final sample chosen for the logistic regression procedure involved only children ages 5 to 12 years (because the proportion of 3- and 4-year-olds with breast development was very small), with no chronic disease or medication. Thus, from among the 11 684 study girls between the ages of 5 and 12 years, 934 girls with either chronic illness or on long-term medications were dropped, leaving a sample of 10 750 girls. For the sake of comparison, the univariate and bivariate analyses are based on this same nonchronically ill sample. There was concern that the results of the overall model would mask variables that were important for 1 racial group as opposed to the other, particularly because 90% of the study sample was white. For this reason, subsequent and separate regressions were run for white and black girls.

	RESULTS

Top Abstract Methods Results Discussion Conclusion References

As shown in Fig 1, for white girls, the mean BMI-ZS for each age are markedly greater in girls with versus without breast development, and these differences all were highly significant (P < .0001). For all age groups, the mean BMI-ZS was close to 0 for the prepubertal girls. Figure 2 presents a graphic comparison of black 6- to 9-year-old girls with and without breast development. For all but the 9-year-old prepubertal subgroup, the mean BMI-ZS were greater than 0, reflecting the heavier than average nature of the black versus white sample overall. Mean BMI-ZS for black girls also were higher in girls with Tanner 2 or greater breast development than in prepubertal girls in 3 of the 4 age groups. However, the differences were statistically significant only for the 9-year-olds. The small number of 6- and 7-year-old black girls with Tanner 2 breast development (8 and 18, respectively), as well as the above-average BMI-ZS in prepubertal girls, may account for the lack of significance of the difference in mean BMI-ZS in those 2 age groups.

View larger version (32K): [in this window] [in a new window]   Fig. 1.   Mean BMI-ZS in 6- to 9-year-old white girls with and without breast development. The number of girls for each age and Tanner stage is shown above the bars. All of the differences are significant at P < .001.

View larger version (32K): [in this window] [in a new window]   Fig. 2.   Mean BMI-ZS in 6- to 9-year-old black girls with and without breast development. The difference between stage 1 and stage 2 or greater girls is significant for the 9-year-olds (P = .03).

We next examined for each age group the relationship between the Tanner stage for breast development and the mean BMI-ZS. For this analysis, white and black girls were combined, and girls who were younger than 7 years were not included. If being overweight were a strong predictor of advanced pubertal status, then one would expect that the line relating Tanner stage and BMI-ZS for each age group would have a strongly positive slope. As shown in Fig 3, the more advanced the breast development was, the higher the mean BMI-ZS was for each age. It also is apparent that for the 3 older age groups, the mean BMI-ZS for stage 1 (prepubertal) girls was progressively less than 0. This suggests that not only are earlier-maturing girls relatively overweight, but also the later-maturing girls are relatively underweight. The same trend was seen when mean BMI-ZS in 11- to 12-year-old girls with and without menses were compared. The white girls who were still premenarchal were, on the average, somewhat underweight (mean BMI-ZS = 0.25 vs 0.29 for postmenarchal girls; P < .0001). For the black girls, the mean BMI-ZS was 0.09 for premenarchal and 0.70 for postmenarchal girls (P < .0001).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Relationship between breast development and mean BMI-ZS in girls between the ages of 7 and 12 years. For this analysis, white and black girls were combined. The numbers of girls represented by each data point are as follows: 7-year-olds: stage I = 1061, stage II = 62, stage III = 5; 8-year-olds: stage 1 = 1148, stage II = 148, stage III = 17, stage IV = 2; 9-year-olds: stage I = 679, stage II = 281, stage III = 70, stage IV = 12; 10-year-olds: stage I = 415, stage II = 392, stage III = 239, stage IV = 62, stage V = 7; 11-year-olds: stage I = 145, stage II = 307, stage III = 395, stage IV = 173, stage V = 57; 12-year-olds: stage I = 30, stage II = 113, stage III = 308, stage IV = 260, stage V = 167.

