Objective. For those with normal body habitus, self-assessments have been reported to yield Tanner stage ratings similar to those found by actual examinations. Little is known about whether such self-assessments are accurate in obese children. We therefore determined the reliability of Tanner stage self-assessments in both nonobese and obese children.
Methods. We studied 244 children age 6 to 12 years, 135 girls and 109 boys, 41% of whom were obese (body mass index ≥95th percentile for age and gender). Girls rated both breast and pubic hair development and boys rated pubic hair development using a standardized series of drawings accompanied by explanatory text. After self-ratings were completed, a pediatric endocrinologist or trained nurse practitioner who was blinded to subjects’ self-ratings examined each subject.
Results. Self-ratings of breast Tanner stage were concordant with actual stage in 48%, overestimated in 25%, and underestimated in 27% of nonobese girls. By contrast, breast Tanner stage was overestimated by 38% of obese girls and was underestimated by only 12%. On average, obese girls overestimated actual Tanner breast stage by 0.47 ± 0.9 stages. Pubic hair ratings were largely concordant with actual pubic hair stage in both obese and nonobese girls. Both nonobese and obese boys significantly overestimated actual Tanner pubic hair stage, by 0.51 ± 1.1 stages and 0.31 ± 0.8 stages, respectively.
Conclusions. We conclude that, in children 6 to 12 years of age, self-assessment cannot be used to determine reliably the breast Tanner stage of obese girls or the pubic hair stage of boys.
Accurate assessment of sexual maturity is an important consideration for many pediatric investigations. Many studies that do not involve actual physical examinations ask children and adolescents to rate their own pubertal development.1–3 Previous small studies examining the accuracy of such self-reported measures have suggested that breast and pubic hair Tanner stage self-assessments in girls have acceptable agreement with physician ratings in children 9 to 17 years.4–7 For example, Duke et al4 studied 43 girls and reported a correlation between subjects and physicians of 0.81 for breast stage and 0.91 for pubic hair. Others have found poor concordance between self-assessments and examinations of testicular development and variable concordance with pubic hair stage in boys.5–8
Between 1976 and 1994, the prevalence of overweight in childhood increased dramatically, and national surveys suggest that >12% of all children and adolescents have a body mass index (BMI) greater than the 95th percentile that was defined by the second and third National Health and Nutrition Examination Surveys.9 Although high BMI seems to be associated with early pubertal development,10 little is known about the effects of fatness on self-assessments of Tanner stage. One previous small study attempted to examine the effect of fatness on self-assessment by assessing concordance of self-ratings to examination ratings in children who had high skinfold thickness, a measure of subcutaneous adiposity.11 Williams et al11 examined 37 children for whom the sum of triceps and subscapular skinfold thickness was greater than the 70th percentile for age and gender and did not demonstrate any statistically significant difference between subjects’ and physicians’ ratings of breast or pubic hair Tanner stage. Because the standard definition of obesity requires skinfold thickness or BMI greater than the 95th percentile, it is unclear how many subjects in the Williams et al study were actually significantly overweight. To the best of our knowledge, no previous studies have examined self-reports of pubertal maturation in significantly overweight children. We, therefore, compared the accuracy of Tanner stage self-assessments in a large cohort of obese and normal-weight children. We hypothesized that, because of lipomastia, Tanner breast stage self-ratings would be significantly overestimated in obese but not normal-weight girls. We further hypothesized that pubic hair stage self-ratings would be assessed accurately in both normal-weight and obese boys and girls.
Subjects were recruited to take part in 2 natural history research studies at the National Institute of Child Health and Human Development. Recruitment was through mailed notices to 6- to 12-year-old children in the Montgomery and Prince Georges County, Maryland, school districts; through local physician referrals of healthy overweight children; and through advertisements in local newspapers (Table 1). Subjects were classified on the basis of BMI percentile, determined from the age-, gender-, and race-specific norms for BMI of the first National Health and Nutrition Examination Survey,12 as being of normal weight when BMI was between the 15th and 85th percentiles, overweight when BMI was ≥85th but <95th percentile, and obese when BMI was ≥95th percentile for their age, gender, and race. All subjects had normal history and physical examinations as well as normal blood chemistry and hepatic and thyroid function. None of the subjects had any significant physical or neurocognitive illness. The National Institute of Child Health and Human Development institutional review board approved the clinical protocol. Informed consent and assent were obtained from parents and their children.
