Cortical Bone Measurements in Turner’s Syndrome ELIZABETH PARK Uniuersity oflroronto, Toronto, Ontario, Cuna& M5S I A l

K E Y WORDS Turner’s syndrome . Cortical width - Osteoporosis

ABSTRACT Cortical bone width measurements taken at midshaft on the sec- ond metacarpal were obtained from 156 hand X-rays of 80 karyotypically docu- mented individuals with Turner’s syndrome age 1 to 25 years. Total shaft width, medullary width, cortical width and percent cortical area were grouped by bone age and compared with normal female standards. Total width was significantly and increasingly below normal; medullary width was not consistently different from normal; cortical width was significantly lower from normal from age 14 onward, al- though it did rise at age 17 (adult bone age); percent cortical area was significantly below normal at ages 14 and 15, but was normal by adulthood. Values for percent cortical area did not indicate severe or widespread osteoporosis.

Within the Turners sample cortical bone measurement were not significantly de- creased in the presence of the XO sex chromosome constitution compared with other sex chromosome variants. Nor were the measurements decreased in the pres- ence of positive metacarpal sign or a combination of typical Turner stigmata (web neck, low posterior hairline, shield chest). There was evidence that cortical width and percent cortical area increased significantly following estrogen treatment or spontaneous menarche.

Turner’s syndrome (Turner, ’38) is characterized by short stature (adult height less than 5’, Haddad and Wilkins, ’59; Lindsten, ’63) and a broad spectrum of congenital abnormalities, notably lymph- edema of the newborn, shield chest, coarctation of the aorta and horseshoe kidney (Smith, ’70). The syndrome has a genetic basis in the absence of one X chro- mosome, or a part thereof, in some or all of the cells, and an anatomical basis in the congenital absence of functioning ovaries, present only as streaks of fibrous tissue. Typically this results in the failure of nor- mal pubertal changes to occur, as well as high urinary gonadotrophin titres and low urinary estrogen and 17-ketosteroid ex- cretion levels in the second decade of life. Occasionally spontaneous pubertal changes occur. In cases of pubertal failure, cyclical estrogen therapy results in the de- velopment of secondary sexual characteris- tics and menstruation.

In addition to delayed bone age after the expected age of puberty, (Acheson and Zampa, ’61; Dalla Palma et al., ’67) a num- ber of structural skeletal abnormalities are present, such as short 4th metacarpal, short metatarsals, enlarged and lowered medial femoral condyle and enlarged medial tibia1 metaphysis. A generalized abnormality of bone structure has been noted by many workers (Haddad and Wilkins, ’59; Levin, ’62; Steinbach, ’64; Engel and Forbes, ’65; Preger et al., ’68). The bone is described as being rarefied and having an abnormal tra- becular pattern with widely spaced tra- beculae. These abnormalities, frequently labelled “osteoporosis”, have been de- scribed for both children and adults al- though they appear more frequently in adults.

Garn and Poznanski (’73) have published quantitative data on the degree of osteo- porosis in 29 Turners, assessed from corti- cal width measurements. These data, how-

455 AM. J. PHYS. ANTHROP., 46: 455-462.

456 ELIZABETH PARK

ever, represent the average of measure- ments taken on individuals of different ages. In the present study a larger sample is used and cortical width measurements are compared with normal standards from childhood to maturity. Furthermore, the association of certain Turner stigmata with cortical width measurements is investi- gated and the effect of estrogen treatment/ menarche on cortical thickness is assessed.