Because in young overweight girls, increased fat in the chest area might be mistaken for breast tissue, which could bias the results, we also examined the relationship between BMI-ZS and breast development in the sample of girls (approximately 40% of the total) in which breast stage was recorded by palpation as well as by visual inspection. To make sure that there was an adequate number of girls in this smaller group, we combined results for 6-, 7-, and 8-year-old girls. For the total sample of white girls with no breast tissue, mean BMI-ZS was 0.029 (N = 3393) versus 0.551 (N = 212) for the girls with Tanner 2 or greater breast development. For the smaller sample in whom Tanner staging by palpation was recorded, mean BMI-ZS in prepubertal girls was 0.073 (N = 1331) versus 0.564 (N = 107) in girls with breast tissue. Both differences were significant at P < .001. For black girls, the difference in mean BMI-ZS between girls with and without breasts was smaller, but again there was little difference between the total sample and the smaller palpated sample.

The hormonal basis of pubic hair development is distinct from that of breast development, as it reflects increased secretion of adrenal androgens (adrenarche), which, unlike ovarian estrogen secretion, is not under control of gonadotropins. Nonetheless, the trend for earlier pubic hair development in girls is very similar to the trend for earlier breast development. To determine whether increased BMI contributed to earlier adrenarche in the PROS study, we examined mean BMI-ZS in both white and black 6- to 8-year-olds with pubic hair only versus no development (3 age groups were pooled to have a large enough sample for the black girls). For comparison, we also looked at girls with breast tissue but no pubic hair as well as girls with both breast tissue and pubic hair (Fig 4). For white girls, mean BMI-ZS was 0.46 ± 1.29 for those with pubic hair alone versus 0.04 ± 1.05 for those with neither breasts nor pubic hair (P < .0001). Compared with girls with pubic hair only, mean BMI-ZS was slightly but not significantly greater for white girls with breasts only (0.53 ± 1.24) or both breasts and pubic hair (0.61 ± 1.05). For black girls, the difference between the mean BMI-ZS of 6- to 8-year-olds with pubic hair only (0.70 ± 1.58) versus girls with neither breasts nor pubic hair (0.26 ± 1.50) was nearly as great as for white girls, but because of the smaller sample size and the higher SD, the difference was not significant (P = .12). Thus, for white and possibly for black young girls, greater BMI is associated with an increased likelihood of early appearance of pubic hair.

View larger version (48K): [in this window] [in a new window]   Fig. 4.   Mean BMI-ZS in white and black 6- to 8-year-old girls with no development versus breast and/or pubic hair development. The numbers for each group are indicated above the bars. The difference in white girls between those with no development and the other 3 groups was very significant (P < .0001), but for the black girls, none of the differences was significant.

As noted previously, to look at the independent and combined associations between the main variables of interest and the presence of breast development, we used logistic regression modeling for the nonchronically ill sample described previously. Main effects models (ie, models including only the effects of the individual variables) indicated that age, BMI-ZS, and race were the most important predictors of breast development. However, when terms that took into account the interactions between variables were considered, a better model (one accounting for a higher proportion of variance in breast development) emerged. As shown in Table 1, this model included significant age (ie, with each successive year, girls were 3.256 times as likely to show signs of breast development as their year-younger counterparts) and race (white girls were only 28% as likely as their black counterparts to show such signs) effects. However, this model also showed the BMI-ZS main effect to be nonsignificant, in favor of a term that reflected the interaction of race and BMI-ZS (heavier white girls were 24.5% more likely to manifest breast development than their heavier black counterparts). The significance of the interaction term is consistent with the above results of bivariate analyses, revealing that for 6- to 9-year-old girls, the difference between BMI-ZS in pubertal versus prepubertal white girls was greater than for black girls.

To examine further the effect of BMI (and other variables) on puberty and to address concerns that the results of the overall model would mask variables that were important for 1 racial group as opposed to the other (particularly because 90% of the study sample was white), we did separate logistic regression modeling for white and black girls (Table 2). For purposes of comparison and because the variables in the race-specific final models were so similar, Table 2 displays the impact of the same 3 variablescalculated age, BMI, and type of visit (well-child vs other). For white girls, only calculated age and BMI were significant predictors. For black girls, all 3 were significant. These results suggest that BMI is a significant predictor of breast development for both white and black girls, although the parameter estimate is larger for white girls. In fact, as noted previously, the differential impact of BMI for white versus black girls is responsible for a significant portion of the variance in the overall model. This interaction of race and BMI seems to have accounted for much of the variance that BMI itself would have accounted for, resulting in the reduction of the BMI main effect to nonsignificance in the overall model. The significant interaction term in the overall model and the significance of the BMI effect in the separate analyses reported here, however, underscore the importance of BMI as an explanatory variable for breast development.

