American Journal of Medical Genetics 36:148-154 (1990)

-~

Autoimmune Thyroiditis Associated With Mild “Subclinical” Hypothyroidism in Adults With Down Syndrome: A Comparison of Patients With and Without Manifestations of Alzheimer Disease Maire E. Percy, Arthur J. Dalton, Vjerica D. Markovic, Donald R. Crapper McLachlan, Edith Gera, Jocelyn T. Hummel, Ann C.M. Rusk, Martin J. Somerville, David F. Andrews, and Paul G. Walfish Departments of Obstetrics and Gynaecology (M.E.P., J.T.H., A.C.M.R., M.J.S.), Physiology (D.R.C.M., M . J . S . ) , Medicine (D.R.C.M., P.G.W.), Statistics and Preventive Medicine and Biostatistics (D.F.A.), University of Toronto; Surrey Place Centre (A.J.D., V.D.M., M.E.P.) and Endocrine Division and Research Institute, Mount Sinai Hospital (E.G., P.G.W.), Toronto, Canada.

Serum tests of thyroid function were compared in Down syndrome (DS) patients with and without manifestations of Alzheimer disease (AD). Relative to control individuals, DS patients had, overall, lower mean total T4 ( P = 0.070) and T3 ( P = 0.015), higher T3U ( P = 0.013) and TSH ( P = 0.020), no difference in free T4, and higher thyroid antithyroglobulin (ATA)( P = 0.033) and antimicrosomal autoantibody (AMA) titres ( P = 0.0097). Similar trends were apparent in DS males and females, and in DS patients off all drugs. In an analysis of case/control pairs with corrections for age and sex, DS patients with AD manifestations (n = 9) had significantly lower T3 ( P = 0.029) and higher AMA ( P = 0.043) than paired control individuals, whereas DS patients without AD manifestations (n = 20) had significantly lower T3 ( P = 0.013) but higher ATA ( P = 0.0065). T3 was significantly lower in the DS patients with AD manifestations than in the unaffected ( P = 0.0013). These data suggest that autoimmune thyroiditis associated with a mild “subclinical” form of hypothyroidism is common in adult DS patients and more pronounced in patients with AD manifestations than in those without. This

“subclinical” hypothyroidism may contribute to cognitive deficits in ageing DS patients.

KEY WORDS: antithyroglobulin antibody, antimicrosomal, thyroid stimulating hormone, serum triiodothyronine

INTRODUCTION A high proportion of Down syndrome (DS) patients who come to autopsy after age 40 show brain changes that closely resemble those characteristic of Alzheimer disease (AD). Moreover, longitudinal studies have indicated that a high proportion also develop clinical symptoms resembling AD [Dalton et al., 1974; Crapper et al., 1975; Wisniewski et al., 1985; Oliver and Holland, 19861. Thus, studies of ageing DS patients may provide clues about factors contributing to the development of AD. A history of thyroid disease has been identified a s a possible risk factor for AD in the general population [Heyman et al., 1983; Mortimer, 19891. Although an increase in the frequency of thyroid disease among patients with DS has been described by many different investigators [Aarskog, 1969; Sare et al., 1978; de Lob0 et al., 1980; Coleman and Abbassi, 1984; Fort et al., 1984; Vladutiu et al., 1984; Cutler et al., 19861, abnormalities of the thyroid function have not been evaluated in ageing DS patients or in the Alzheimer process in this population. For a n investigation of the red cell oxygenReceived for publication September 15, 1988; revision received metabolizing enzymes in DS, our group identified subOctober 26, 1989. sets of severely to profoundly mentally retarded DS paAddress reprint requests to Maire E. Percy, Ph.D., a t the present tients with and without clinical manifestations of AD address of Neurogenetics Laboratory, Room 423, Surrey Place Cen[Percy et al., in press]. In the present study, we have tre, 2 Surrey Place, Toronto, Ontario, M5S 2C2, Canada. Arthur J. Dalton’s present address is The New York State Insti- compared serum tests for thyroid function in these two tute for Basic Research in Developmental Disabilities, Staten Is- groups of DS patients and in DS patients overall relative to paired control individuals. We have also measured land, NY.

0 1990 Wiley-Liss, Inc.

