1475

5. European Collaborative Study. Children born to women with HIV-1 infection: natural history and risk of transmission. Lancet 1991; 337: 253-60. 6. Andiman WA, Simpson J, Olson B, Dember L, Silva TJ, Mille G. Rate of transmission of human immunodeficiency virus type 1 infection from mother to child and short-term outcome of neonatal infection. Am J Dis Child 1990; 144: 758-66. 7. Lindgren S, Anzén B, Bohlin A-B, Lindman K. HIV and child-bearing: clinical outcome and aspects of mother-to-infant transmission. AIDS 1991; 5: 1111-16. 8. Lewis SH, Reynolds-Kohler C, Fox HE, Nelson JA. HIV-1 in trophoblastic and villous Hofbauer cells, and haematological precursors in eight-week fetuses. Lancet 1990; 335: 565-68. 9. Courgnaud V, Lauré F, Brossard A, et al. Frequent and early in utero HIV-1 infection. AIDS Res Hum Retroviruses 1991; 7: 83-88. 10. Ehrnst A, Lindgren S, Dictor M, et al. HIV in pregnant women and their offspring: evidence for late transmission. Lancet 1991; 338: 203-07. 11. Centers for Disease Control. Classification system for human immunodeficiency virus (HIV) infection in children under 13 years of age. MMWR 1987; 36: 225-30; 235-36. 12. Patterson RM, Wood RC. What is twin birthweight discordance? Am J Perinatol 1990; 7: 217-19. 13. Bishop YMM, Fienberg SE, Holland PW. Discrete multivariate analysis, theory and practice. Cambridge, Massachusetts: MIT Press, 1976. 14. Menez-Bautista R, Fikrig SM, Pahwa S, Sarangadharan MG, Stoneburner RL. Monozygotic twins discordant for the acquired immunodeficiency syndrome. Am J Dis Child 1986; 140: 678-79. 15. Park CL, Streicher H, Rothberg R. Transmission of human immunodeficiency virus from parents to only one dizygotic twin. J Clin Microbiol 1987; 25: 1119-21. 16. Mok JQ, Giaquinto C, DeRossi A, Grosch-Womer I, Ades AE, Peckham CS. Infants born to mothers seropositive for human immunodeficiency virus: preliminary findings from a multicentre European study. Lancet 1987; i: 1164-68.

C, Tovo P-A, deMartino M, Galli L, for the Italian Multicentre HIV-1 transmission rate in first born children to seropositive mothers and interfering factors (abstract WC3241). In: VIIth International Conference on AIDS, 1991: 356. Chiodo F, Ricci E, Costigliola P, Michelacci L, Bovicelli L, Dallascasa P. Vertical transmission of HTLV-III. Lancet 1986; i: 739. Wong VCW, Lee AKY, Ip HMH. Transmission of hepatitis B antigens

17. Gabiano

Study.

18. 19.

20.

21.

22. 23. 24. 25. 26.

27.

28. 29.

from symptom free carrier mothers to the fetus and infant. Br J Obstet Gynaecol 1980; 87: 958-65. Growdon WA, Apodaca L, Cragun J, Peterson EM, delaMaza LM. Neonatal herpes simplex virus infection occurring in a second twin of an asymptomatic mother: failure of a modern protocol. JAMA 1987; 257: 508-11. Plaeger-Marshall S, Lewis K, Sullivan-Bolyai J, Bryson YJ. Immunologic assessment of neonatal herpes simplex virus infection in one dizygotic twin. Pediatr Infect Dis J 1989; 8: 171-75. Shearer WT, Schreiner RL, Marshall RE, Barton I.L. Cytomegalovirus infection in a newborn dizygous twin. J Pediatr 1972; 81: 1161-65. Amstey MS, Monif GRG. Genital herpesvirus infection in pregnancy. Obstet Gynecol 1974; 44: 394-97. Franciosi RA, Knostman JD, Zimmerman RA. Group B streptococcal neonatal and infant infections. J Pediatr 1973; 82: 707-18. Minkoff HL. Care of pregnant women infected with human immunodeficiency virus. JAMA 1987; 258: 2714-17. Johnson FD, MacCallum LR, Brettle RP, et al (abstract WC3239). In: VIIth International Conference on AIDS, 1991: 355. Hino S, Sugiyama H, Doi H, et al. Breaking the cycle of HTLV-I transmission via carrier mothers’ milk. Lancet 1987; ii: 158-59. Benirschke K. The placenta in twin gestation. Clin Obstet Gynecol 1990; 33: 18-31. Strandskov HH, Edelen EW. Monozygotic and dizygotic twin birth frequencies in the total white and colored US population. Genetics

1946; 31: 438.