The race-specific analyses also revealed that type of visit is significant for black girls. Among black girls only, those who presented for nonwell-child visits were more developed than their well-child visit counterparts. The significance of the finding only for black girls, coupled with the large preponderance of white girls in the overall study sample, diluted the effect in the overall model, where it was nonsignificant.

	DISCUSSION

Top Abstract Methods Results Discussion Conclusion References

Several earlier studies indicated that as a group, overweight girls tend to mature earlier than girls who are not overweight.^7,8,14 The results presented here provide further support for this relationship in normal 6- to 9-year-old white and (to a lesser degree) black girls, many of whom are starting puberty at ages previously considered precocious. The original hypothesis of Frisch and McArthur⁸, that there is a critical body weight or body fatness that is required to trigger the onset of menstrual cycles, has not been confirmed by subsequent studies (eg, reference 14). However, there is strong evidence that body fat and the initiation of the hormonal events of puberty are in some way related. An important question is whether the trend for earlier puberty in US girls is likely to be related to the increasing prevalence of obesity. Alternatively, one could view the increase in body fat and BMI occurring in the peripubertal period as the result of the hormonal changes of early puberty rather than the cause. A review of the literature may help suggest which of these formulations is more likely.

Several studies documented a change in both body fat and fat distribution as girls progress through puberty,^10,15,16 although none provided evidence that the hormonal changes of puberty cause these changes. Rosenbaum and Leibel¹⁷ summarized data that suggest that estrogen accounts for the greater degree of body fatness in adult women, as well as the storage of fat in more peripheral versus central adipose tissue depots. However, because estrogens are present at low levels in early puberty, one cannot assume that it is estrogens that cause early-pubertal girls to be more obese than their prepubertal peers. Data to support the alternative explanation, that the amount of body fat is 1 of the factors that determine which girls mature early, also are lacking. A longitudinal study of 68 Dutch school-age girls concluded that body fat mass or fat distribution was not related to the age at onset of puberty.¹⁸ To what extent their findings are relevant to the trends noted in US girls is unclear, because none of their participants entered puberty before age 9.5 and no BMI data were provided.

Two recent cross-sectional studies^15,16 and 1 longitudinal study¹⁹ that looked at plasma leptin levels in healthy children and adolescents in relation to pubertal stage are relevant to this discussion. Leptin, a protein made in adipocytes, is a key regulator of body weight, and serum levels correlate very strongly with both BMI and fat mass. In the mouse, leptin also is required for normal reproductive function.²⁰ Leptin levels rise significantly in early female puberty, consistent with the hypothesis that increasing leptin levels may be 1 of the events which trigger the onset of puberty in girls. Log-transformed serum leptin levels correlated well (r = 0.73-0.85) with BMI in both early-pubertal and mid-pubertal girls.¹⁵ Leptin levels began to rise 2 to 3 years before clear pubertal increases in estradiol, luteinizing hormone, and follicle-stimulating hormone were detected.¹⁶ A longitudinal study¹⁹ confirmed the trends of these 2 cross-sectional studies and found no independent relationship among leptin, sex steroids, and gonadotropins when fat mass and fat-free mass were included in the model (as might have been expected if the rising hormone levels of early puberty caused the increase in fat mass and leptin levels). Taken together, these studies favor the hypothesis but do not prove that increasing body fat is a trigger for rather than a result of the onset of puberty.