Alzheimer Disease in Down Syndrome

the levels of thyroid antithyroglobulin antibody (ATA) and antimicrosomal antibody (AMA) in the study participants since “autoimmune thyroiditis” has been observed in previous studies of thyroid function in DS [Aarskog, 19691. The objectives of the study were to determine if there was a consistent type of abnormality in thyroid function in our population of DS patients overall and if there were any differences in thyroid function between DS patients with and without clinical manifestations of AD.

MATERIALS AND METHODS Patients The study was carried out in accordance with a University of Toronto approved protocol. Forty-six individuals with DS from group homes or institutions in Ontario, ranging in age from 31 to 70 years (average age, 45.3 2 11.2 years) and 36 healthy control individuals of the same age range were included in the present study. All but five DS patients, who were carriers of hepatitis B (two females and three males), were karyotyped. With one exception [Percy et al., in press], they were nonmosaic trisomy 21. The hepatitis carriers were not excluded from the study. Twenty of the DS patients had no clinical manifestations of AD, whereas nine were classified as having AD manifestations. Three of the latter (two females and one male) were hepatitis carriers. The patients comprising these two groups were identical to those participating in our previous study, and their characteristics have been described IPercy e t al., in press]. Each of these 29 patients had a “paired” control individual, matched for sex, age ( i2 years) and time of blood sampling ( 2 1 hour). The DS patients without AD manifestations were significantly younger (46.0 i 10.4 years) than those with AD manifestations (54.7 i 8.3 years). Biochemical testing was carried out independently without knowledge of the clinical status of the patients. Serum Tests for Thyroid Function and Autoantibodies to Thyroid The serum thyroxine (T4)level was determined by the double radioimmunoassay of Chopra and Irvine 119721, the serum triiodothyronine (T3) level by the radioimmunoassay method of Chopra et al. 119721, and the serum TSH level by the radioimmunoassay method of Pate1 et al. [1971] usinga human TSH assay kit supplied by the National Institute of Arthritis, Metabolism and Digestive Diseases (NIAMDD), Bethesda, Maryland, and TSH international reference standard 68138, supplied by the World Health Organization (WHO)International Laboratories of Biological Standards, Holly Hill, England. The T3 uptake (T3U) test was performed using a standard silica-talc tablet and measuring the percentage uptake of the added T3 tracer; the uptake was then expressed a s a percentage of ‘‘normal,’’ which in this case was a control sample, run with each assay, of pooled serum from persons with normal thyroid function. Measurements of serum free thyroxine (FT4) and free triiodothyronine (FT3) were performed by standard radioisotopic methods [Walfish et al., 19821. The serum

149

titres of ATA and AMA were determined by quantitative hemagglutination using Sera-Tek kits (Ames Company, Division of Miles Laboratories, Elkhart, IN). Blood samples were collected by venipuncture over a 21/2 year period. Serum samples were stored a t - 80°C in small aliquots and assayed after a single thaw. One sample from each individual was analyzed.

Statistical Analyses Means of the serum thyroid function tests were compared using Student’s two-tailed t test. Because of their skewed distribution, all data (except ages) were logarithmically transformed. To eliminate variation due to reagents, season, and other factors, the comparison of DS patients with and without AD manifestations was based on data from the 29 caseicontrol pairs. Differences between pairs of DS patients and control individuals were thought possibly to depend on the common age and sex of the pair. To study these relations and to allow for such effects in assessing the differences between the groups of DS patients with and without manifestations of AD, as in our previous study [Percy et al., in press], multiple linear regression methods were also used [Snedecor and Cochran, 19801. A linear model for case/ control differences as a function of age, sex, and manifestation of AD was fitted and the significance of the last term assessed. The results of 7 tests are presented. No formal allowance for multiplicity of tests has been made. The significance of the prevalence of abnormal test results in caselcontrol comparisons was determined using standard contingency table analysis. RESULTS The comparison of means using data on all available subjects is given in Table I, and on DS patients with and without AD manifestations relative to the paired control individuals in Table 11.The prevalences of elevated TSH, ATA, AMA, and clinical hyper- or hypothyroidism in the caseicontrol pairs are shown in Table 111. Table IV summarizes the caselcontrol differences corrected for age and sex. Individual T4, T3, TSH, and AMA results in the DS patients with and without AD manifestations and paired control individuals are shown in Fig. 1. The data in Table I indicate that in DS, overall, there is a reduction in total T4 and T3 compared with controls (P = 0.07 and 0.015, respectively) and a n elevation in T3U (P = 0.034). However, if free T4 and T3 indices are computed by multiplying the T4 and T3 by the T3U, then there is no difference between the control and patient groups. Although there is a statistically significant difference overall between patients and control individuals in serum TSH (P = 0.0201, the difference is not large, the effect being more pronounced in males (P = 0.025) than in females (P = 0.33). This suggests t h a t the pituitary is not sensing the hypothyroidism (i.e., the low circulating levels of free T4 and T3) as in primary hypothyroidism. These observations are supported by the free T4 values measured by the rapid nondialysis technique, which are not different in the patients and control individuals. Nevertheless, although there is a higher T3 uptake and a lower T4 with no difference in the free T4 values, the higher than normal