MATERIALS AND METHODS

A total of 156 pairs of standard bone age X-rays of 80 Turners of known karyotype and age 1 to 25 years were studied. Most in- dividuals had one or two X-rays, at different ages, and a few had three. The age at initiation of estrogen treatment (%=17.0 years, N=29), or more rarely, spontaneous menarche (x =13.4 years, N =lo), was known. Using dial-reading calipers, total bone width (TW) and medul- lary cavity width (MW) were measured by the author to the nearest 0.05 mm at mid- shaft on the second metacar a1 according to the method of Garn ('70) rfig. 1). Corti- cal width (CW) was calculated by subtract- ing MW from TW, and percent cortical area (%CA) was calculated, again after Garn ('70). Measuring crror, the average difference b e b ~ e n measurements taken on two separate occasions, for TW and MW

Fig. 1 Hand X-ray of a girl with Turner's syn- drome age 18.7 years (bone age 14.8 years). Estrogen treatment began at age 17.3 years. Note thin cortical bone.

TABLE 1

Signaficant differences an cortical width meusurements in Turners compared with normal females

Measurements

Total Meddlary Cortiral Percent cortical width width width area

Age N

6 6 NS NS NS NS 7 3 NS NS NS NS 8 4 NS NS NS NS

NS NS 9 5 NS

NS NS NS NS NS NS NS NS NS N S NS

NS

NS NS * Q * 0 * I

0

0

* (I

0

* * *

10 11 11 21 12 16 13 28 14 14 15 12 16 8 17 13

* 0 0

In all cases of significant differences, means for Turners were l e r b than for normal females. '. indicates significance.

CORTICAL BONE MEASUREMENTS IN TURNERS SYNDROME 457

8

7

6

5

E E

: 4 o_ 3

3

2

1 2 4 6 8 10 12 14 16

AGE

Fig. 2 Total shaft width and cortical width of the second metacarpal in Turners and normal frmales.

was 0.084 mm and 0.123 mm respectively. The x-rays were grouped by Greulich

and Pyle ('59) bone age and TW, M W , CW, and %CA compared statistically to recent standards for North American white females (Poznanski, '74). Because sample sizes were very small before age 6, statisti- cal comparisons began at age 6 and involved 141 X-rays of 71 individuals. Since differences between right and left sides were not significant at any of the ages, only the left side was used in this and subse- quent tests. In addition to the t-test two nonparametric tests, the signed rank and sign test, were used. For the former, Tur- ners' values were tested for skewness and kurtosis, and the Turners' and normal values tested for homogeneity of variance; heteroscedastic t-tests were used in those instances where variances were not homo-

geneous. The 0.05 level of significance was used . Z-scores were calculated for TW, CW and %CA.

Within the Turners sample, the effect of the following variables on cortical width parameters was studied using t-tests: (1) Sex chromosome constitution. Typical XO Turners were compared with Turners with mosaic and/or structural X abnormalities. (2) Metacarpal sign. Individuals with a short 4th metacarpal, and hence a positive metacarpal sign (Archibald et al., '59) were compared to those with a negative metacarpal sign. (3) Typical Turner pheno- type. The combination of stigmata which best discriminated more typical from less typical Turners in this sample was web neck, low posterior hairline and shield chest. Individuals with all and none of these stigmata were compared.

458 ELIZABETH PARK

I 1 I N 2 4 6 8 10 12 14 16

AGE

Fig. 3 Percent cortical area of the second metacarpal in Turners and normal fvmales.

.,* ,'.,

'... ._ I -. -. -. J'

- total width cortical width percent cortical area _ - _ _

t 2 4 6 8 10 12 14 16 2

Turners' Z-scores for total width, cortical width and percent cortical area of the second AGE

Fig. 4 metacarpal.

To study the effect of estrogen treat- ment on cortical width measurements the X-rays of 29 individuals who received cy- clical estrogen therapy and ten individuals with spontaneous menarche were grouped on the basis of 1 ,2 and 3 years before treat- ment/menarche and I, 2 and 3 years after treatment/menarche. As differences be- tween the treated and menarche groups were not significant, the two groups were not kept separate. T-test comparisons were made between the time groups.

