| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
Department of Medicine (I.L.S., J.N.H., M.R.P.), Division of Endocrinology, Children’s Hospital, and Department of Genetics (J.N.H.), Children’s Hospital and Harvard Medical School, Boston, Massachusetts 02115; Center for Genome Research (J.N.H.), Whitehead/MIT, Cambridge, Massachusetts 02139; and Department of Pediatrics (M.R.P.), Division of Pediatric Endocrinology and Metabolism, Rainbow Babies and Children’s Hospital, University Hospitals of Cleveland, Case Western Reserve University School of Medicine, Cleveland, Ohio 44106
Address all correspondence and requests for reprints to: Mark R. Palmert, M.D., Ph.D., Division of Pediatric Endocrinology and Metabolism, Rainbow Babies and Children’s Hospital, University Hospitals of Cleveland, 11100 Euclid Avenue, Cleveland, Ohio 44106. E-mail: mrp13{at}po.cwru.edu
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
2 SD beyond the mean) and relaxed criteria (pubertal delay 1 SD beyond the mean). These pedigrees were compared with 25 control pedigrees. Mean age of menarche was 14.3 ± 1.4 yr for mothers of CD probands vs. 12.7 ± 1.4 yr for mothers of controls (P < 0.0001). Thirty-eight percent of CD mothers met the strict 2 SD criteria, and an additional 29% met the relaxed 1 SD criteria for pubertal delay. By contrast, among the control mothers, 12% met the strict and an additional 8% met the relaxed criteria (P < 0.0001 for comparison with CD mothers). CD fathers were also more likely than the control fathers to have a history of pubertal delay. For first-degree relatives, the estimated relative risk of meeting the 2 SD and 1 SD criteria for delay in CD vs. control pedigrees were 4.8 and 4.9, respectively; estimated relative risk for second-degree relatives were 3.2 and 4.4, respectively. Inheritance patterns varied, but many families showed an apparent autosomal dominant pattern, with or without incomplete penetrance. Although many genes may underlie CD, the inheritance patterns suggest that there are also single genes with major effects whose penetrance is likely affected by genetic or environmental modifiers. The future identification of these major and modifying genes is an exciting prospect that would improve our understanding of the factors that regulate human pubertal timing and modulate the human reproductive endocrine axis. |
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Although populations of individuals with CD should theoretically be enriched for genes that modulate pubertal timing, no formal analysis of the genetic basis of this condition has been performed. Modes of inheritance and the degree of familial aggregation are unknown. Because identification of the genes that underlie CD would not only further our understanding of this condition but also identify important modulators of the human reproductive endocrine axis, we studied 53 CD pedigrees to determine general pattern(s) of inheritance and estimate the degree of familiarity. These data represent important first steps toward identifying the genes that underlie CD.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Using endocrine division and hospital-wide databases (17, 18), we identified adolescents seen for delayed puberty in the endocrine clinic at Children’s Hospital in Boston. Identification was based on problem list and diagnosis coding, and only individuals meeting strict criteria for pubertal delay were considered for participation. Individual records were then reviewed, and eligible families were contacted regarding participation in a genetic study that involved acquisition of an extensive family history of pubertal timing and blood samples for DNA isolation from probands and parents. Forty-one families agreed to participate and met our inclusion criteria for CD probands [no underlying medical conditions; girls with lack of breast development by age 13 yr and boys with lack of testicular enlargement (testis size <2.5 cm in length or <4 cc in volume) by age 14 yr; and longitudinal growth patterns that were consistent with those typical of CD (19)]. Twelve other families met the inclusion criteria and had sufficient data documented in their medical records to permit analysis. All 53 probands (40 boys and 13 girls) had documented spontaneous pubertal development. The vast majority of subjects were drawn from clinic visits that occurred between 1995 and 2000, and we estimate that 20% of eligible subjects were included in the study. To estimate the degree of familial aggregation of CD and compare the histories of delay seen among our CD subjects with those found among the general population, we also interviewed 25 families of other patients seen in our endocrine clinic for disorders not related to pubertal timing.