Stressful life events and Graves’ disease

The role of stressful life events in the onset of Graves’ disease (toxic diffuse goitre) is controversial. However, the numerous early clinical reports that supported such an association were not adequately controlled and specificity of the diagnosis could be questioned. Later studies have not shown a causal relation, but these studies were small, did not have proper controls, or epidemiological methods were inappropriate. To assess possible associations between life events, heredity, social support, and Graves’ disease, we have done a population-based case-control study in a defined area with about 1 million inhabitants. Over 2 years, 208 (95%) of 219 eligible patients with newly-diagnosed Graves’ disease and 372 (80%) of all selected matched controls answered an identical mailed questionnaire about marital status,

occupation, drinking and smoking habits, physical activity, familial occurrence of thyroid disease, life events, social support, and personality. Compared with controls, patients claimed to have had more negative life events in the 12 months preceding the diagnosis, and negative life-event scores were also significantly higher (odds ratio 6·3, 95% confidence interval 2·7-14·7, for the category with the highest

negative score). Individuals who had relatives with thyroid disease (especially first-degree and seconddegree relatives) were more likely to have Graves’ disease (3·6, 2·2-5·9). Slightly more patients than controls were divorced (1 ·8, &1middot;0-3·3) and reported a less frequent intake of alcohol (0·4, 0·2-0·8). When results were adjusted for possible confounding factors in multivariate analyses, risk estimates were almost unchanged. These findings indicate that negative life events and hereditary factors may be risk factors for Graves’ disease.

Introduction The

largely

cause of Graves’ disease (diffuse toxic goitre) is unknown. Hereditary factors linked to the HLA

ADDRESSES Departments of Internal Medicine (B. Winsa, MD, Prof A Karlsson, MD) and Statistics (Prof R Bergstrom, PhD), and Cancer Epidemiology Unit (Prof H. O. Adami, MD), Uppsala University, Uppsala; and Departments of Internal Medicine, Örebro Hospital (A. Gamstedt, MD), Västerås Hospital (R. Jansson, MD), and Danderyd Hospital (U. Adamson, MD, P. A. Dahlberg, MD), Sweden. Correspondence to Dr Brita Winsa, Department of Internal Medicine, University Hospital, 751 85, Uppsala, Sweden.

1476

locus’ and Gm allotypes’2 have been associated with an increased susceptibility to the disease and there is a five-fold higher incidence in women than in men. Among exogenous factors, infections have been thought to be a possible caused Additionally, since thyrotoxicosis was first described by Parry in 1825, there have been many reports about a possible association between psychological stress and onset of Graves’ disease,5-9 although such an association has not been proven. In Norway and Denmark the incidence of hyperthyroidism increased in the first years of the 1939-45 war. 8,9 By contrast, Hadden et a110 found no change in incidence of the disease during the civil unrest in Ireland, and results from a prospective study in which the control group consisted of patients with non-toxic goitre, suggested that stress was not a cause."I To see whether life events influence the development of Graves’ disease, and to assess any possible interactions between life events and hereditary factors, we have done a large population-based case-control study.

Subjects and methods Patients Patients with newly diagnosed Graves’ disease who were referred the departments of internal medicine in four county hospitals in central Sweden from July 1,1987, to June 30,1989, were eligible for the study. The total population, from which the controls were selected, was almost 1 million in the four counties, and the catchment area of the county hospitals included 860 000 of these inhabitants. Thus, it is reasonable to assume that the study included about 86% of incident cases in the source population. There were 219 eligible patients. Diagnosis of Graves’ disease was based on clinical signs, raised thyroid hormone, and suppressed thyrotropin (measured by a sensitive immunometric assay). Thyrotropin-receptor antibodies, measured by a radioreceptor binding assay (’TRAK’, Henning, Berlin, Germany) were found in 195 of the 219; the other 24 patients (with undetectable thyrotropinreceptor antibodies) had clinical signs typical of Graves’ disease. Thyroid scanning in 131 patients (60%) showed an increased, to

homogeneous uptake. Patients were asked to take part in the study immediately on referral to the hospital, which was generally less than 2 weeks after

diagnosis. 208 patients (95 %) completed the questionnaire. Missing were completed by phone in 54 of the cases. There were 171 women (82%) (mean [SD] age 45 years [14’1]) and 37 men (18%) (47-9 [14-5]). observations

Controls For each patient with Graves’ disease, 2 controls were chosen from the population register, which includes all Sweden. The controls were randomly selected among individuals matched for sex, date of birth, and current county residence. For every 2 controls, 2 additional controls were selected for replacement in the event that a control should refuse to take part. Within a month of patient selection, 2 controls were informed by post (with one reminder) about the study and shortly after contacted by phone. 463 controls were contacted and 372 (80%) completed questionnaires. The responding cases and controls were evenly distributed within the area of referral. Missing observations were completed by phone in 79 controls. Among the 91 non-responders, 39 did not give a reason for not replying, 21 did not answer because they were afraid to have their names on the register, 17 declined to take part but did not explain why, 9 found the questions too personal, 3 questionnaires were lost in the mail, 1 person was too ill to take part, and 1 did not complete the form correctly.