Subjects were studied during an outpatient visit at the Warren Grant Magnuson Clinical Center of the National Institutes of Health. Before they underwent physical examination, subjects were given a standardized series of drawings with explanatory text to assess their own pubertal development. Girls were given line drawings of the 5 stages of breast and female pubic hair development with appropriate written descriptions accompanying the drawings. Boys were given line drawings of boys showing the 5 stages of pubic hair development, with appropriate written descriptions. The description of each stage was read to the subject, and then each subject was asked to select the drawing and stage that best indicated his or her own development. Subjects then underwent a physical examination by either a trained nurse practitioner or a pediatric endocrinologist, both experienced in Tanner stage evaluation. The examiner was blinded to the subjects’ self-assessments. Interrater reliability was 100% for Tanner breast stage and >98% for pubic hair stage. Substantial training was done to achieve this high level of interrater reliability.
The data were analyzed by nonparametric procedures using StatView 5.01 software (Abacus Concepts, Inc, Berkley, CA). The correlations between actual and self-reported Tanner stage ratings were examined using Kendall rank correlations, and the differences between actual and self-reported Tanner stage ratings were examined with the 1-sample sign test. Tanner stage data were analyzed separately for girls and boys. Additional analyses examined whether Tanner stage was estimated with equal accuracy in obese and nonobese subjects.
Description of Study Population
A total of 244 subjects were studied, 135 of whom were girls and 109 of whom were boys. The age range of the girls was 5.9 to 12.9 years, and the age range of the boys was 5.5 to 12.9 years. Fifty-seven percent were white; 36% were black; and the remainder were of Hispanic, Asian, or mixed ethnic background. When categorized by their BMI, 60 of the girls (44%) and 39 of the boys (36%) were obese, with a BMI ≥95th percentile for age, gender, and race; 22 girls (16%) and 13 boys (12%) had BMI between the 85th and 95th percentiles; and the remainder (110 children) had BMI between the 15th and 85th percentiles. For the purposes of analyses, all subjects with BMI <95th percentile were combined as a nonobese group. In addition, because there were no significant differences when data were analyzed by race, all data were analyzed including all races together. Mean BMI was similar for the nonobese girls and boys (17.6 ± 2.7 vs 17.3 ± 1.8 kg/m2; P=.36) but was significantly lower for obese girls than for obese boys (28.0 ± 5.6 vs 31.4 ± 8.0 kg/m2; P = .015). A large portion of the cohort was severely obese, with 17% of the children (19 boys and 23 girls) having BMI >30 kg/m2(Table 1).
Self-Assessment of Breast Development in Girls
Of the entire group of 135 girls, 49% accurately estimated their breast Tanner stage; 43 (32%) of 135 overestimated their Tanner stage, and 27 (20%) of 135 underestimated their Tanner stage. Of the obese girls (BMI ≥95%), 30 (50%) of 60 accurately assessed their breast Tanner stage, whereas 23 (38%) overestimated and 7 (12%) underestimated their breast Tanner stage. Of the nonobese girls, 36 (48%) of 75 accurately assessed breast Tanner stage, 19 (25%) overestimated breast Tanner stage, and 20 (27%) underestimated breast Tanner stage. When examined separately, obese girls significantly overestimated their actual breast stage by 0.47 ± 0.9 stages (1-sample sign test, P=.0052). By contrast, there was no significant bias of estimation in nonobese girls (P = .99). Kendall rank correlation coefficient between breast Tanner stage self-reports and actual ratings for obese girls was 0.37 (P < .0001) and for nonobese girls was 0.535 (P < .0001; Fig 1A, Table 2).