We interviewed probands and parents, but the total study group included 394 relatives of CD probands and 204 relatives of control probands for whom information was available regarding the timing of puberty. Information regarding ethnicity was not available for six of the CD probands for whom data regarding family history were derived from medical records. Of the remaining CD probands, 42 of 47 (89%) declared themselves as white, non-Hispanic; 2 of 47 (4%) as Hispanic; 2 of 47 (4%) as Cape Verdian; and 1 of 47 (2%) as mixed Asian/white. Similarly, 23 of 25 (92%) of the control pedigrees declared themselves as white, non-Hispanic and 2 of 25 (8%) as Hispanic. The 12 CD probands whose histories were analyzed through medical records and 21 of 41 CD probands whose families were interviewed were among the 232 subjects included in a retrospective case series regarding the etiologies and characteristics of adolescents with delayed puberty (14); however, none of the characteristics of the controls, and no data from the analysis of inheritance patterns, have been reported previously.
Study design
The record review and interviews were approved by the Children’s Hospital Committee on Clinical Investigation. Informed consent and, when appropriate, assent was obtained before all interviews, which were conducted by a single interviewer using a structured questionnaire.
Knowing that the timing of puberty is influenced by genetic background suggests that subjects with CD might have families that are enriched for individuals with later-than-average pubertal development, even if not necessarily late enough to meet diagnostic criteria for delayed puberty. Because previous data supported this concept (14), we assessed family histories of pubertal delay in all pedigrees using two different classifications. In the first analysis, strict criteria for delay were employed, and family members were designated as affected if pubertal development occurred 2 SD or more beyond the mean (men with pubertal onset after age 14 yr and/or growth spurt after age 16 yr; women with menarche after age 15 yr) (2, 20, 21, 22, 23, 24, 25). In the second analysis, relaxed criteria were employed wherein affected status required pubertal development 1 SD or more beyond the mean (men with pubertal onset after age 13 yr and/or growth spurt after age 15 yr; women with menarche after age 14 yr).
Histories of pubertal timing were obtained by first asking about the timing of pubertal events in relationship to peers, then by asking age of growth spurt and attainment of final height, and finally by asking for the timing of specific events (such as development of pubic hair, onset of breast development, age of menarche, and timing of genital growth). Families were aware that extensive histories of pubertal timing would be taken during the research visit, and any missing information was collected, if possible, during follow-up phone calls. The available data were used to classify the timing of puberty as detailed previously.
Data analysis
All three investigators analyzed each pedigree, and the most likely inheritance pattern for the CD pedigrees was determined by inspection. None of the families reported a history of consanguinity. Apparent autosomal dominant transmission was characterized by involvement of two generations or more (usually successive, although a generation may have been skipped in some instances of incomplete penetrance); additional features may have included observation of affected females and male-to-male transmission. Characteristics of apparent autosomal recessive transmission included males and females being affected, all affected individuals deriving from one generation, and both parents being unaffected. A family was classified as X-linked if only males were affected, transmission occurred through female carriers, and male-to-male transmission did not occur. When only the proband was affected, inheritance was defined as sporadic. Pedigrees were considered bilineal/unclassifiable if, for all individuals in the pedigree with available parental phenotypes, one of the following two patterns was observed: 1) all siblings of the individual were affected and both parents were affected, or 2) any unaffected parents themselves had an affected sibling or parent.
To calculate the estimated relative risk of meeting criteria for pubertal delay among CD family members (
), the percentage of affected family members among the CD population was divided by the percentage of affected family members within the control group (26). Primary relatives (full siblings and parents of the proband) and secondary relatives (half-siblings of the proband, siblings of the proband’s parents, and the grandparents of the proband) were considered. To determine the 95% confidence intervals for , we sampled 10,000 times from a pair of binomial distributions, using sample sizes that were the same as the actual numbers of primary (or secondary) relatives in our two populations. One of these distributions had the expected frequency of affected relatives estimated from the CD pedigrees, and the other distribution had the expected frequency estimated from control pedigrees. For each of the 10,000 simulated pairs of samples, was calculated. The 95% confidence interval was defined as the range between the 2.5% lowest and the 2.5% highest relative risks in the 10,000 simulations. Ages of menarche among CD and control groups were compared using a two-tailed t test for independent samples; percentages of mothers in the two groups with delayed pubertal timing were compared using a 
2 test with 2 degrees of freedom. The t test and 2 test were performed using the Complete Statistical System: Statistica from StatSoft, Inc. (Tulsa, OK); significance was attributed to two-tailed P less than 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We next estimated the of meeting the 2 SD and 1 SD criteria for delay in CD vs. control pedigrees for first- and second-degree relatives of the probands. For first-degree relatives, was 4.8 for 2 SD and 4.9 for 1 SD criteria, for second-degree relatives was 3.2 for 2 SD and 4.4 for 1 SD criteria (Table 1
). Because can be confounded by increased ascertainment of families with multiple affected siblings, we also calculated P, which limited the analysis to estimating the relative risk for parents of the probands. Similar to the original calculation, P for strict 2 SD criteria was estimated to be 5.2 and P for relaxed 1 SD criteria was estimated to be 4.3 (data derived from those parents with available data regarding pubertal timing, 100 CD parents and 47 control parents).