HESULTS

The ages at which Turners differ sig- nificantly from normal are summarized in table 1. Total width in Turners is sig- nificantly below normal at all ages and the discrepancy increases with age up to age14 (fig. 2). Medullary width does not differ consistently. Cortical width is significantly lower after age 13, although it does show an increase by adulthood (bone age 17) (fig. 2). Percent cortical area falls sig-

CORTICAL BONE MEASUREMENTS IN TURNER'S SYNDROME 459

TABLE 2

Efect of estrogen treatment and menarche on cortical width mPusurements

Years before treatment Years after treatment nr mrnarc~he or iiicnarrhc

Measurement 3 2 I 1 2 3

Y 5 8 26 15 14 14

Total width (mm) 6.15 6.56 6.61 6.80 6.91 6.75

Cortical width (rnrn) 3.51 3.21 3.43 3.23 3.91 4.42 Percent cortical area 80.5 76.3 76.6 76.7 80.2 85.8

Medullary width (mm) 2.70 3.31 3.17 3.54 2.77 2.33

nificantly below normal at ages 14 and 15, but is not significantly different at ages 16 and 17 (fig. 3). Figure 4 shows these results plotted as Z-scores.

Within the Turners sample none of the variables examined, sex chromosome con- stitution, metacarpal sign or the more typi- cal Turner phenotype, was associated with significant differences in cortical width measurements.

The results of the examination of estrogen treatment/menarche are sum- marized in table 2. Total width increases slightly from three years before to three years after, but none of the differences are significant. Medullary width means show a slight increase from three years before to one year before and then decrease from two to three years after, with the result that the value for three years after is even less, although not significantly so, than the value for three years before. The mean value for cortical width changes little from three years before to one year before, but increases from one year before to two and three years after are significant. Mean values for percent cortical area are approx- imately the same from three years before to two years after, increasing markedly at three years after. The increases in percent cortical area from one year before to three years after and one year after to three years after are highly significant.

DISCUSSION

The amount of bone tissue present at any one time is the net result of the processes of resorption and apposition, and deficient bone mineralization can result from de-

creased apposition, or increased resorp- tion, or both. In situations where there is a net loss of bone, trabecular bone, with its higher surface to bone volume ratio, is lost faster than cortical bone. Bone loss, or osteoporosis, is usually considered to be a generalized phenomenon, so that a de- crease in skeletal mass is accompanied by a decrease in the width of compact bone in the appendicular skeleton (Collins, '66). Studies of the normal sequence of develop- mental changes at the two surfaces of tubu- lar bone (Garn, '70) show that the outer subperiosteal surface has a general pre- dilection for apposition and an adolescent spurt, especially in males, but with apposi- tion continuing throughout life. The inner endosteal surface exhibits a long period of resorption throughout childhood until the steroid-mediated phase of adolescent de- velopment, with apposition thereafter, and to a greater degree in females than in males.

Turners appear to follow the pattern of normal females, with the exception that they show a lag around the expected time of puberty. This might speculatively be re- lated to the relatively late age of treat- ment/menarche (X =16.4 years). Turners are small people, so it is not surprising that total width is less than in normal girls. Total width falls off after age 13 and stays low (fig. 2). Interestingly, this is also the age at which stature in Turners shows the greatest difference from normal, when Turners fail to manifest a growth spurt (Brook et al., '74). Cortical width and per- cent cortical area also falloff at the time of expected puberty, in agreement with the

