This Article Abstract Full Text (PDF) Submit a related Letter to the Editor Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Haverkamp, F. Articles by Hoppe, C. Search for Related Content PubMed PubMed Citation Articles by Haverkamp, F. Articles by Hoppe, C.

Growth Retardation in Turner Syndrome: Aneuploidy, Rather Than Specific Gene Loss, May Explain Growth Failure

Department of Pediatrics (F.H., J.W.), Institute of Human Genetics (K.Z.), and Department of Statistics (C.H.), University of Bonn, 53113 Bonn; Department of Pediatrics, University of Munich (O.B.), Munich; Department of Pediatrics, University of Berlin (P.A.), Berlin; Department of Pediatrics, University of Essen (B.P.H.), Essen; Department of Pediatrics, University of Frankfurt (E.W.), Frankfurt; Department of Pediatrics, University of Heidelberg (M.B.), Heidelberg; Department of Pediatrics, University of Leipzig (E.K.), Leipzig; Department of Pediatrics, University of Kiel (C.J.P., W.G.S.), Kiel; and Department of Pediatrics, Municipal Hospital (R.M.), Krefeld, Germany

Address all correspondence and requests for reprints to: Fritz Haverkamp, M.D., Department of Pediatrics, Adenauerallee 119, 53113 Bonn, Germany. E-mail: f.haverkamp{at}uni-bonn.de

	Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
The etiology of short stature (SST) in Turner syndrome (TS) is still a subject of speculation. A variety of hypotheses have been put forward, from SST as a result of increased intrauterine tissue pressure after fetal lymphedema to haploinsufficiency of a specific growth gene(s). These hypotheses have various statistical-auxological implications on the growth distribution in TS. Empirical research has provided no clear evidence for any of these theories, but the well known correlation between patients’ and midparental height (MPH) could be established. The influence of undetected mosaic status has often been cited as a major problem in the investigation of growth in TS. However, an assessment of mosaic status (simultaneous analysis of karyotype and phenotype) and its effect on growth with inclusion of MPH has been not yet carried out for a large sample. The aim of this study was to evaluate growth and its complex relationship to mosaic status and MPH in TS.

In a mixed cross-sectional and longitudinal study we retrospectively analyzed the auxological and clinical data of 447 patients with a pure loss of X-chromosomal material (n = 381 with 45,X0; n = 66 mosaics). The 447 patients were selected from a series of 609 consecutive patients with TS. To assess the effect of mosaic status on growth, we computed a bifactorial analysis of variance (phenotype, karyotype), including MPH as a covariate.

In line with the mosaic hypothesis, we found a correlation between individual loss of X-chromosomal material and phenotypical expressivity. In contrast, no correlation was found with respect to growth. With respect to MPH, we found growth retardation (GR) even in those patients with "normal" height above the third percentile (-2 or more SD score).

The interindividual variance of GR in TS (comparable to growth variance in the normal population) seems to be unrelated to other TS-specific factors (e.g. mosaic status or single gene loss). Instead, both interindividual variance and the global growth shift distribution are best explained by the presence of an unspecific aneuploidic effect. Furthermore, consideration of patient height in relation to MPH should lead to a better understanding of the nature of GR in TS than the commonly used, strictly qualitative definition of SST.

	Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
DESPITE THE fact that patients with Ullrich-Turner syndrome (TS) were first described more than 60 yr ago (1, 2), the mechanisms underlying short stature (SST) are still far from being understood. Several hypotheses on the etiology of SST in TS have been suggested (for a review, see Ref. 3). These hypotheses have proposed 1) global genetic factors, such as chromosome imbalance with impaired cell proliferation due to aneuploidy; 2) specific genetic factors, such as nonrandom X-inactivation or the effects of various growth genes [e.g. haploinsufficiency of the SHOX gene (4)]; and 3) specific nongenetic factors such as fetal lymphedema, resulting in increased tissue pressure, which may be followed by intrauterine growth retardation (GR).