|
2 SD = 2.3% of a normally distributed population), we expected a small number of the control pedigrees to have members with pubertal delay. In this regard, it is interesting to note that if these two families are excluded from the analysis, 3.2% of the subjects in control pedigrees met strict criteria 2 SD for pubertal delay and a total of only 5.9% met the relaxed 1 SD criteria. Thus, in total, data from CD and control pedigrees indicate that families of subjects with CD are very highly enriched for individuals who themselves have delayed pubertal development.
Finally, we examined the apparent mode of inheritance of CD in the 53 pedigrees according to the criteria set forth in Subjects and Methods. As seen in Table 2, the apparent inheritance pattern varied among the CD pedigrees, but there was clear evidence of autosomal dominant inheritance, with either complete or incomplete penetrance. Examples of pedigrees consistent with different modes of inheritance are shown in Figs. 1 and 2. The most commonly observed patterns were similar to those exhibited in Fig. 1, A, B, and E.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We acknowledge that our analysis has limitations. It is possible that families with more dramatic histories of delayed puberty may have been more willing to participate in our study than families with negative histories. If so, our results may not be fully representative of the general CD population. For example, sporadic cases of CD would be underrepresented and the estimated relative risk of delay among CD relatives () would be falsely elevated. However, because our primary goal was to investigate modes of inheritance in the more familial cases of CD, this potential bias, if present, may have aided our ability to define the full spectrum of inheritance patterns. In addition, the examples of the two control pedigrees with strong clustering of CD cases (and the general paucity of sporadic CD cases in the remaining control pedigrees) suggested that a high degree of familial aggregation of CD is not atypical in the general population.
The designation of the apparent inheritance pattern for each pedigree was not always clear-cut. The high number of bilineal or unclassifiable pedigrees suggests complex, multifactorial inheritance of pubertal timing. However, some of these pedigrees could have arisen from assortative, nonrandom mating, a possibility supported by our observation that in 17 of 53 (32%) of the CD probands, both parents met at least the relaxed 1 SD criteria (Fig. 2A
). Alternatively, the high rate of both parents being affected could stem from ascertainment bias because of increased penetrance of phenotype in the proband, increased likelihood of referral to the Endocrine program, and/or increased willingness to participate in this study. Regardless of the cause, the presence of two affected parents made determination of the mode of inheritance difficult for certain pedigrees and may have led to our underestimating the number of pedigrees showing autosomal dominant inheritance.
In addition, some pedigrees could theoretically be attributed to more than one mode of inheritance. For example, some pedigrees classified as autosomal recessive could actually represent autosomal dominant inheritance with incomplete penetrance in the parent carrying the causal genetic variant. To quantify the number of instances with no family history, we classified pedigrees with only one affected individual as sporadic. We recognize, however, that some sporadic pedigrees could have arisen from autosomal recessive inheritance or from autosomal dominant inheritance with incomplete penetrance. Finally, it is formally possible that some pedigrees with apparent dominant inheritance are actually pseudodominant, a pattern that can be seen when a recessive, causative allele is present at high frequency within the general population and when an affected individual marries a carrier. Until the underlying genes and responsible allelic variants are identified, it will be impossible to determine definitively the modes of inheritance in all CD pedigrees. However, it seems unlikely that a high number of autosomal recessive pedigrees were misclassified as forms of autosomal dominant or sporadic inheritance because such misclassification would depend on an unexpectedly high frequency of a small number of causative allele(s) within the general population.
Determination of inheritance patterns can also be confounded by potentially imprecise data. With the exception of extremely long-term prospective studies, family history data regarding pubertal timing can be obtained only through recall, which introduces an element of inaccuracy into the data. However, as we have discussed previously (14), recalled data are valuable because, even decades after the event, studies indicate that 75–90% of women remember their age of menarche and 50% of men remember the timing of their pubertal growth spurt within a year (27, 28, 29). Moreover, the overall validity of the histories obtained from the CD pedigrees is evidenced by the striking contrast between these histories and those obtained from the control pedigrees. Although the possibility of an ongoing secular trend toward earlier pubertal development (2, 30, 31, 32) could complicate assessment of pubertal timing among parents and grandparents, the control families report histories of pubertal timing that are consistent with current general population norms. None of the control fathers reported a history of late puberty and the control mothers had an average age of menarche (12.7 ± 1.4 yr) that is consistent with recent U.S. population data [12.9 ± 1.2 yr, (2)].
