FERTII.ITY AI\D STERILITY
Vol. 24, No.1. January 1973
Copyright © 1973 by The Williams & Wilkins Co.
Printed in U.S.A.
DIABETES MELLITUS AND THYROID AUTOIMMUNITY IN GONADAL DYSGENESIS J. VAN CAM PENH OUT, M.D., F.R.C.S.(C)., A. ANTAKI, M.D.,
E. RASIO, M.D., PH.D.
Department of Obstetrics and Gynecology, and Department of Medicine, Notre-Dame Hospital, University of Montreal, P.Q., Canada
The association of gonadal dysgenesis and various endocrinopathies such as Addison's disease,1 hypothyroidism, 2 and diabetes 3 has been reported. One of the still controversial questions is related to the frequency of diabetes mellitus in patients with gonadal dysgenesis and in their relatives. The type of diabetes and its pathogenesis in gonadal dysgenesis are poorly understood because of the many possible hormonal factors involved, the peculiarities of body conformation, and the resulting lack of appropriate control studies. In this presentation, 18 patients with clinically and cytogenetically well-defined types of gonadal dysgenesis were investigated. Thyroid and pituitary-adrenal functions were studied in all patients. Diabetes mellitus was defined on the basis of blood glucose and plasma immunoreactive insulin levels measured before and after a standard oral load of glucose. The results were compared to those observed in a control group of healthy female subjects. Furthermore, thyroglobulin antibodies were titrated in most patients.
gonads devoid of follicles. The first 13 patients listed in Table 1 were considered as Turner's syndrome because of the short stature and associated somatic malformations. The 2 patients, K. J. and D. H., although without malformations, were classified as Turner's syndrome because of their short stature. The remaining 5 patients were diagnosed as pure gonadal dysgenesis syndromes. Chromosome analysis was made on leukocytes cultured according to the method of Moorhead et al.,4 and on cells from gonadal streak culture in 1 case. The urinary excretions of 17 keto- and 17 hydroxy-steroids before and during a metyrapone test were normal in all patients. Tomographies of the sella turcica did not reveal radiologic anomalies, except for Patient K. J., who had a Rathke pouch cyst removed at the age of 14. At that time, her stature was short and she had primary amenorrhea with no secondary sex characters; the pituitary functions were normal. The plasma growth hormone response to insulin-induced hypoglycemia was normal before and after the transsphenoidal resection of the Rathke MATERIAL/' AND METHODS pouch cyst. Following the operation, the The clinical and cytogenetical data of patient's height did not change and the the 18 patients with gonadal dysgenesis diagnosis of gonadal dysgenesis was fully are summarized in Table 1. The diagnosis documented at the age of 16. of gonadal dysgenesis was based on the The thyroid function was evaluated by clinical findings of primary amenorrhea measuring the serum concentration of with absence of secondary sex characters, PBI, T3 and T4, the thyroid uptake of on the laboratory report of high urinary radioiodine, and the levels of circulating excretion of gonadotropins, and on the thyroid antibodies. Thyroglobulin antihistologic observation of bilateral streak bodies were titrated by the latex precipitin Received June 12, 1972; revised August 16, 1972. test;5 because of the poor sensitivity of 1 of
VAN CAMPENHOUT ET AL.