Percy et al.

150

TABLE I. Comparisons of Means of Thyroid Function Test Results in all Available Adult Down Syndrome Patients and Control Individuals Using Student's Two-Tailed t Test* -

~~

Males Patients Controls P values

~

No. 24 15

T4

-

~

7.4::: 7.7Ti.Z 0.57

~~~

T3 T3~U TSH -~ .~~~

1.6;: 1.9'; 0.056

34.7';; 33.4':; 0.27

3.3'::: 1.91;; 0.015

1.2 :2 1.2':; 0.47

0.015

35.0';; 31.51g.i 0.036 34.8::: 32.4';; 0.013

2.2-?; 1.6';: 0.33 2.8':: 1.7:,!,: 0.020

1. 1. 0 1.2:;; 1.2.:; 0.53

P value

P value

0.070

FT4-

ATA

~~

~~~

~

AMA

~

..~

Age-

'2:

::

1:32.841:zz 1:13.8- 'if 0.041

44.8-cl0.9 46.8211.8 0.59

';,A

1: 50.4-'""" 40 1 1:21.8I 0.067 1:40.31';:: 1: 18.0 3129 45 0.0097

45.7 -t 11.6 47.5i10.6 0.60 45.3211.2 47.2 2 11.0 0.43

1:13.3 1:lO.O 0.081

A

~~~

1: 16.77 1:12.4. E y 0.11 1:14.81'+: 1:11.41 2; 0.033

'

*The thyroid data, but not ages, were logarithmically transformed for the comparisons (refer to Materials and Methods). The means -t standard m +-antilog SD) - antilog (m)} deviations for age, and antilogs of the logarithmic m z SD values for the other parameters[ i.e. antilog (m) +{antilog -{antilog ((m) (m SD))

3

are shown

mean TSH suggests that there may be a subtle form of mild hypothyroidism in the DS patients. What is more impressive are the increased antibody titres in the DS patients. The AMA titres are higher in both male and female DS patients in spite of no significant age differences between controls and patients. Moreover, there are no differences between males and females within the control or DS group in their antibody titres (analyses not shown). When the comparisons were limited to the 20 DS patients with no AD manifestations, the 9 DS patients with AD manifestations, and the paired control individuals, the trends that were observed in the larger groups were again apparent (Fig. 1, Table 11). Although TSH was not significantly elevated in these DS patients overall, more DS patients than control individuals had serum TSH values greater than 4.0, and therefore above the upper limits ofnormal (Fig. 1).Relative to the paired control individuals, the mean TSH was higher in the DS patients with AD manifestations than in those without, although not significantly (Table 11).Overall, the ATA and AMA titres were significantly elevated relative to the control individuals (P = 0.0021 and 0.031, respec-

tively), although none of the ATA titres in the patients exceeded 1: 80, and the AMA titres (with one exception) were between 1: 80 and 1: 640. ATA was significantly elevated in the DS patients without AD manifestations (P = 0.0055), whereas AMA was significantly elevated in those with AD manifestations (P = 0.0048). The prevalence of abnormally high AMA titres was significantly greater in DS patients overall than in the paired controls (P = 0.036); that of elevated TSH values differed only marginally (P = 0.070) (Table 111). The prevalence of clinical hyper- or hypothyroidism was slightly, but not significantly, greater in DS patients than in control individuals. There were no significant differences in prevalences of abnormal test results in comparisons of DS patients with and without AD manifestations, but this finding in part reflects the small sample sizes of the groups as well as the mildness of the abnormalities. The application of multiple linear regression in a n analysis of these 29 caseicontrol pairs corroborated the above conclusions and indicated that sex effects were not different in DS patients and control individuals, but that the dependence of T3 and AMA titre upon age