Self-Assessment of Pubic Hair Development in Girls
For girls, self-reported pubic hair Tanner stage did not show significant bias relative to actual Tanner stage. Pubic hair Tanner stage was accurately estimated by 80 (59%) of 135 girls, overestimated by 31 (23%) of 135, and underestimated by 24 (18%). There were no weight-related differences in reports of girls’ pubic hair stage. Of the 60 obese girls, 34 (57%) accurately assessed, 14 (23%) overestimated, and 12 (20%) underestimated their pubic hair Tanner stage. Of the 75 nonobese girls, 46 (61%) accurately assessed, 17 (23%) overestimated, and 12 (16%) underestimated pubic hair Tanner stage. There was no significant bias of estimation in either obese (P=.85) or nonobese (P=.45) girls. Kendall rank correlations between actual and self-reported Tanner pubic hair stage showed correlation coefficients of 0.66 (P < .0001) and 0.64 (P < .0001) for the nonobese and obese girls, respectively (Fig 1B, Table 3).
Self-Assessment of Pubic Hair Development in Boys
Self-assessment of pubic hair Tanner stage was accurately assessed in 52 (48%) of 109 boys, underestimated in 12 (11%), and overestimated in 45 (41%). In both obese and nonobese boys, there was a significant bias toward overestimation of pubic hair Tanner stage. Sixteen (41%) of 39 obese boys and 29 (41%) of 70 nonobese boys overestimated their pubic hair Tanner stage. Both nonobese and obese boys significantly overestimated actual Tanner pubic hair stage, by 0.51 ± 1.1 stages (P=.0001) and 0.31 ± 0.8 stages (P=.05), respectively. Kendall rank correlation showed significant correlations between actual and self-reported pubic hair stage in boys (0.35 for nonobese and 0.45 for obese boys, both P < .0001; Fig 1C, Table 3).
Relationship Between Age and Tanner Stage Estimates
We found no significant association between age of boys or girls and the accuracy of their assessments of breast or pubic hair Tanner stage. The majority of overestimations were found in those with Tanner stage 1 or 2 development, but overestimations were not limited to the youngest children.
Previous studies have reported that children of normal weight can accurately assess their own stage of sexual maturation using drawings and pictures,4–8 but extremely limited data exist for overweight children. The one previous study11 that attempted to examine the effects of fatness on accuracy of self-assessment of Tanner stage likely evaluated few subjects with significant obesity and stated that no “morbidly obese” children were studied.11 In the present investigation, we found that both obese and nonobese girls assessed pubic hair Tanner stage without significant bias, but when obese and nonobese groups were analyzed separately, only nonobese girls estimated their Tanner breast stage without significant bias. By contrast, obese girls significantly overestimated their Tanner breast stage.
Although our study population consisted of children who were younger than those in some previous investigations of Tanner stage self-assessments, we do not believe that the children’s age significantly contributed to the overestimation of breast Tanner stage. We found no correlation between age and the ability to correctly assign either breast or pubic hair Tanner stage. Furthermore, that girls were able to assign their pubic hair stage with good accuracy suggests that a lack of the cognitive ability to understand the rating diagrams did not contribute to these findings.
We propose that the tendency for overestimation of breast development in obese girls may be the result of their difficulty to distinguish lipomastia from true breast tissue. The widely used method of pubertal stage assessment described by Tanner was originally based on visual inspection.13 However, to assess breast Tanner stage accurately in overweight girls, it may be necessary to palpate the breasts to determine the amount of actual breast tissue present. This technique is used in boys to distinguish gynecomastia from lipomastia.14 If the breasts are not palpated, then adipose tissue can be mistaken for breast tissue, causing an overestimation in breast Tanner stage. Because the self-assessment is done by visual comparison of the subject’s breasts to a picture representation, the subject’s assessment may overestimate Tanner stages 1 to 4 when lipomastia is present.