The significant interracial difference in the onset of both breast development and pubic hair noted in the PROS study raises the question of whether the earlier maturation in black girls is attributable to genetic or environmental factors. We initially hypothesized that this difference was because of the greater obesity of young black versus white girls. The review of Troiano et al⁹ documented that the prevalence of overweight girls rose sharply in national surveys done in 1988 to 1991 (National Health and Nutrition Examination Survey III) compared with 1963 to 1965 (National Health Examination Survey). This increase was most dramatic (more than 3-fold) in 6- to 11-year-old black females. In a study of 2379 9- and 10-year-old girls, black girls were taller and heavier than white girls and more likely to have entered puberty (64% vs 33%), and those who were pubertal had a higher mean BMI.¹¹

The findings reported here document an important effect of obesity, as reflected by BMI-ZS, on pubertal maturation, but this effect was stronger for white than for black girls. In the multivariate analysis, when age and BMI were controlled for (Table 1), there is still an independent and significant association between race and breast development. Although obesity may contribute to the earlier onset of puberty in black girls, genetic and/or environmental factors specific to the black population seem to be important in determining how early puberty starts. Evidence consistent with this hypothesis was reported by Wong et al²¹ in white and black 8- to 17-year-old girls. Mean BMI and fat mass were somewhat greater in black girls, but mean leptin levels were almost twice as high; even after leptin levels were adjusted for pubertal stage and fat mass, the interracial difference in leptin persisted. Thus, genetic factors within the black population may account for their higher leptin levels for a given level of fatness, which could contribute to their observed earlier onset of puberty. In this context, it is noteworthy that above-average mean BMI-ZS in white 6- to 8-year-olds were seen only in girls who had breast development (Fig 1), whereas in black 6- to 8-year-old girls, even prepubertal girls had positive mean BMI-ZS of 0.2 to 0.5 (Fig 2). It may be speculated that the greater body fat in prepubertal black girls, as well as the higher leptin levels even after controlling for body fat, may increase their likelihood of early onset of puberty. The importance of ethnic-based genetic factors in the relationship between puberty and obesity is underscored by findings in Hispanic girls, who between the ages of 6 and 11 years have a higher mean BMI than both white and black girls¹³ but no apparent tendency toward earlier pubertal maturation.^22,23

Parallel findings were noted when we examined the influence of obesity on the development of pubic hair, which is attributable to the activation of adrenal androgen production. For white 6- to 8-year-old girls, those with pubic hair but no breast development had significantly higher mean BMI-ZS than those with no pubic hair or breast development, whereas for black girls, the difference in BMI- ZS was similar but not statistically significant because of smaller sample size. The factors that regulate the onset of adrenal androgen secretion (adrenarche) are not well understood. However, a recent study provided evidence that nutritional status is an important influence, as there was a good correlation in prepubertal boys and girls between increasing obesity during a 1-year period (measured as the change in BMI) and the increase in urinary adrenal androgen excretion.²⁴ In addition, girls with a history of premature adrenarche have an increased risk of developing functional ovarian hyperandrogenism in adolescence, a condition characterized by obesity and insulin resistance,²⁵ and insulin resistance has been demonstrated in prepubertal girls with premature adrenarche.²⁶ Arslanian et al²⁷ reported that black prepubertal children are relatively insulin resistant compared with age- and BMI-matched white children. Thus, hyperinsulinism may be a key factor in stimulating an early increase in adrenal androgen secretion and thus the early appearance of pubic hair in black girls. Our results thus provide additional evidence that both obesity and race may influence the timing of the onset of not only ovarian estrogen secretion but also adrenal androgen secretion.

The meaning of the association between nonwell-child visits and pubertal development found among black girls but not among white girls is unclear. Although black girls in the study made a slightly higher proportion of nonwell-child visits (5.38% vs 4.30% for white girls), this does not explain a finding of this magnitude. A review of the data confirms that the overwhelming majority of nonwell-child visits were for acute illnesses rather than for chronic problems. It is difficult to come up with an explanation for why the parents of black girls (but not white girls) who show evidence of early pubertal development would be more likely to bring their children to the doctor for acute illnesses.