TABLE 11. Comparisons of Means of Thyroid Function Test Results in Down Syndrome (DS) Patients With and Without AD Manifestations and Matched Control (C Individuals Using Student's Two-Tailed t Test* T4 T3 T3 U ~~No. DS patients without AD manifestations (m/f= 1119) DS 20 6.4;:-i 1.6+:2 33.6 C 20 8.31:: 1.9::; 32.4 P value 0.047 0.050 0.3 DS patients with AD manifestations (mif= 415) DS 9 7.2?';,: 1.5:;: 33.6':; C 9 8.8::; 2.0lhy 31.5Ig: P value 0.25 0.10 0.41 DS patients overall (mif = 15/14) DS 29 6.7':; 1.6':; 33.6I:.: C 29 8.4::; 1.9::; 32.0":; P value 0.020 0.0093 0.19

TSH ~~

~

FT4 ~~

AM* -~ -

ATA

-

2.11:; 2.2::; 0.77

1.1::; 1.2';; 0.61

1: 1:

3.4';; 2.31;; 0.30

1.3':; 1.2::; 0.82

1:18.2 1:12.2 0.20

2.4';; 2.2';); 0.74

1.21!.; 1.2';; 0.79

~

Age -

0.21

0.92

:i

1:64.3'%: 1 : 1 7 . 1 ??: 0.048

5 4 . 7 t 8.3 5 3 . 7 1 8.9 0.80

1:18.1+':; 1:11.21 iE 0.0021

1:47.7 1:20.91 7:' 0.031

48.7i10.5 48.7110.6 0.84

*Refer to Table I legend. None of the participants were taking thyroid medications.The control individual for each DS patient was matched for sex, age ( z 2 years), and time of blood sampling ( 2 1 hour). Individual T4, T3, TSH, and AMA results are given in Fig. 1.

Alzheimer Disease in Down Syndrome

151

TABLE 111. Prevalence of Abnormal Test Results for Thyroid Function in Paired Down Syndrome (DS) Patients and Control (C) Individuals* Elevated TSH“

No. 29 29

.~

DS C P value

(%I 7129 = 24.1 2129 = 6.9 0.070

Elevated ATA~ (%a)

2129 = 6.9 0129 = 0.0 0.150

Elevated AMA’

(%I 11129=37.9 4129 = 13.8 0.036

Clinical hyperthyroidism ~~~

(%I

Clinical hypothyroidism

~~~~

( 7 C

~

2129 = 6.9 1129= 3.5 >0.5

~

3129 = 10.0 2129 = 6.9 >0.5

* The groups being compared are the same as in Table 11. >4.0 ~ U l m l . > 1 : 40. ‘ > 1 : 80. a

differed significantly in these two groups (analysis not shown). When corrections were made for the effects of age and sex (though minor), relative to the matched control individuals the DS patients without AD manifestations had lower T3 (P = 0.0013) and higher ATA titres (P = 0.0065), whereas the DS patients with AD manifestations had lower T3 (P = 0.029) but higher AMA titres (P = 0.043) (Table IV). (It should be noted that the ATA titres were the same in the DS patients with and without AD manifestations, whereas the AMA titre was higher in those with AD manifestations than in those without.) Also, as is evident from Table IV, the DS patients with AD manifestations had significantly lower T3 than those without (P = 0.0013).

DISCUSSION Our study suggests that autoimmune thyroiditis associated with a mild ‘‘subclinical” form of hypothyroidism is common in adult DS patients. Although our sample sizes were small, this was more pronounced in DS patients with manifestations of AD than in those without. Clinical hypothyroidism is known to mask as AD in the general population. Accordingly, we are proposing that the “subclinical” hypothyroidism that we have characterized may contribute to the cognitive deficits that are characteristic of ageing DS patients, and possibly also to the mental retardation that is characteristic of DS in general. According to standard criteria, the proportion of DS patients requiring treatment for thyroid disorders was only slightly higher than that of the control group (Table 111).However,whether persons with such “subclinical” hypothyroidism should be treated or