Hypotheses of types 2 and 3 propose specific, sometimes genetic, differences within the TS population that may perhaps explain the variance in growth or phenotype within this group. In contrast, hypotheses of type 1 refer to global genetic factors that define the syndrome. Such factors are clearly incapable of explaining the variance in growth or phenotype within the TS population, but they may account for the growth shift of the statistical height distribution in the whole TS population compared to healthy peers. SST is usually defined as a qualitative feature and is based on the statistical final height distribution of the corresponding age and sex group (cut-off, -2 SD score; third percentile). A few rare, contradictory articles have reported less obvious GR in patients with a karyotype of 45,X0 (5, 6) or, in contrast, in patients with double Xp (7, 8, 9, 10). However the vast majority of studies on growth and karyotype have failed to confirm correlations between specific genetic factors and GR within TS samples (11, 12, 13, 14, 15, 16, 17).

Instead, a correlation between patients and midparental height (MPH; a combined measure of father’s and mother’s height; see below), comparable with the normal population has been confirmed in several studies (8, 10, 12, 18). The investigation of an influence of X-chromosomal aneuploidy on growth is generally complicated by the possible existence of a still undetected mosaic status in TS. However, it is known that some patients with pure monosomy may have a minor phenotype, and some patients with a mosaic in the chromosomal analysis may have a severe phenotype. Both conditions are explained by interindividual variance in undetected mosaic status (19). Thus, an estimation of the degree of mosaic is still possible because patients with a more severe phenotype seem to have a larger amount of affected aneuploidic tissue and cells (20, 21). Both for this reason and because of the above-mentioned correlation between individual and parental height, our study included MPH as well as phenotype and karyotype, an approach that had not yet been systematically carried out in a large patient cohort.

	Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
A retrospective national multicenter study was carried out in nine German pediatric endocrinology centers (Berlin, Bonn, Essen, Frankfurt, Heidelberg, Kiel, Krefeld, Leipzig, and Munich). Cytogenetic, phenotypical, and auxological data and parental heights were collated from the records of 609 consecutive in- and out-patients at various developmental stages (repeated measures; years of recording, 1967–1992). As patients from Bonn (n = 124) were screened for external and internal malformations using a standardized protocol, we found a higher proportion of minor malformations and congenital lymphedema compared to the retrospective data from other clinics (n = 485). To increase the statistical power, we decided to merge both data files, but to refer exclusively to karyotype, internal malformations, and webbed neck.

Patients’ height measurements were compared to statistical height distribution of healthy German female peers and expressed as SD score. Patients’ final height was predicted by means of various models (projected final height, Bayley-Pinneau prognosis) (22, 23). Auxological data recorded after initiation of any growth-promoting therapy and estrogen replacement therapy were excluded from the analysis.

Karyotype analyses were carried out by standard methods in all patients (analysis of not less than 100 lymphocytes). To isolate the influence of X-chromosomal aneuploidy, patients with additional X-chromosomal (e.g. 45, X0/47, XXX; n = 22) or additional Y-chromosomal (n = 15) material as well as patients with complex karyotypes (X-iso- or X-ring chromosomes; n = 96) were excluded from the analysis. A total of 447 patients remained in the study.

Based on a bifactorial approach, we classified the degree of mosaicism by rating the phenotypical expressivity in patients belonging to the same karyotype group. Two karyotypical groups were defined: monosomic individuals (45,X0; n = 381) and all others (n = 66). The latter had either mosaic karyotype (45,X0/46,XX: n_mosaic = 51) or deletions [46X, del(Xp); 46X, del(X)(pter-p22); 46X, del(X)(pter-21); 46X, del(X)(q21-q24); 45X/46X, del(Xq); 46X, del(X)(pter-p12); 46X, del(X)(pter-p11); n_deletion = 15]. Three phenotypical groups (factor stages) were defined as follows: 1) minor phenotype (n = 182) with neither webbed neck nor internal (renal and/or cardiac) malformations, 2) medium phenotype (n = 172) with either webbed neck or internal malformations, and 3) severe phenotype (n = 93) with both webbed neck and internal malformations.