Several limitations of this study need to be mentioned. The sample was not selected randomly, as only offices in the PROS network that volunteered for the study participated and only patients who came to the office for a visit that required a complete physical examination during the study period were enrolled. The accuracy of calculation of BMI depends on the precision of height and weight measurements, and errors in measuring height in office settings (especially undermeasurement as a result of poor positioning) are common. Breast development was staged primarily by inspection, which could have introduced a bias if a large number of overweight girls were scored as Tanner 2 when they had only fat tissue in the region of the breasts. However, 39% of the girls had breasts examined by palpation as well, and when the BMI-ZS of those girls were analyzed separately, the difference between the girls with no breast tissue and those with breast tissue persisted. Although the results are interpreted as suggesting a direct effect of body fat on increasing the likelihood of early puberty, they do not rule out the possibility that the hormonal changes of early puberty trigger the increase in BMI noted in this study. It should also be noted that body fat may be a marker for other factors that are not yet well understood, such as environmental exposures to hormonally active estrogen-like agents,²³ a lifestyle that emphasizes inactivity, and differences in diet, all of which might independently influence the onset of puberty.

	CONCLUSION

Top Abstract Methods Results Discussion Conclusion References

The results presented here support the hypothesis that the earlier onset of breast development and pubic hair documented previously¹ likely is related to increased BMI in white girls. The explanation for earlier puberty in black girls, whose average onset of breast development is 1 year earlier than white girls, is more complex; there is an interaction between race and BMI such that higher mean BMI-ZS does not account for all of the interracial difference in onset of breast development. It may be speculated that genetic factors and/or environmental factors specific to this ethnic group also may help to account for the earlier maturation of black girls.

Implications for primary care clinicians include the following. White and black girls who are overweight are more likely to show signs of pubertal development. This should be taken into account when deciding either simply to follow early-maturing 6- to 8-year-old girls closely or to refer them to search for a pathologic cause for early puberty and for treatment. Although it may be useful to arrange dietary counseling for obese, early-pubertal girls, there are no data to indicate that slowing or stopping their weight gain would slow or arrest the progression of puberty. Finally, because our study did not include sufficient Hispanic or other ethnic groups for analysis, clinicians should exercise great caution before extrapolating these conclusions to other ethnic groups.

	ACKNOWLEDGMENTS

This study was supported by Genentech, Inc, and by Grant MCJ-177022 from the Health Resources and Services Administration, Maternal and Child Health Bureau.

A list of practices that participated in the data collection for this study can be found in the original study report. We acknowledge the dedicated work of PROS practitioners in the study, as well as the efforts of the PROS chapter coordinators and PROS steering committee members in manuscript review. The views expressed in this paper are those of the authors, and no official endorsement of the American Academy of Pediatrics is intended or should be inferred.

	FOOTNOTES

Received for publication May 15, 2000; accepted Dec 28, 2000.

Address correspondence to Paul B. Kaplowitz, MD, PhD, Box 980140, Richmond, VA 23298-0140. E-mail: pkaplowitz{at}hsc.vcu.edu

	ABBREVIATIONS

PROS, Pediatric Research in Office Settings; BMI, body mass index; BMI-ZS, body mass index z score.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion References