TABLE 1. Clinical Data on Patients with Gonadal Dysgenesis Name
146 45 X
R. P. P. F.
143 45 X 134 45 X
146 45 X
135 45 X
45 X/46 XX
45 X/46 XX
K. J. D.H. G. S.
16 16 20
37.5 61 50
141 45 X/46 XY 140 45 X/46 XY 137 45 X/46 XX
162 46 XX 161 46 XX
D. V. O. R.
Family history of Diabetes
2 maternal aunts
-Webbing of the neck-cubitus valgus-shield shaped chest -Cubitus valgus -Cubitus valgus-high arched palate-horseshoe kidneyshort 4th metacarpal-malrotation of colon-shield shaped chest -Cubitus valgus-high arched palate-wide spaced nipples -Cubitus valgus-short 4th metacarpal-webbing of the neck-hemivertebrae-high arched palate-dorsal scoliosis -Cubitus valgus-webbing of the neck-high arched palate-Cubitus valgus -Cubitus valgus-mental retardation-high arched palate-short 4th metatarsal -Cubitus valgus-wide spaced nipples-genu valgummental retardation-webbing of the neck-uterus subseptus-short 4th metacarpal -Cubitus valgus-high arched palate
134 45 X/46 XX
45 Xiso X
-Cubitus valgus-lumbar scoliosis-genu valgum-high arched palate-atrial septal defect
Paternal aunt and cousin
46 XX(blood) 45 X/46 XX (gonadal streaks) 152 46 XX 160 46 XY
the method, a titer greater that 1: 40 was considered significantly elevated. The oral glucose tolerance test (GTT) was performed by the administration of 100 gm. of glucose in 200 m!. of water, in a 2-min. period, after an overnight fast. During the 3 days preceding the test, the
patients were given a high carbohydrate diet (200 gm./day). Blood was sampled through a catheter inserted in an antecubital vein. Plasma glucose was measured by the method of Brown 6 adapted to a Technicon autoanalyser, m which a cupric chelate is reduced by glucose in a cuprous
complex. The plasma insulin concentration was determined in duplicate by a double antibody radioimmunoassay technic 7 using crystalline bovine insulin as a standard (courtesy of Dr. J. Schlichtkrull, Novo Laboratories, Copenhagen). The glucose tolerance was defined as being abnormal if the plasma glucose value exceeded 170 mg./lOO ml. at the 1st hr., 140 mg./l00 ml. at the 2nd hr., and 120 mg./lOO ml. at the 3rd hr. of the test. Patient G. S., a known case of diabetes, with a history of diabetic coma, was not tested. On the basis of glucose tolerance, the patients with gonadal dysgenesis were divided into two groups, one with normal GTT and one with a diabetic GTT. Both groups happened to be similar in respect to the average age, height, weight, frequency of familial diabetes, and use of oral contraceptive hormones. Control studies of glucose tolerance and plasma insulin response were performed on 7 healthy female, medical staff volunteers, whose biologic and therapeutical characteristics corresponded to those of the gonadal dysgenesis groups (Table 2). Four patients with gonadal dysgenesis and abnormal glucose tolerance were further investigated by the use of a tolbutamide test; 1 gm. of sodium tolbutamide was rapidly injected into an antecubital
vein, and the decrease in blood glucose concentrations was followed for 3 hr. at various time intervals. RESULTS
Thyroid Autoimmunity. The tests of evaluation of the thyroid function were normal in all patients at the time of diagnosis of gonadal dysgenesis. However, Patient A.O. became hypothyroid 1 year after the initial investigation; she was treated with thyroid extracts. During the treatment, the titer of circulating thyroid antibodies increased from 1: 80-1: 5120. A biopsy of the enlarged thyroid showed the typical histologic pattern of Hashimoto's disease. A second patient, G.L., was euthyroid and had a titer of circulating thyroid antibodies greater that 1: 40. Thus, for the 16 tested patients, the incidence of thyroid autoimmunity is 12.5%. Frequency of Diabetes Mellitus. Among the 18 patients investigated in our study, 9 had a normal glucose tolerance (Fig. 1). One patient, G.S., had clinical diabetes, while 8 others had chemical diabetes as diagnosed by an abnormal tolerance to oral glucose (Fig. 2). This represents an overall incidence of diabetes of 50%. The nondiabetic and diabetic groups compare in respect to age, height, and weight (Table 3). The incidence of diabe-
TABLE 2. Clinical Data on Control Subjects Name
Family history of diabetes
Oral contraceptive therapy
T.L. A. D.
P. H. L. c. F.M. B. D. B. M.
22 23 21 22 21
50 45 51 45 53
165 165 172 165 170
168 (165~ 172)
Mother Maternal aunts
No Yes Yes No No Yes Yes
28 28 28
VAN CAMPENHOUT ET AL.