The difficulties in assessing breast Tanner stage in obese girls with lipomastia may also potentially lead to incorrect assignment of Tanner stage during routine physical examinations. If a health care professional is determining Tanner breast stage by visual inspection alone, without palpation, then the examiner may also overestimate Tanner breast stage in overweight girls. In 1997, a study that reported assessments of pubertal maturation by clinicians in pediatric practices15 concluded that the mean age of onset of breast development was earlier than the current norms for the start of breast development. This study reported that 15.4% of black girls and 5.0% of white girls had breast development by age 7 years. However, the investigators did not find a concordant advancement in the age of the onset of menses.15 In this study, the evaluation of breast Tanner stage was done by visual inspection of the breast alone, and the study approach did not specify a need for palpation. Given that >12% of all girls are now classified as having BMI ≥95th percentile9 and that the prevalence of overweight in black girls may be even greater,16 we propose that the earlier onset of breast development identified by Herman-Giddens et al15 may be attributable, at least in part, to overestimation of the Tanner breast stages of children with lipomastia. As the presence of lipomastia does not indicate activation of the hypothalamic-pituitary-gonadal unit, incorrect assignment of early Tanner stages could account for the lack of difference in age of menses observed. Herman-Giddens et al15 also reported that the mean age of onset of pubic hair development was advanced. Our data on pubic hair self-assessment in girls do not suggest that pubic hair Tanner stage was incorrectly assessed. However, breast tissue is widely considered the indication of pubertal onset, not pubic hair. Adrenarche may very well be advanced in the population studied by Herman-Giddens et al,15 but that does not indicate an earlier age of onset of puberty.
The issue of the use of palpation to assess breast tanner stage in the study by Herman-Giddens et al15 was addressed in a special article by Kaplowitz and Oberfield.17 That article reported that 39% of the subjects in the study by Herman-Giddens et al15 had Tanner staging done by both visual inspection and palpation, and overestimation by inspection was found in 4% of girls of all Tanner stages. It is unclear from that article whether the Tanner staging by palpation and visual inspection was done independent of one another. In addition, the article by Kaplowitz and Oberfield17 reported that 15% of the girls who were staged Tanner 2 by inspection had no palpable breast tissue on palpation.17 That the percentage of subjects who were at Tanner stage 1 and whose Tanner stage was overestimated by visual inspection is higher than the overall percentage of patients at all Tanner stages and whose Tanner stage was overestimated by visual inspection may further support the need for palpation to distinguish Tanner stage 2 from Tanner stage 1. The earlier onset of breast development in the study by Herman-Giddens et al15 may be attributable in part to overestimation of Tanner stage 2 because palpation was not used for assessment of breast Tanner stage in all patients.
In our study, both nonobese and obese boys significantly overestimated their pubic hair Tanner stage. Our observations are similar to previously reported results by Schlossberger et al,6 who found that boys tended to overestimate pubic hair stage at earlier stages of development. Thus, self-assessed pubic hair Tanner stage does not seem to be suitable for use in boys 6 to 12 years.
We conclude that although half of all 6- to 12-year-old girls can accurately assess breast Tanner stage, self-assessments of breast stage should not be considered an unbiased measure of breast development in young obese girls. Although never a substitute for a good physical examination in the clinical care setting, self-assessment of breast Tanner stage by nonobese girls, in addition to self-assessment of pubic hair Tanner stage by obese and nonobese girls, may be a reasonable substitute for Tanner stage determined by physical examination, for the purpose of large-scale, epidemiologic research studies when physical examination may not be feasible. However, boys’ self-ratings of pubic hair Tanner stage do not seem to reflect actual Tanner pubic hair stage and would not be a reliable, unbiased alternative.
This study was supported by Z-01-HD-04-00641 (to Dr Yanovski) and the National Center on Minority Health and Health Disparities.
- ↵Duke PM, Litt IF, Gross RT. Adolescents’ self-assessment of sexual maturation. Pediatrics.1980;66 :918– 920
- ↵Neinstein LS. Adolescent self-assessment of sexual maturation: reassessment and evaluation in a mixed ethnic urban population. Clin Pediatr (Phila).1982;21 :482– 484
- ↵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
- ↵Must A, Dallal GE, Dietz WH. Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2) and triceps skinfold thickness. Am J Clin Nutr.1991;53 :839– 846
- ↵Marshall WA, Tanner JM. Variations in pattern of pubertal changes in girls. Arch Dis Child.1969;44 :291– 303
- ↵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
- ↵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. Pediatrics.1999;104 :936– 941
- Copyright © 2002 by the American Academy of Pediatrics