not is an important matter requiring further investigation [Cooper, 19871. Lob0 et al. [1980] have previously described the benefits of treating teenage DS patients with gross hypothyroidism. As judged from the changes in TSH and AMA, the degree of severity of thyroid hypofunction is milder and the degree of rise in antibody titres is less severe than in some of the more affected persons reported in general population surveys [Walfish and Gryfe, 1982; Sawin et al., 1985; Cooper, 1987; Rosenthal et al., 1987; Schroffner, 19871.In our study, 24% of the DS patients and 7% of the matched control individuals had TSH values greater than 4.0 (the upper limit of normal), only 1129 (3.5%) of the DS patients having a TSH value above 10 pUiml. Thirty-eight percent of the DS patients had elevated AMA titres; only one patient’s titre exceeded 1:1,280. In a recent survey of 258 individuals over 60, Rosenthal et al. [19871 found that 13.5% of the population had TSH values greater than 4.0,4% having TSH values above 10 FUiml, whereas 14 ofthe 258 (5.5%)had AMA titres exceeding 1:1,280 (Fig. 1, Table 111). The abnormalities in thyroid function in DS thus resemble those in normal “ageing,” but they are more prevalent, and less severe. The frequency of thyroid dysfunction in DS is thought to be lower in children than in adults, although a rigorous comparison has not yet been made [Cutler et al., 19861. These authors have stressed the need for a longitudinal screening of thyroid function in DS. Thyroid autopsy in DS would help to clarify the nature of the abnormalities that our serum tests have identified. Of considerable interest is our finding that DS males

TABLE IV. Paired Analysis of Thyroid Function Test Results Using Multiple Linear Regression, Case Versus Control, Correcting for Ageisex” ~~

DS patients Coefficient SE P value DS patients Coefficient SE P value DS patients Coefficient

SE P value

T4 T3 T3U TSH FT4 without AD manifestations vs. DS patients with AD manifestations -0.0355 0.0129 0.00591 0.865 0.0146 0.212 0.104 0.0649 0.518 0.172 0.081 0.0013 0.29 0.46 0.93 without AD manifestations vs. matched control individuals -0.209 - 0.213 0.0406 - 0.218 - 0.0264 0.115 0.059 0.0377 0.290 0.0941 0.081 0.0013 0.29 0.46 0.78 with AD manifestations vs. matched control individuals - 0.244 - 0.200 0.0465 0.647 - 0.0118 0.168 0.0865 0.0497 0.401 0.137 0.16 0.029 0.36 0.12 0.93

ATA

AMA ~~

- 0.177

- 1.541

0.324 0.0048

1.046 0.15

0.521 0.176 0.0065

0.228 0.568 0.69

0.404 0.257 0.13

1.77 0.830 0.043

Percy et al.

152

T4 5

0

.

18

31

12

;i 0 ,

2

0

,

DS C -AD

I,

,

DS C +AD

DS

C

-AD

1 12560

-AD

+AD

DS C +AD

AMA

! -AD

+AD

Fig. 1. Serum T4, T3, TSH, and AMA in Down syndrome (DS) patients with and without Alzheimer disease ( + AD/ - AD) manifestations and paired control (C) individuals. The arrows denote the means of the logarithmic distributions (refer to Table 11).The numbers on the ventricle axis represent the untransformed units of Tq,T3, and TSH, and titre of AMA in individual cases and controls.

and females had higher than normal TSH and higher antibody titres, since in the general population the prevalence of autoimmune thyroid disease is higher in females [Rosenthal et al., 19871. This finding suggests that the hypothyroidism and thyroid immunity in DS may be causally related. Of the 20 DS patients without AD manifestations, 3 of 7 with elevated AMA were female; ofthe 9 DS patients with AD manifestations, 4 of 5 with elevated AMA were female. These data raise the possibility that elevated AMA titres and the female sex may be associated in the category of DS patients with AD manifestations. A possible association of thyroid disease and AD has previously been suggested for affected women in families with senile dementia of the Alzheimer type [Heyman et al., 19831. The changes that were observed in the DS patients overall were not biased by the fact that some of the participants were taking some form of medication; when these individuals were excluded from the comparisons, very similar results were obtained (data not shown). (Medications taken by the study participants have been described elsewhere [Percy et al., in press].) The trends observed in the DS patients overall, and in the smaller group of DS patients without AD manifestations, were also not biased by the fact that five of the DS study participants were carriers of hepatitis, a state known to be associated with higher T4 and lower T3 uptake, and autoimmune thyroid disease [Schussler et al., 19781. A possibility of bias cannot be excluded in the case of the