4 PLASMA GLUCOSE
mg., 100m'. 200
110 PLASMA IRI .,U/ml.
o H+---~---T--------r-------~ 120 110 min. 60 30
_ _ _ _ GLUCOSE
.. _______ 1 R I
FIG. 1. Evolution of plasma glucose and insulin levels during GTT in 9 nondiabetic patients with gonadal dysgenesis. Mean values with S.E.M.
diabetic group, the insulinogenic index (~ insulin above base line in ~U./ml./ ~ glucose above base line in mg./lOO ml.) was manifold higher in nondiabetic than in diabetic patients (Fig. 3). In a control group of 7 patients without gonadal dysgenesis, glucose tolerance was normal and plasma insulin levels were of the same order of magnitude as those observed in the two groups of gonadal dysgenesis (Table 5). Tolbutamide Test. In Fig. 4 are shown the results of four tolbutamide tests performed in patients with Turner's syndrome and diabetic GTT. According to the criteria of Unger and Madison,8 the blood sugar concentrations decreased m all cases in a normal fashion. DISCUSSION
tes in relatives was similar; 11 patients Thyroid Autoimmunity. Isolated cases were taking various oral contraceptives, of Hashimoto's disease and Turner's synfor periods of time longer than 3 months, drome 2, 9 or of hypothyroidism associated at the time of the glucose tolerance eval10 uation; 6 had a normal blood glucose curve with gonadal dysgenesis have been reported. and 5 were diabetic. The incidence of thyroid autoimmunity All diabetic patients but one presented in gonadal dysgenesis 1 0 -13 varies from the somatogonadal dysgenesis of the Tur12.5%-52%. Nevertheless, Irvine et al. l ner's syndrome (Table 4). The only case of pure gonadal dysgenesis with abnormal GLUCOSE GTT was characterized by a 45 X/46 XX PLASMA mg.' 100m •. mosaic blood karyotype. All 4 patients 200 with the chromosomal constitution 46 110 XX or 46 XY and pure gonadal dysgenesis had a normal glucose tolerance. PLASMA lRI pU/ml. The Plasma Insulin Response to Oral 160 Glucose Load. In patients with gonadal dysgenesis, the mean values of plasma T ...} ......... insulin concentrations following the ingestion of 100 gm. of glucose are shown in 120 ~ _10 Fig. 1 for the group with normal glucose tolerance and in Fig. 2 for the group with 100 , / abnormal tolerance. At any given time, o 0 110 min. 120 o 30 60 there was no significant difference between the insulin values in the two groups; _ _ _ _ GLUCOSE the insulin surfaces above base line values _______ .. I R I were calculated by planimetry and were FIG. 2. Evolution of plasma glucose and insulin comparable. Because of the much higher levels during GTT in 8 diabetic patients with golevels of blood glucose achieved in the nadal dysgenesis. Mean values with S.E.M.
TABLE 3. Comparison between Diabetic and Nondiabetic Patients with Gonadal Dysgenesis ~
24 (16-37) 20 (16-28) 20
Family history' of diabetes
148.1 (134-162) 143 (131-167) 137
45.3 (30-55.5) 51.7 (34-71.5) 50
2 (22%) 2 (25%) No
6 (66%) 4 (50%) Yes
TABLE 4. Incidence of Diabetes Mellitus in Patients with Turner's Syndrome and Pure Gonadal Dysgenesis Patients with pure gonadal dysgenesis
Patients with Turner's syndrome
Total 45 X
Patients total Patients with diabetes mellitus
45 X!46 XX 46 X iso X 45 X!46 XY
45 X!46 XX
~-----.----------~----------O+---~----~---------r--------~ o 30 60 120 180
_ _ NON DIABETIC ... ___ ... DIABETIC
FIG. 3. Insulinogenic index during GTT in 9 nondiabetic and 8 diabetic patients with gonadal dysgenesis. Mean values. The insulinogenic index is the ratio of plasma insulin increment above base line (I'U./ml.) to the corresponding plasma glucose increment above base line (mg./lOO mI.).
were unable to detect thyroid antibodies in a series of 43 cases of gonadal dysgenesis. With the latex thyroglobulin autoprecipitin test, which is poorly sensitive, we found an incidence of 12.5% of thyroid autoimmunity. Hill I. and Dallaire and Flynn 15 reported an incidence of 3% and 10%, respectively, among normal females 20-40
years old. Williams, Engel, and Forbes 11 and Valloton and Forbes,'3 using the tanned red cells agglutination technic, had a frequency of 52% and 51% of thyroid antibodies in gonadal dysgenesis, as opposed to 19% and 15.5% in normal controls. Several patients with gonadal dysgenesis developed primary myxedema or typical Hashimoto's disease. Furthermore, McHardy, Doniach, and Polani, 10 studying the response to thyroid-stimulating hormone in gonadal dysgenesis, demonstrated a low thyroid reserve in patients with significant titer of thyroid antibodies; these patients behave like those with chronic thyroiditis and may develop hypothyroidism. As in 1 of our cases, histologic evidence of thyroiditis was noted on thyroid biopsies,2, 9. 11, 12 and correlated well with immunologic and clinical studies, It seems, thus, that the incidence of Hashimoto's disease, primary myxedema, and thyroid antibodies in patients with gonadal dysgenesis is higher than in any control group. Furthermore, it has been shown that thyroid antibodies were more common in the parents of patients with gonadal dys-
VAN CAMPENHOUT ET AL.