Our findings expand on our previous observation that many family members of CD probands experienced later-than-average pubertal development, although not always late enough to meet strict criteria for pubertal delay. As seen in Table 2 and Fig. 2, this finding has important implications when assessing inheritance patterns. Regardless of the criteria employed, several different inheritance patterns were observed in the study population. However, when the relaxed 1 SD criteria were employed, the number of families with sporadic and autosomal recessive patterns decreased, whereas the number with dominant patterns increased. This could be an artifact of the relaxed criteria but more likely suggests that dominant inheritance, with variable penetrance, is a predominant inheritance pattern within CD families.
That several different inheritance patterns were observed in the study population suggests that multiple genes (possibly acting through different genetic models) likely modulate the timing of puberty in humans. This is consistent with the hypothesis that multiple genes, perhaps with additive effects, underlie the genetic regulation of complex traits (33, 34). However, the high proportion of families with apparent autosomal dominant inheritance in our study was an unexpected result and indicates that some of the genetic variation in the timing of puberty may stem from variation in a few genes with major effects. If so, the apparently heterogeneous inheritance patterns could also reflect the action of modifiers of the major genes (either other genes or environmental factors). The possibility that such major genes underlie some cases of CD is an important insight that suggests that the genetic basis of CD may be tractable to traditional methods of genetic analysis, such as genome-wide linkage studies.
Many genes are good candidates to harbor variants that underlie the genetic component of CD. These could include subtle variation in the genes encoding GnRH or its receptor or variation in upstream regulators such as leptin or its receptor. The subsequent identification of major and modifying genes is an exciting prospect that would improve our understanding of the factors that regulate human pubertal timing and modulate the human reproductive endocrine axis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: CD, Constitutional delay of growth and maturation; , relative risk for pubertal delay; p, relative risk for pubertal delay among parents of probands.
Received June 3, 2002.
Accepted September 12, 2002.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  I Banerjee, D Hanson, R Perveen, A Whatmore, G C Black, and P E Clayton Constitutional delay of growth and puberty is not commonly associated with mutations in the acid labile subunit gene. Eur. J. Endocrinol., April 1, 2008; 158(4): 473 - 477. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Wehkalampi, E. Widen, T. Laine, A. Palotie, and L. Dunkel Patterns of Inheritance of Constitutional Delay of Growth and Puberty in Families of Adolescent Girls and Boys Referred to Specialist Pediatric Care J. Clin. Endocrinol. Metab., March 1, 2008; 93(3): 723 - 728. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Bhasin Experiments of Nature -- A Glimpse into the Mysteries of the Pubertal Clock N. Engl. J. Med., August 30, 2007; 357(9): 929 - 932. [Full Text] [PDF]
|
|
|
|
 |
|
 A. Dempfle, S. A. Wudy, K. Saar, S. Hagemann, S. Friedel, A. Scherag, L. D. Berthold, G. Alzen, L. Gortner, W. F. Blum, et al. Evidence for involvement of the vitamin D receptor gene in idiopathic short stature via a genome-wide linkage study and subsequent association studies Hum. Mol. Genet., September 15, 2006; 15(18): 2772 - 2783. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Rothenbuhler, D. Fradin, S. Heath, H. Lefevre, C. Bouvattier, M. Lathrop, and P. Bougneres Weight-Adjusted Genome Scan Analysis for Mapping Quantitative Trait Loci for Menarchal Age J. Clin. Endocrinol. Metab., September 1, 2006; 91(9): 3534 - 3537. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. L. Sedlmeyer, C. L. Pearce, J. A. Trueman, J. L. Butler, T. Bersaglieri, A. P. Read, P. E. Clayton, L. N. Kolonel, B. E. Henderson, J. N. Hirschhorn, et al. Determination of Sequence Variation and Haplotype Structure for the Gonadotropin-Releasing Hormone (GnRH) and GnRH Receptor Genes: Investigation of Role in Pubertal Timing J. Clin. Endocrinol. Metab., February 1, 2005; 90(2): 1091 - 1099. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. M. Plant and M. L. Barker-Gibb Neurobiological mechanisms of puberty in higher primates Hum. Reprod. Update, January 1, 2004; 10(1): 67 - 77. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