Parental height was also considered as a covariate. MPH was calculated as: MPH = ((H_F + H_M)/2) - 6.5 cm, with H_F as the father’s height and H_M as the mother’s height (height in centimeters) (24).

We defined five developmental stages: perinatal (birth), early childhood (1–6 yr), prepuberty (7–11 yr), puberty (12–15 yr), and postpuberty (16 yr +). These developmental stages correspond to the major stages of GR process in TS as revealed in our sample (see Fig. 1). GR is increasingly pronounced in early childhood and puberty, but seems to be stationary in pre- and postpuberty.

View larger version (25K): [in this window] [in a new window]   Figure 1. Growth retardation in TS. Height measurements are expressed as the SD score, referring to statistical height distributions of German female peer groups. Vertical dotted lines represent predefined developmental phases. In box plots, bars represent group medians, box edges represent 25th and 75th percentiles, and lines represent minimum and maximum after exclusion of outliers (distance from box edges >1.5-fold of box length).

The influences of karyotype and phenotype on growth were tested sequentially using single one-factorial ANOVAs. Due to the very low cell counts in mosaics with severe phenotype (maximum, n = 6), testing of interactional effects of karyotype and phenotype on growth in TS was impossible.

Increasing the power of statistical tests is of particular importance in cases where effects are weak or absent, as expected in this study. Therefore, we indirectly increased statistical power 1) by merging diverse data files to increase sample sizes, 2) by defining the significance level () as = 0.10, and 3) by renouncement of the significance level correction for multiple testing (e.g. according to the Bonferroni method). Note that by these steps the probability of confirming significant effects of karyotype or phenotype on growth was increased as far as justifiable.

	Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Correlation of karyotype and phenotype

For patients with complete loss of one X-chromosome (45,X0), there was a slight, but significant, increase in the risk of severe phenotypical malformations (Pearson’s ² = 17.63; df = 2; P < 0.001). Nonparametric correlation of phenotype and karyotype (Spearman) was r = 0.19 (P < 0.01), indicating a higher risk among 45,XO patients of presenting a more severe phenotype. This confirms the empirical basis for the dual approach of this study.

Effects of karyotype and phenotype on growth at different developmental stages

The mean height SD scores of the different karyotypical, phenotypical, and developmental groups are listed in Table 1. No effects of phenotype or karyotype on growth could be detected at any developmental stage using one-factorial ANOVAs, with the exception of a main effect of karyotype on growth in early childhood. However, the sample sizes of mosaic patients in early childhood and postpubertal phasis were very small (early childhood, n = 3; postpuberty, n = 6). Therefore, confirmation of a main effect of karyotype on growth in early childhood could not be taken as reliable.

Significant correlations between patient height and MPH were confirmed at all developmental stages (Table 2). In consequence, parents of short statured patients were significantly smaller than parents of normal statured patients as expressed in MPH differences (MPH_normal - MPH_{short statured}) at most developmental stages: MPH at birth, 1,78 cm (P < 0.05); in early childhood, 5.01 cm (P < 0.01); at prepubertal stage, 6.15 cm (P < 0.05); at puberty, 5.87 cm (P < 0.05); and at postpubertal stage, 6.61 cm (P = NS). Correlations seemed to be lower at birth and in the more progressive phases of the growth retardation process during early childhood and puberty. Further analyses revealed that the height of the father did not correlate with the SD score at birth and in early childhood (P > 0.10). In early developmental stages the height of the mother seemed to be a better predictor, even slightly better than MPH, which is the best predictor in later development.

According to the standard definition (height SD score below -2), 21% of all patients were short statured at birth (total available measurements, n = 325), 62.9% in early childhood (n = 62), 93.0% in the prepubertal phase (n = 100), 97.0% in puberty (n = 133), and 94.1% in the postpubertal phase (n = 51). Figure 2 shows the relation between patient height SD score and MPH. MPH is expressed as SD score based on the height distribution of healthy female peers.