TABLE 5. Blood Glucose and Plasma Insulin Levels during an Oral Glucose Tolerance Test in Control Subjects' 30'
Blood glucose 85.6 mg./100 m!. Plasma insulin /lU'/m!. 12.6 • Mean values
S.E.M. Number of subjects, 7.
FIG. 4. Evolution of plasma glucose levels in 4 diabetic patients with Turner's syndrome, following the rapid intravenous injection of 1 gm. of tolbutamide. Individual values.
genesis. 13 As postulated by Valloton and Forbes, 13 it might be that the high incidence of thyroid autoimmunity in gonadal dysgenesis is due to an X-linked mode of inheritance of the tendency to form that antibody. It is also possible, but hitherto not demonstrated, that this type of autoimmunity in the parents is the cause of chromosomal defects. On the other hand, both thyroid autoimmunity and chromosomal anomalies could simply result from the same physical, viral, or biochemical mutagenic influence. Further research is required to answer these questions. Diabetes Mellitus. The association of diabetes and gonadal dysgenesis has received attention during the past decade. A total of approximately 100 patients have been investigated by various authors, using different diagnostic procedures. The number of patients reported in each series
is small, the age range is wide, the chromosome aberrations and somatic malformations are multiform. Because of the already complex problems involved in the etiology and pathogenicity of common diabetes, a clear understanding of the diabetic syndrome in patients with gonadal dysgenesis is hardly conceivable at the present time. Our series of 18 patients comprises 13 Turner's and 5 pure gonadal dysgenesis syndromes. In contrast to studies of Engel and Forbes l2 and Menzinger, Fallucca, and Andreani, 16 who found a high frequency of overt diabetes mellitus both in patients with gonadal dysgenesis and in their relatives, a family history of diabetes was positive in only 3 of our 13 cases of Turner's syndrome and in 1 of the 5 subjects with pure gonadal dysgenesis; in no instance were the parents or siblings known diabetics. These data resemble those of Jackson et aI. 17 and Nielsen, Johansen, and Y de. 18 It appears therefore that, contrary to the suggestion of Lindsten et aI.,19 there is no direct relationship between the frequency of overt diabetes in relatives and the incidence of latent, chemical, or overt diabetes in patients with gonadal dysgenesis. However, the possibility remains that close relatives may have unrecognized forms of carbohydrate intolerance which, if searched for, may help to resolve the above mentioned disparity. There is general agreement, further sustained by the results of our study, that patients with gonadal dysgenesis have a higher than expected incidence of diabetes mellitus. The diagnosis of diabetes has generally been made on clinical observations, on the basis of an abnormal toler-
ance to a glucose load, or on the detection of anomalies of the insulin response to glucose. In this study, the patients with gonadal dysgenesis were divided into nondiabetics and diabetics according to their tolerance to an oral load of 100 gm. of glucose. All subjects in the diabetic group complied with the diagnostic criteria of the United States Public Health Service 20 or with those of the Joslin Clinic,21 even when having taken into account that plasma, rather than blood glucose levels, was measured. Among our 13 patients with Turner's syndrome, 1 had overt diabetes and 7 had a diabetic glucose tolerance curve, for an over-all incidence of diabetes of 61 %. This is the highest figure ever reported. All our patients were young, and ageing does not appear a primary factor in the genesis of diabetes. It has been proposed by Jackson et al. 17 that the loss of an X chromosome may render a patient susceptible to diabetes. In regard to this hypothesis, it is of interest that in our study all 4 patients with pure gonadal dysgenesis and a normal karyotype (46 XX or 46 XV) had a normal glucose tolerance as well as a normal plasma insulin curve. The data shown in Figs. 1 and 2 clearly indicate that circulating insulin is present in diabetics in similar amounts as in nondiabetics before and after the glucose load. This has also been observed in the common form of maturity-onset mild diabetes. As demonstrated by Kipnis,22 normal or even exaggerated insulin levels during GTT in mild diabetics need not be interpreted as indicating insulin resistance. Indeed, the insulinogenic stimulus represented by the blood glucose level is not equivalent in normal and diabetic subjects; thus, in respect to the high levels of blood glucose achieved in the diabetics, the plasma insulin level is quantitatively inadequate. This phenomenon can be estimated by the insulinogenic index,23 and it can be seen in Fig. 3 that the index is significantly