DS patients with AD manifestations, as three of these nine (including two women) were hepatitis carriers; all had AMA autoantibodies, one also having a very high T4 level (Fig. 1). A fundamental question is why thyroid autoantibody titres are higher in DS. A spectrum of autoimmune abnormalities has been observed in DS, and it has been suggested that this phenomenon may be related to the fact that trisomy 21 patients have a triple gene dosage for the a, p-interferon receptor, which might interfere with the maturation of monocytes [Walford, 19821. Recent surveys suggest that there may be a n association between certain HLA antigens and autoimmune thyroiditis [reviewed by Farid and Thompson, 19861. Reports of possible restrictions in HLA heterogeneity in DS may be relevant to this issue LMayr et al., 19851. Virus infections, endocrinologic disturbances, or chromosomal abnormalities are other possible causes [Fialkow, 19701. As the triple gene dosage for superoxide dismutase-1 is believed to cause excessive production of hydrogen peroxide in DS [Percy et al., in press], we speculate that the elevated AMA and ATA production in this population may be a consequence of the latter process, since peroxide is a substrate for the iodide peroxidase. (This hypothesis could be tested in mice transgenic for the superoxide dismutase-1 gene.) In the general population, the presence of antithyroid antibodies can be a reflection of a n autoimmune syndrome in which autoantibodies with other specificities are also produced [Rousset et al., 19831. Of possible relevance is the finding by Tanaka and Miyatake 119831 of anti-acetylcholine antibodies in aged individuals and in DS patients. F'urther investigations of the immune and autoimmune status in DS and AD are clearly warranted, as there is evidence for depressed cellular immune responses to mitogenic stimulation in both DS and AD in the general population, suggesting some type of immune deficiency in patients with these two disorders [Singh et al., 19871. That immune responses differ in these two disorders, however, is suggested by the presence of anti-brain antibodies in AD patients in the general population but not in DS patients LSingh and Fudenberg, 19861. It has been established that thyroid AMA reacts with the peroxidase that catalyzes iodination of thyroglobulin in the presence of hydrogen peroxide LCzarnecka et al., 1985; Portmann et al., 1985; Ruf et al., 19871. Thus, antibody to this enzyme might account for the lowered T4 and T3 levels apparent in DS. T3 is the metabolically active thyroid hormone. It binds to receptors in the nucleus that are bound to chromatin and increases the expression of genetic information by increasing the synthesis of specific mRNAs. Because T3 levels are reduced in DS, a n expected consequence would be reduced brain mRNA levels. In the general population, AD is associated with altered chromatin structure, which accounts, in part, for a nonrandom reduction in mRNA [McLachlan et al., 19881. A possible consequence of altered chromatin structure is that the DNA consensus sequence recognized by the thyroid receptor may be heterochromatized for some genes. If heterochromatization is similar in DS, a form of

Alzheimer Disease in Down Syndrome

end-organ myxoedema might result, compounding the effect of the minimal circulating T3 level. We noted that although the T3 level was significantly reduced in DS, the TSH level was not elevated in proportion, as though the pituitary was not sensing the hypothyroidism. Thyroid hormone is important in the regulation of synthesis and secretion of TSH in the anterior pituitary. In aged male and female rats, lower basal serum thyroid hormone levels and an impaired TSH response to TRH have been demonstrated. These age-related effects are due not only to a reduction in thyroid gland function but also to a concomitant hyporesponsiveness of the aged rat pituitary thyrotroph to negative feedback and TRH stimuli LChen and Walfish, 1977,1979;Khalil and Walfish, 1983; Pekary et al., 19871. In the rat, i t is thought that hypothyroidism may induce both transcription and translation of the hypothalamic thyrotropin releasing prohormone (pro-TRH) in the paraventricular nucleus. Because epinephrine and norepinephrine have effects on TRH secretion, resulting in the stimulation of both TSH and TRH release in the rat, it is possible that the feedback control of thyroid hormone on the biosynthesis of TRH may be mediated, a t least in part, indirectly through central catecholamine pathways [Seyerson et al., 19871. As the cholinergic pathway is known to be deficient in adults with DS [Yates et al., 19801,as well as in AD [Bowen, 19881, it is possible that the stimulation of TSH release is affected.