View larger version (24K): [in this window] [in a new window]   Figure 2. Distribution of parental and patients’ height in SD score by karyotype groups. Dotted lines represent the limit of short stature according to standard definition, whereas the diagonal line represents equal patients’ and MPH SD score. Only 3 of 261 patients tended to be larger than their parents. These parents were from Essen (no. 12), Berlin (no. 324), and Munich (no. 455). SD scores were detemined at the age of 10 yr (no. 455), 14 yr (no. 324), and 15 yr (no. 455). In two of these cases patients’ measurements indicate short stature in patients as well as in parents. Cases 12 and 324 represent extreme values with regard to MPH or patient’s SD score, respectively.

Note that MPH distribution differed from that in the normal population (group mean of MPH SD score, -0.78; t test for simple samples; critical value, 0; t = -13.14; P < 0.001), mainly due to a larger proportion of short statured mothers (77 of 447, or 17.2%).

	Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
In our study on the influences of phenotype, karyotype, and MPH on growth in TS, there were two major findings: 1) individual growth is independent of karyotype/phenotype, but individual height correlates well with MPH, as previously observed in the normal population; 2) with respect to MPH, GR in TS is also present in patients with normal stature (height -2 SD score or more). Regarding the various hypotheses on the etiology of growth in TS, one may therefore conclude that GR in TS is explained neither by the presence of intrauterine fetal lymphedema and the resulting increased pressure on tissues nor by any undetected variable mosaic status. With respect to karyotype, our results are in line with the great majority of studies on growth (11, 12, 13, 14, 15, 16, 17). By analyzing both karyotype and phenotype and increasing statistical power, as described above, we were able to demonstrate that there is, in fact, not even a weak effect on growth. Thus, there is no empirical evidence that an individual’s degree of mosaic status influences growth in TS.

Our empirical findings have far-reaching implications for the recent discussion concerning the role of single growth gene(s) on the X-chromosome. Our data show that interindividual growth variance in TS cannot be explained by an individual’s degree of loss of X-chromosomal DNA material. The observed empirical growth pattern clearly contradicts the hypotheses that assume interindividual variation in haploinsufficiency for a growth gene(s) as postulated for the SHOX gene (4). In this case, one would expect patients with normal appearance and mosaic karyotype (e.g. 45,X0/46,XX) to be less growth retarded (at least relative to MPH), because in these patients a lesser degree of haploinsufficiency would be expected than in patients with classical monosomy and extreme phenotype.

Our finding of similar growth retardation in relation to MPH regardless of whether individual height is above or below -2 SD score contradicts those hypotheses based on the qualitative definition of SST (height below -2 SD score) commonly used in TS. This finding again calls into question recent speculation about the possible etiological relevance of the SHOX gene (4), as in this model growth retardation has been strictly defined as height below -2 SD score. There have been several reports of patients with Turner features and height -2 SD score or more who simultaneously had an intact pseudoautosomal region, including double dosage of the SHOX gene (4). However, neither are TS features exclusively restricted to patients with TS nor did these reports contain any information about the MPH of these patients. It thus remains unclear whether these patients are comparable with TS patients and really had normal growth in view of parental height.

The use of MPH in patients with SST may occasionally be problematic, as there could be a correlation in cases where one or both parents carry the same growth gene mutation as the child. However, in TS haploinsufficiency a mutation for SHOX gene is assumed (4). Furthermore, parents of TS patients are usually not growth retarded and are therefore extremely unlikely to carry a growth gene mutation such as SHOX.

Pediatric studies on growth in TS should be assessed for specific selective effects. For example, TS patients with normal stature, normal pubertal development, and no external/internal malformations may never be presented and could thus have been systematically excluded from our investigation. Several studies have demonstrated that patients with spontaneous puberty more often present mosaic karyotype, but do not differ significantly with respect to their final height from those patients with either no or reduced pubertal development (10, 25). Furthermore, one might assume that of the TS subpopulation with normal appearance, only short statured patients were recorded. Based on the established correlation with MPH, one would then expect that the parental height of these patients would tend to be shorter than the parental height of patients with distinct features or with 45,X0 karyotype. This was, however, not the case, indicating that these subpopulations were, in fact, representative [MPH: monosomy (n = 301), 163.2 ± 5.7 cm; mosaics (n = 55), 162.3 ± 5.5 cm; P > 0.05, by Student’s t test].