460 ELIZABETH PhHK

findings of Preger t-: al. ('68) but they show an increase towards adulthood. The best measure of the relative amount of bone tis- sue is percent cortical area, since it is inde- pendent of absolute size. As adults, Turners are normal for percent cortical area. Garn ('70) suggests that percent cortical area reduced more than 2 SD below the mean for age and sex be taken as evidence of osteoporosis. In the present series the re- duction was not generally of this mag- nitude. The lowest mean value was -1.6 SD at age 16 (fig. 4). At age 15,25% of indi- viduals had percent cortical area < -2 SD. This was the highest frequency for any age; for example, at ages 13 and 17 the fre- quencies were 7.1% and 8.3% respectively. Garn and Poznanski ('73) present com- posite data for 29 Turners of unstated age, in which the mean percent cortical area was -1.6 SD, indicating a greater tenden- cy towards osteoporosis than in the present series. This figure of -1.6 SD was obtained by averaging the measurements of Turners of different ages. The great variation in percent cortical area with age found in the present study suggests that this is not justified, and may obscure maturity-related trends, which, in turn, could be important in understanding why Turners have a ten- dency to decreased percent cortical area at certain ages.

Cortical width parameters were not found to be significantly affected by side, sex chromosome constitution, typical Turn- er phenotype or positive metacarpal sign. However, there was evidence that some of the changes were associated with chang- ing estrogen levels. Albright et al. ('42) ini- tially suggested that the symptoms of osteoporosis in Turners were due to estrogen deficiency, drawing a parallel be- tween osteoporosis in Turner's syndrome and post-menopausal osteoporosis. In nor- mal females rising estrogen levels at puber- ty lead to endosteal apposition and an increase in cortical width and percent cor- tical area. Turners show a marked decrease in these parameters relative to normal starting at bone age 12. The fact that these parameters rise to near normal levels at

adulthood is suggestive of an association with estrogen, but more direct evidence comes from the observation that cortical width and percent cortical area increase significantly following the onset of estrogen therapy/menarche. Percent cortical area increased from 76.6% at one year prior to treatmedmenarche to 80.2% at two years after and 85.8% at three after. Total width increased little, following treatment/men- arche, while medullary width decreased markedly, indicating that the rise in per- cent cortical area is due almost entirely to increased endosteal apposition. In this re- spect it differs from the adolescent cortical width recovery phase in Down's syndrome, which is due to increased subperiosteal and endosteal apposition (Garn et al., '72).

Previous authors have reported that estrogen treatment has little or no effect on the symptoms of osteoporosis in Turners. Lindsten ('63) found that some patients had less evidence of widely spaced tra- beculae after treatment, while others did not change. Haddad and Wilkins ('59) and Finby and Archibald ('63) found no sig- nificant improvement in bone mineraliza- tion, trabecular pattern or cortical width following long periods of hormonal therapy. The conclusions of these workers, however, were based on scanty and unsystematic data. While the data in the present study are not based on large Sam- ples, the measurements are quantified and systematically analysed, and therefore con- stitute g,ood evidence for an association be- tween estrogen therapy and increased cor- tical width and percent cortical area in Turners.

One of the arguments put forth against the importace of estrogen deficiency in causing osteoporosis in Turners is that the symptoms of osteoporosis, particularly widely spaced trabeculae, are present in children with Turner's syndrome, whereas in normal females, osteoporosis does not become apparent until about ten years after the menopause (Finby and Archibald, '63; Preger et al., '68). There are at least two possible explanations for this apparent contradiction. One is that, although nor-

CORTICAL BONE MEASUREMENTS IN TURKER'S SYNDROME 46 1

mally in childhood gonadal hormones are produced in very small quantities, little is known about their functions (Hubble, '69). It is possible that estrogens are necessary for normal mineralization of the skeleton early in development, either stimulating bone formation directly or counterbalanc- ing the resorptive action of parathyroid hormone (Brown et al., '74; Dequeker, '75). A second possible explanation is that the widely spaced trabeculae constitute a defect in bone structure distinct from osteoporosis, and therefore a result of the operation of different factors.

ACKNOWLEDGMENTS

I would like to thank the Medical Re- search Council of Canada for financial assistance during this research. I am also indebted to Doctor K. 0. McCuaig and Doctor J. D. Bailey for their help.

LITEHATUFE CITED

Acheson, R. M., and G. A. Zampa 1961 Skeletal maturation in ovarian dysgenesis and Turner's s p - drome. Lancet, I: 917-920.