ACKNOWLEDGMENTS This work was supported in part by the Ontario Ministry of Health (grant #01099), the Ontario Mental Health Foundation (grant #934-85/87], the Ontario Ministry of Community and Social Services Lottery Grants Program, and the Clinical Research Support Unit, Department of Preventive Medicine and Biostatistics, University of Toronto. We are very grateful to the participants, their families, and their care-givers for their cooperation in this study, to Medical Diagnostic Laboratories (MDS) for assistance in the collection of blood samples, and to Sharon Bauer for assistance with typing of the manuscript. M.E.P. was the recipient of a National Health Research Scholar Award (Canada Health and Welfare). REFERENCES Aarskog D (1969): Autoimmune thyroid disease in children with mongolism. Arch Dis Child 44:454-460. Bowen DM (1988): Neurotransmitters in Alzheimer’s disease. Age 11:104-110. Chen H J , Walfish PG (1977):Effects of age and ovarian function on the pituitary-thyroid system in female rats. J Endocrinol 78:225-232. Chen HJ, Walfish (1979): Effects of age and testicular function on the pituitary-thyroid system in male rats. J Endocrinol 82:53-59. Chopra IJ, Irvine J 11972):A radioimmunoassay for T4 measurement of thyroxine in unextracted serum. J Clin Endocrinol Metab 34:938-947. Chopra IJ, Ho RS, Lam R (1972): An improved radioimmunoassay of triiodothyronine in serum: Its application to clinical and physiological studies. J Lab Clin Med 80:729-739. Coleman M, Abbassi V (1984): Down’s syndrome and hypothyroidism: Coincidence or consequence? Lancet 1569. Cooper DS (1987): Editorial: Subclinical hypothyroidism. JAMA 258:246-247.

153

Crapper DR, Dalton AJ, Skopitz M, Scott JW, Hachinski V (1975): Alzheimer degeneration in Down syndrome: Electrophysiologic alterations and histopathologic findings. Arch Neurol 32618-623. Cutler AT, Benezra-Abeiter R, Brink SJ (1986): Thyroid function in young children with Down syndrome. Am J Dis Child 140:479-483. Czarnecka B, Ruf J , Ferrand M, Carayon P (1985):Antigenic relation between thyroid peroxidase and the microsomal antigen implicated in auto-immune diseases of the thyroid. CR Acad Sci (111) 300:577-580. Dalton AJ, Crapper DR, Schlotterer GR (1974): Alzheimer’s disease in Down syndrome: Visual retention deficits. Cirtex 10:366-377. Farid HR, Thompson C (1986): HLA and autoimmune endocrine disease 1985. Mol Biol Med 3235-97. Fialkow PL f 1970):Thyroid autoimmunity and Down’s syndrome. Ann NY Acad Sci 171:500-511. Fort P, Lifshitz F, Bellisario R, Davis J, Lanes R, Pugliese M, Richman R, Post EM, David R (1984): Abnormalities of thyroid function in infants with Down syndrome. J Pediatr 104545-549. Heyman A, Wilkinson WE, Hurwitz BJ, Schmechel D, Sigmon AH, Weinberg T, Helms MJ, Swift M (1983): Alzheimer’s disease: Genetic aspects and associated clinical disorders. Ann Neurol 14:507-515. Khalil A, Walfish PG (1983):The effects of aging on several stimuli to the male rat hypothalamic-pituitary-thyroid system. In Ui N, Torizuka K, Negataki S, Miyai K (eds): “Current Problems in Thyroid Research.” Amsterdam: Elsevier Science Publishers, pp 37-40. Lob0 E de H , K h a n M, Tew J (1980): Community study of hypothyroidism in Down’s syndrome. Br Med J 280:1253-1255. Mayr WR, Kirnbauer M, Paush V, Andrle M, Rett A, Zdansky R (1985): HLA and trisomy 21. Am J Hum Genet 37:606-607. McLachlan DRC, Lukiw WJ, Wong L, Bergeron C, Bech-Hansen T 11988): Selective mRNA reduction in Alzheimer’s disease. Mol Brain Res 3:255-262. Mortimer J A 11990): Genetic and environmental risk factors for Alzeheimer disease: Key questions and new approaches. In Altman H (eds):“Alzeheimer Disease and Dementia. Problems, Prospects, and Perspectives.” New York: Plenum Press (in press). Oliver C, Holland J (1986): Down’s syndrome and Alzheimer’s disease: A review. Psycho1 Med 16:307-322. Patel YC, Burger HG, Hudson BM (1971):Radioimmunoassay ofserum thyrotropin: Sensitivity and specificity. J Clin Endocrinol Metab 33:768-774. Pekary AE, Mire11 CJ, Turner LF J r , Walfish PG, Hershman J M (19871: Hypothalamic secretion of thyrotropin releasing hormone declines in aging rats. J Gerontol 47:447-450. Percy ME, Dalton AJ, Markovic VD, Crapper McLachlan DR, Hummel JT, Rusk ACM, Andrews DF (1989):Red cell superoxide dismutase, glutathione peroxidase and catalase in Down syndrome patients with and without manifestations of Alzheimer disease. Am J Med Genet (in press). Portmann L, Hamada N, Heinrich G, De Groot U (1985):Antithyroid peroxidase antibody in patients with autoimmune thyroid disease: Possible identity with antimicrosomal antibody. J Clin Endocrinol Metab 61:1001-1003. Rosenthal MJ, Hunt WC, Garry PJ, Goodwin R (1987):Thyroid failure in the elderly. Microsomal antibodies a s discriminant for therapy. JAMA 258:209-247. Rousset B, Bernier-Valentin F, Poncet C, Orgiazzi J , Madec AM, Monier JC, Mornex R (1983): Anti-tubulin antibodies in autoimmune thyroid disorders. Clin Exp Immunol 52:325-332. Ruf J , Czarnocka B, De Micco C, Dutoit C, Ferrand M, Carayon P (1987):Thyroid peroxidase is the organ-specific “microsomal” autoantigen involved in thyroid autoimmunity. Acta Endocrinol [Suppl] 281:49-56. Sare Z, Ruvalcaba RHA, Kelley VC (1978):Prevalence ofthyroid disorders in Down’s syndrome. Clin Genet 14:154-158. Sawin CT, Bigos ST, Land S,Bacharach P (1985): The aging thyroid: Relationship between elevated serum thyrotropin level and thyroid antibodies in elderly patients. Am J Med 79:591-595. Schroffner WG 11987): The aging thyroid in health and disease. Geriatrics 42:41-52.