The global growth shift in TS is probably caused by genetic factors that all TS girls have in common. In contrast, the interindividual variance in growth retardation in TS obviously cannot be explained by TS-specific genetic factors (e.g. individual loss of specific X-chromosomal material such as the SHOX gene) (4). The lack of influence of mosaic status and the simultaneous correlation between height and MPH, comparable to that in the normal population, empirically underline the old view that GR in TS must be defined as a complex qualitative trait resulting from aneuploidy or other forms of chromosomal imbalance, as is presumed for other chromosomal disorders often associated with pathological GR [e.g. Down’s syndrome (26)], but not with rearrangements, single gene deletions, or SHOX gene haploinsufficiency (4) in pseudoautosomal region.

In the future, a more comprehensive understanding of the nature of GR in TS may be expected if both patient height and MPH are considered, in contrast to a strictly qualitative definition of SST.

Received March 8, 1999.

Revised August 12, 1999.

Accepted August 19, 1999.

	References

Top Abstract Introduction Subjects and Methods Results Discussion References

 

1. Ullrich O. 1929 Über typische Kombinationsbilder multipler Abartungen. Monatschr Kinderheilkd. 49:271–276.
2. Turner HH. 1938 A syndrome of infantilism, congenital webbed neck and cubitus valgus. Endocrinology23 :566–574.
3. Ogata T, Matsuo N. 1995 Turner syndrome, and female sex aberrations: deduction of the principal factors involved in the development of clinical features. Hum Genet. 95:607–629.[Medline]
4. Rao E, Weiss B, Fukami M, et al. 1997 Pseudoautosomal deletions encompassing a novel homeobox gene cause growth failure in idiopathic short stature and Turner syndrome. Nat Genet. 16:54–63.[CrossRef][Medline]
5. Mazzanti L, Nizzoli G, Tassinari D, et al. 1994 Spontaneous growth and pubertal development in Turner’s syndrome with different karyotypes. Acta Pædiatr Scand. 83:299–304.
6. Lenko HL, Perheentupa J, Söderblom A. 1979 Growth in Turner’s syndrome: spontaneous and fluoxymesterone stimulated. Acta Pædiatr Scand. 277(Suppl):57–63.
7. Low LC, Sham C, Kwan E, et al. 1997 Spontaneous growth in Chinese patients with Turner’s syndrome and influence of karyotype. Acta Paeiatr Scand. 86:18–21.
8. Cohen A, Kauli R, Pertzelan A, et al. 1995 Final height of girls with Turner’s syndrome: correlation with karyotype and parental height. Acta Paeiatr Scand. 84:550–554.
9. Snider ME, Solomon IL. 1974 Ultimate height in chromosomal gonadal dysgenesis without androgen therapy. Am J Dis Child. 127:673–674.[Medline]
10. Massa G, Vanderschueren-Lodeweyckx M, Malvaux P. 1990 Linear growth in patients with Turner syndrome: influence of spontaneous puberty and parental height. Eur J Pediatr. 149:246–250.[CrossRef][Medline]
11. Rochiccioli P, David M, Malpuech G, et al. 1994 Study of final height in Turner’s syndrome: ethnic and genetic influences. Acta Paeiatr Scand. 83:305–308.
12. Bernasconi S, Larizza D, Benso L, et al. 1994 Turner’s syndrome in Italy: familial characteristics, neonatal data, standards for birth weight and for height and weight from infancy to adulthood. Acta Paeiatr Scand. 83:292–298.
13. Naeraa RW, Nielsen J. 1990 Standards for growth and final height in Turner’s syndrome. Acta Paeiatr Scand. 79:182–190.[Medline]
14. Karlberg J, Albertsson-Wikland K, Nilsson KO, Ritzen EM, Westphal O. 1991 Growth in infancy and childhood in girls with Turner’s syndrome. Acta Paeiatr Scand. 80:1158–1165.[Medline]
15. Ranke MB, Grauer M-L. 1994 Adult height in Turner syndrome: results of a multinational survey 1993. Horm Res. 42:90–94.[Medline]
16. Sybert VP. 1984 Adult height in Turner syndrome with and without androgen therapy. J Pediatr. 104:365–369.[Medline]
17. Pelz L, Timm D, Eyermann E, et al. 1982 Body height in Turner’s syndrome. Clin Genet. 22:62–66.[Medline]
18. Brook CG, Gasser T, Werder E, Prader A, Vanderschueren-Lodewykx MA. 1977 Height correlations between parents and mature offspring in normal subjects and in subjects with Turner’s and Klinefelter’s and other syndromes. Ann Hum Biol. 4:17–22.[CrossRef][Medline]
19. Ferguson-Smith MA. 1993 Genotype-phenotype correlations in the Turner syndrome. In: Hibi I, Takano K, eds. Basic and clinical approach to Turner syndrome. Amsterdam: Elsevier; 17–25.
20. Zinn AR, Page DC, Fisher EMC. 1993 Turner syndrome: the case of the missing sex chromosome. Trends Genet. 9:90–93.[CrossRef][Medline]
21. Amiel A, Kidron D, Kedar I, et al. 1996 Are all phenotypically-normal Turner syndrome fetuses mosaics? Prenat Diagn. 16:791–795.[CrossRef][Medline]
22. Lyon AJ, Preece MA, Grant DB. 1985 Growth curve for girls with Turner syndrome. Arch Dis Child. 60:932–935.[Abstract]
23. Bayley N, Pinneau S. 1952 Tables for predicting adult height from skeletal age: revised for the use of Greulich-Pyle standards. J Pediatr. 40:423–441[CrossRef][Medline]
24. Tanner JM, Goldstein H, Whitehouse RH. 1970 Standards for children‘s height at ages 2–9 years allowing for height of parents. Arch Dis Child. 45:755–762
25. Page LA. 1993 Final heights in 45,X Turner’s syndrome with spontaneous sexual development. Review of European and American reports. J Pediatr Endocrinol. 6:153–158.[Medline]
26. Shapiro B. 1997 Whither Down syndrome critical regions? Hum Genet. 99:421–423.[CrossRef][Medline]