higher in nondiabetic than in diabetic patients with gonadal dysgenesis. Evidence for an insufficient secretion of insulin as a primary factor in the pathogenesis of diabetes in patients with gonadal dysgenesis is also afforded by the studies of Lindsten et al.: 19 in 14 subjects with normal intravenous glucose tolerance test, the insulin response was in most cases diminished when compared to that of control subjects with similar blood glucose levels. Qualitative defects in the insulin response to plasma glucose were also found among our 8 tested diabetics: 4 had indeed a delayed rise of plasma insulin concentration, whereby the peak value occurred at the 2nd or 3rd hr. of the GTT. Four patients with a diabetic GTT were tested with intravenous sodium tolbutamide; the hypoglycemic response was rapid and important. There is ample evidence that tolbutamide increases insulin secretion by mechanisms different from those concerned with glucose; thus, in the mild diabetes of the Turner's syndrome, the capacity of the B-cell to respond to tolbutamide may be retained while the response to glucose is decreased. If confirmed, these results indicate that tolbutamide may prove a valuable therapeutical tool for the management of chemical diabetes in patients with gonadal dysgenesis, where the weight is usually low and diets with restricted calories would seem inappropriate. Factors other than insulin may also play a role in the pathogenesis of diabetes. Occasionally, high levels of plasma growth hormone have been reported in the fasting state or, unexpectedly, during the early phase of a GTT. 19 It is possible that anomalies of the growth hormone secretion may account for the true hyper-response of insulin which has been observed in few cases of gonadal dysgenesis with normal glucose tolerance,17 and that tissue insulin resistance may be associated to some
VAN CAMPENHOUT ET AL.
extent with low insulin secretion in our cases of mild diabetes. Among the patients with gonadal dysgenesis and normal glucose tolerance, it is likely that some will develop diabetes. At first glance the insulin response to glucose in this group of patients seems low when compared to that reported by many authors in healthy nonobese subjects of both sexes and various ages. However, control studies in 7 female volunteers of the same age, body conformation, diabetic heredity, and history of oral contraceptives intake displayed the same type of low plasma insulin profile. This could be due to the unusual type of population tested, and these results emphasize the need of carefully designed control studies in the evaluation of patients with gonadal dysgenesis. SUMMARY
Thirteen patients with Turner's syndrome and 5 with pure gonadal dysgenesis were investigated. Adrenal-pituitary function was normal in all subjects; thyroglobulin antibodies were detected in two. A standard oral glucose tolerance test was normal in 9 patients; 8 patients had a diabetic curve and 1 had overt diabetes. Among the patients with Turner's syndrome, the frequency of diabetes was 61 %; in this group, the mean age was 20 years and the frequency of overt diabetes in close relatives was negligible. In the 5 patients with pure gonadal dysgenesis, 1 had mild diabetes and 4 had a normal glucose tolerance as well as a normal karyotype. The plasma insulin response to glucose was quantitatively comparable in the diabetic and nondiabetic groups of patients; in the diabetics, the insulin peak was either delayed or brisk. No anomalies of the insulin response to glucose could be detected in nondiabetic patients with gonadal dysgenesis by comparison with the results obtained in a carefully designed control study.