154

Percy et al.

Schussler GC, Schaffner F, Korn F (19781: Increased serum thyroid hormone binding and decreased free hormone in chronic active liver disease. New Engl J Med 229:510-515. Seligson H, Seligson D (1972): Measurement of thyroxine by competitive protein binding. Clin Chim Acta 38:199-205. Seyerson TP, Kauer J, Wolfe HC, Mobtaker H, Wu P, Jackson IMD, Lechan RM (19871: Thyroid hormone regulates TRH biosynthesis in the paraventricular nucleus of the rat hypothalamus. Science 238:78-80. Singh VK, Fudenberg HH (1986): Detection of brain autoantibodies in the serum of patients with Alzheimer’s disease but not Down’s syndrome. Immunol Lett 12:277-280. Singh VK, Fudenberg HH, Brown FR I11 (1987): Immunologic dysfunction; Simultaneous study of Alzheimer’s and older Down’s patients. Mech Ageing Dev 37257-264. Snedecor GW, Cochran WG (1980): “Statistical Methods.” 7th ed. Ames, Iowa: Iowa State University Press. Tanaka M, Miyatake T (1983): Anti-acetylcholine antibody in aged

individuals and in patients with Down’s syndrome. J Neuroimmunol 4:17-24. Vladutiu AO, Chun TC, Victor A, Gienau L, Bannerman RM (1984): Down’s syndrome and hypothyroidism: A role for thyroid autoimmunity? Lancet 1:1416. Walfish PG, Gryfe CI (1982):Testing your older patients’ thyroid function. Geriatrics 37:135-150. Walfish PG, Gottesman IS, Baxter J L (1982): Graves’ ophthalmopathy and sabclinical hypothyroidism: Diagnostic value of the thyrotropin releasing hormone test. Can Med Assoc J 127:291-293. Walford RL (1982): Immunological studies of Down’s syndrome and Alzheimer’s disease. Ann NY Acad Sci 396:95-106. Wisniewski KE, Dalton AJ, McLachlan DRC, Wen GY, Wisniewski HM (19851: Alzheimer’s disease in Down syndrome: Clinicopathologic studies. Neurology 35957-961. Yates CM, Simpson J , Maloney AFJ, Gordon A, Reid AH (1980): Alzheimer-like cholinergic deficiency in Down’s syndrome. Lancet II:979.