1. Herman-Giddens ME, Slora EJ, Wasserman RC, 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 [Abstract/Free Full Text]
2. Marshall WA, Tanner JM Variations in the pattern of pubertal changes in girls. Arch Dis Child 1969; 44:291-303
3. Reynolds EL, Wines JV Individual difference in physical changes associated with adolescence in girls. Am J Dis Child 1948; 75:329-350 [Abstract/Free Full Text]
4. Nicolson AB, Hanley C Indices of physiological maturity: derivation and interrelationships. Child Dev 1953; 24:3-38 [Medline]
5. Kaplowitz PB, Oberfield SE, the Drug and Therapeutics and Executive Committees of the Lawson Wilkins Pediatric Endocrine Society Reexamination of the age limit for defining when puberty is precocious in girls in the United States: implications for evaluation and treatment. Pediatrics 1999; 104:936-941 [Abstract/Free Full Text]
6. Wyshak G, Frisch RE Evidence for a secular trend in age of menarche. N Engl J Med 1982; 306:1033-1035 [Medline]
7. Garn SM, Haskell JA Fat and growth during childhood. Science 1959; 130:1711-1712 [Abstract/Free Full Text]
8. Frisch RE, McArthur JW Menstrual cycles: fatness as a determinant of minimum weight for height necessary for their maintenance and onset. Science 1974; 185:949-951 [Abstract/Free Full Text]
9. Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, Johnson CL Overweight prevalence and trends for children and adolescents: the National Health and Nutrition Examination Surveys, 1963 to 1991. Arch Pediatr Adolesc Med 1995; 149:1085-1091 [Abstract/Free Full Text]
10. Hammer LD, Wilson DM, Litt IF, Impact of pubertal development on body fat distribution among white, Hispanic and Asian female adolescents. J Pediatr 1991; 118:975-980 [CrossRef][Medline]
11. Morrison JA, Barton B, Biro F, Sexual maturation and obesity in 9 and 10-year old black and white girls: the National Heart, Lung, and Blood Institute Growth and Health Study. J Pediatr 1994; 124:889-895 [CrossRef][Medline]
12. Daniels SR, Khoury PR, Morrison JA The utility of body mass index as a measure of body fatness in children and adolescents: differences by race and gender. Pediatrics 1997; 99:804-807 [Abstract/Free Full Text]
13. Rosner B, Prineas R, Loggie J, Daniels SR Percentiles for body mass index in U. S. children 5 to 17 years of age. J Pediatr 1998; 132:211-222 [CrossRef][Medline]
14. Garn SM, LaVelle M, Pilkington JJ Comparison of fatness in premenarcheal and postmenarcheal girls of the same age. J Pediatr 1983; 103:328-331 [CrossRef][Medline]
15. Blum WF, Englaro P, Hanitsch S, Plasma leptin levels in healthy children and adolescents: dependence on body mass index, body fat mass, gender, pubertal stage, and testosterone. J Clin Endocrinol Metab 1997; 82:2904-2910 [Abstract/Free Full Text]
16. Garcia-Mayor RV, Andrade A, Rios M, Serum leptin levels in normal children: relationship to age, gender, body mass index, pituitary-gonadal hormones, and pubertal stage. J Clin Endocrinol Metab 1997; 82:2849-2855 [Abstract/Free Full Text]
17. Rosenbaum M, Leibel RL Role of gonadal steroids in sexual dimorphisms in body composition and circulating concentrations of leptin. J Clin Endocrinol Metab 1999; 84:1784-1789 [Free Full Text]
18. De Ridder CM, Thijssen JHH, Bruning PF, Body fat mass, body fat distribution, and pubertal development: a longitudinal study of physical and hormonal sexual maturation of girls. J Clin Endocrinol Metab 1992; 75:442-446 [Abstract]
19. Ahmed ML, Ong KKL, Morrell DJ, Longitudinal study of leptin concentrations during puberty: sex differences and relationship to changes in body composition. J Clin Endocrinol Metab 1999; 84:899-905 [Abstract/Free Full Text]
20. Barash IA, Cheung CC, Weigle DS, Leptin is a metabolic signal to the reproductive system. Endocrinology 1996; 137:3144-3147 [Abstract]
21. Wong WW, Nicolson M, Stuff JE, Serum leptin concentrations in Caucasian and African-American girls. J Clin Endocrinol Metab 1998; 83:3574-3577 [Abstract/Free Full Text]
22. Villarreal SF, Martorell R, Mendoza F Sexual maturation of Mexican-American adolescents. Am J Hum Biol 1989; 1:87-95
23. Wolff MS, Berkowitz GS, Brower S, Hormonally active environmental exposures and their relationship to risk for reproductive cancer among women. Women Cancer 1998; 1:8-20
24. Remer T, Manz F Role of nutritional status in the regulation of adrenarche . J Clin Endocrinol Metab 1999; 84:3936-3944 [Abstract/Free Full Text]
25. Ibanez L, Potau N, Zampolli M, Postpubertal outcome in girls diagnosed with premature pubarche during childhood: increased frequency of functional ovarian hyperandrogenism. J Clin Endocrinol Metab 1993; 76:1599-1603 [Abstract]
26. Dimartino-Nardi J Premature adrenarche: findings in prepubertal African-American and Caribbean-Hispanic girls . Acta Paediatr Suppl 1999; 88:67-72 [Medline]
27. Arslanian S, Suprasongsin C, Janosky J Insulin secretion and sensitivity in black vs. white prepubertal healthy children. J Clin Endocrinol Metab 1997; 82:1923-1927 [Abstract/Free Full Text]