Acknowledgment. The authors wish to acknowledge the assistance of Dr. Y. Ainmelk. REFERENCES 1. IRVINE, W. J., CHAN, M. M., SCARTH, L., KOLB, F. 0., HARTOG, M., BAYLISS, R. I. S., AND DRURY, M. 1. Immunological aspects of premature ovarian failure associated with idiopathic Addison's disease. Lancet II:883, 1968. 2. SPARKES, R S., AND MOTULSKY, A. G. Hashimoto's disease in Turner's syndrome with isochromosome X. Lancet 1:947, 1963. 3. FORBES, A. P., AND ENGEL, E. The high incidence of diabetes mellitus in 41 patients with gonadal dysgenesis, and their close relatives. Metabolism 12:428, 1963. 4. MOORHEAD, P. S., NOWELL, P. C., MELLMAN, N. J., BATTIPS, D. M., AND HUNGERFORD, D. A. Chromosome preparation of leucocytes cultured from human peripheral blood. Exp Cell Res 20:613, 1960. 5. ANDERSON, J. R, BUCHANAN, W. W., GOUDIE, R B., AND GRAY, K. G. Diagnostic tests for thyroid antibodies: a comparison of the precipitin and latex-fixation tests. J Clin Path 15:462, 1962. 6. BROWN, M. E. Ultra-micro sugar determinations using 2, 9-dimethyl-1, 10-phenanthroline hydrochloride (neocuproine). Diabetes 10:60, 1961. 7. MORGAN, C. R, AND LAZAROW, A. Immuno-assay of insulin: two antibody system. Plasma insulin levels of normal, sub diabetic and diabetic rat. Diabetes 12:115, 1963. 8. UNGER, R H., AND MADISON, L. L. A new diagnostic procedure for mild diabetes mellitus. Evaluation of an intravenous tolbutamide response test. Diabetes 7:455, 1958. 9. GRUMBACH, M. M., AND MORISHIMA, A. J. X Chromosome abnormalities in gonadal dysgenesis; DNA replication of structurally abnormal chromosome; relation to thyroid disease. J Pediat 65:1087, 1964. 10. McHARDY, S., DONIACH, D., AND POLANI, P. Thyroid function in Turner's syndrome and allied conditions. Lancet II: 1161, 1970. 11. WILLIAMS, E. D., ENGEL, E., AND FORBES, A. P. Thyroiditis and gonadal dysgenesis. New Eng J Med 270:805, 1964. 12. ENGEL, E., AND FORBES, A. P. Cytogenetic and clinical findings in 48 patients with congenitally defective or absent ovaries. Medicine 44:135, 1965. 13. VALLOTON, M. B., AND FORBES, A. P. Autoimmunity in gonadal dysgenesis and Klinefelter's syndrome. Lancet 1:648, 1967. 14. HILL, O. W. Thyroglobulin antibodies in 1297
patients without thyroid disease. Brit Med J I: 1798, 1961. 15.DALLAIRE, L., AND FLYNN, D. K., Auto-immunite de la thyroide et aberrations chromosomiques. Un Med Canada 96:1371, 1967. 16. MENZINGER, G., FALLUCCA, F., AND ANDREANI, D. Gonadal dysgenesis and diabetes. Lancet 1:1269, 1966. 17. JACKSON, M. D., BUCHANAN, K. D., McKIDDIE, M. T., AND PRENTICE, C. R M. Carbohydrate metabolism and pituitary function in gonadal dysgenesis (Turner's syndrome). J Endocr 34: 289,1966. 18. NIELSEN, J., JOHANSEN, K., AND YDE, H. The frequency of diabetes mellitus in patients with Turner's syndrome and pure gonadal dysgenesis. Acta Endocr (Kobenhaun) 62:251, 1969 .
19. LINDSTEN, J., CERASI, E., LUIT, R, AND HULTQUIST, G. The occurrence of abnormal insulin and growth hormone (HGH) responses to sustained hyperglycemia in a disease with sex chromosome aberrations (Turner's syndrome). Acta Endocr (Kobenhaun) 56:107, 1967. 20. REMAIN, Q. R, AND WILKERSON, H. L. C. The efficiency of screening test for diabetes. J Chronic Dis 13:6, 1961. 21. MARBLE, A. Joslin's Diabetes Mellitus (ed. 11). Lea and Febiger, Philadelphia, 1971, p 191. 22. KIPNIS, D. M. "Diabetes." In Excerpta Medica Foundation, 1CS 172. Amsterdam, 1969, p. 251. 23. SELTZER, H. S., AND SMITH, W. L. Plasma insulin activity after glucose. An index of insulinogenic reserve in normal and diabetic man. Diabetes 8:417, 1959.