Albright, F., P. H. Smith and R. Fraser 1942 A syn- drome characterized by primary ovarian insuffi- ciency and decreased stature. Amer. J. Med. Sci.,

Archibald, R. M., N. Finby and F. DeVito 1959 Endo- crine significance of short metacarpals. J. Clin. En- docrinol., 42: 168-186.

Brook, C. G. D., G. Murset, M. Zachmann and A. Prader 1974 Growth in children with 45x0 Turner's syndrome. Arch. Dis. Child:, 49: 789-795.

1974 Osteoporosis in ovarian dysgenesis. J. Pediat., 84:

Collins, D. H. 1966 Pathology of Bone. Butter- worths, London.

Dalla Palma, I., C. Cavina and A. Borghi 1967 Skele- tal development in gonadal dysgenesis, female in phenotype. Amer. J. Roentgenol., 101: 876-883.

204: 625-649.

Brown, D. M., J. Jowsey and D. S. Bradford

816-820.

Dequeker, J. 1975 Bone and ageing. Ann. Rheiim. Dis., 34: 100-1 15.

Engel, E., and A. P. Forbes 1963 Cytological and clinical findings in 48 patients with congenitally defective or absent ovaries. Medicine, 44: 135-164.

Finby, N., and R. M. Archibald 1963 Skeletal abnor- malities associated with gonadal dysgenesis. Amer. J. Roentgenol., 89: 1222-1 235.

Garn, S. M. 1970 The Earlier Gain arid Later Loss of Cortical Bone. Chas. C. Thomas, Springfield.

Garn, S. M., J. C. Call Jr., and J. M. Nagy 1972 Prelim- inary radiogrammetric analysis of the bone recov- ery phase in adolescents with Down's syndrome. Invest. Radiol., 7: 97-101.

Garn, S. M., and A. K. Poznanski 1973 The simulta- neous radiogrammetric approach to the two bone surfaces. In: Clinical Aspects of Metabolic Bone Disease. B. Frame, A. M. Parfitt and H. Dunran, eds. Excerpta Medica hmsterdam, pp. 20-24.

Greulich. W. W., and S. I. Pple 1959 Radiographic Atlas of Skcletal Development of the Hand and Wrist. Stanford University Press, Stanford.

Haddad, H. M., and L. Wilkins 1959 Congenital anomalies associated with gonadal aphid . Review of 55 cases. Pediatrics, 23: 885-902.

Hubble, D. 1969 Endocrine control of growth and skeletal maturation. In: Pediatric Endocrinology. D. Hubble, ed. Blackwell Scientific Publications, Oxford and Edinbnrgh.

Levin, B. 1962 Gonadal dysgenesis, clinical and roentgenological manifestations. Amer. J. Koentgen-

Lindsten, J. 1963 The Nature arid Origin of X Chromosome Aberrations in Turner's Syndrome. A cytogenetical and clinical study of 57 patients. Alm- quist and Wiksell, Uppsala.

Poznanski, A. K. 1974 The Hand in Radiologic Diagnosis. W. B. Saunders Co.. Philadelphia and London.

Preger, L., M. B. Howard, L. Steinbach, P. Moskowitz, A. L. Scully and M. B. Goldberg 1968 Roentgen- ographic abnormalities in phenotypic females with gonadal dysgenesis. Amer. J. Roentgenol., 104: 899- 910.

Smith, D. W. 1970 Recognizable Patterns o f Human Malformation. W. R. Saunders Co., Philadel- phia and London.

Steinbach, H. L. 1964 Roentgen appearance of osteopcirosii'. Radiol. Clin. N. Amer., 2: 191-207.

Turner, H. H. 1938 A syndrome of infantilism, con- genital webbed neck and cubitus valgus. Endocri-

oI., 87: 1116-1127.

1101., 23: 566-574.