This article has been cited by other articles:

				M. R. Perkiomaki and L. Alvesalo Palatine ridges and tongue position in Turner syndrome subjects Eur J Orthod, April 1, 2008; 30(2): 163 - 168. [Abstract] [Full Text] [PDF]

				W. F. Blum, B. J. Crowe, C. A. Quigley, H. Jung, D. Cao, J. L. Ross, L. Braun, G. Rappold, and for the SHOX Study Group Growth Hormone Is Effective in Treatment of Short Stature Associated with Short Stature Homeobox-Containing Gene Deficiency: Two-Year Results of a Randomized, Controlled, Multicenter Trial J. Clin. Endocrinol. Metab., January 1, 2007; 92(1): 219 - 228. [Abstract] [Full Text] [PDF]

				V. P. Sybert and E. McCauley Turner's Syndrome N. Engl. J. Med., September 16, 2004; 351(12): 1227 - 1238. [Full Text] [PDF]

				E. O. Reiter, S. L. Blethen, J. Baptista, and L. Price Early Initiation of Growth Hormone Treatment Allows Age-Appropriate Estrogen Use in Turner's Syndrome J. Clin. Endocrinol. Metab., May 1, 2001; 86(5): 1936 - 1941. [Abstract] [Full Text]

This Article Abstract Full Text (PDF) Submit a related Letter to the Editor Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Haverkamp, F. Articles by Hoppe, C. Search for Related Content PubMed PubMed Citation Articles by Haverkamp, F. Articles by Hoppe, C.