Data collection The four-part questionnaire consisted of 95 items. The first part asked about marital status, occupation (by nine categories:

TABLE I-RISK OF GRAVES’ DISEASE ACCORDING TO NEGATIVE AND POSITIVE Ll FE- EVENT SCORES

multivariate analysis for total negative/positive in categorised forms life-event score

*Adjusted

m

scores

and

negative/

positive scores LES

=

industrial workers, miners, and other manual workers; individuals employed in jobs requiring university education; unemployed people; individuals in service occupations; hospital staff and laboratory personnel; housewives; retired people; owners of enterprises; and others), drinking habits (defmed as frequency of alcohol intake; never, < once a month, 1-3 times a month, once a week), smoking habits (never, 1-10 cigarettes daily, > 10 daily, snuff/pipe), and physical activity (none, once a week, or twice a week). Familial occurrence of thyroid disease was ascertained in detail. The responders were asked to list all relatives with thyroid diseases and to specify type of disease and treatment. Finally, information was obtained about other diseases, present medication, and reproductive history (including age at menarche and menopause, parity, abortions, and miscarriages). In the second part, modified from Sarason’s life-event survey,12 29 possible life events, such as marriage, bereavement, pregnancy, and educational and occupational successes or failures, were listed. Additionally, responders were asked to indicate any important event not included in the list (83 additional events were given by the case group and 23 by the control group). The date of the event was recorded and its positive or negative effect was classified by the responder on a 5-point scale from very positive (+2) to very negative ( - 2). These data were analysed according to principles agreed on in advance. The total positive and total negative scores for events within the previous 12 months for each subject were added together to give a total change score-ie, an estimate of the total amount of rated change experienced during the previous year. The third part of the questionnaire was a shortened version of Sarason’s social support questionnaire (SSQ)13 in which the responders listed the initials of all the people they could depend on, in nine specified situations, as well as their degree of satisfaction with these relationships. The number (N) score for each items of the SSQ is the number of support persons listed, and the satisfaction (S) score is based on ratings on a scale ranging from 4 ("very satisfied") 1 ("very dissatisfied"). The last part of the auestionnaire consisted of 10 statements about the degree of mastery (satisfaction and personal influence on life/work) as well as 3 questions about work mastery and, finally, 9 statements about personality traits obtained from the Jenkins’ activity survey," to measure type A behaviour. An English translation of the questionnaire is available on request. to

Statistics The logistic regression model was used in both univariate and multivariate analyses. This model assumes that the logarithm of the odds of getting diffuse toxic goitre depends on different explanatory variables, such as stress and heredity. Since the data collection procedure was matched, estimates were obtained by the conditional maximum likelihood method.15 This method allows for a variable

1477

TABLE II-RISK OF GRAVES’ DISEASE ACCORDING TO NEGATIVE LIFE-EVENT SCORE IN THREE SEPARATE PERIODS BEFORE

DIAGNOSIS/PSEUDODIAGNOSIS

Fig 1-Mean negative score for cases and controls months before diagnosis/pseudodiagnosis.

in the 12

Observation times, adjusted as described in methods, resulted apparently low incidence of life events in months 12and 11

in

number of controls. Models were estimated with variables in both continuous and categorised form. The latter alternative implies that non-linear effects of the variables are allowed. We used the odds ratio (OR), a measure of the relative risk, to assess the association between the different variables and Graves’ disease. Controls were assigned a date of "pseudodiagnosis" according to the date on which they answered the questionnaire. Only events occurring before diagnosis or pseudodiagnosis were considered in the analysis. Although the questionnaire asked about events during the 12 months before the answering date, in some cases the date of diagnosis was a few months before the answering date. Thus, the observation times for the matched controls were appropriately

adjusted.

Results Life events, social support, and personality Life events with a positive or negative effect during the 12 months before diagnosis for the cases or date of reply for controls were used to calculate positive, negative, and total change scores. These scores were analysed to evaluate an association between any of the three scores and onset of Graves’ disease. The relative risk of disease rose with

*Analysis including all three periods tUnlvanate analysis OR

=

odds ratio, C)=conf!dence Interval

increasing negative scores but there was no association between positive life events and Graves’ disease (table 1). When the life-event score was analysed in continuous form, the risk of Graves’ disease increased significantly by 30% for each negative point whereas there was no association for positive events. The results were largely unchanged in a multivariate analysis with mutual adjustment for negative and positive events (table 1). The total change score also pointed to an increased relative risk, but this was entirely due to the increase caused by the negative score (data not shown). The mean negative score (fig 1), but not the mean positive score (fig 2) or the distribution of events over time, differed between the

two

groups. The distribution pattern of the

mean number of

negative and positive events was almost the the mean scores (data not shown) revealing that the grading of events were similar in both groups. To see whether the increased negative score was merely a result rather than a cause of disease, these scores were analysed for three periods (1-4, 5-8, 9-12 months) before date of diagnosis or pseudodiagnosis. The association between negative life events (continuous) and Graves’ disease was significant for each period (table 11). The death of a close relative or friend was reported by 15% of cases and 10% of controls (p > 005). A further analysis of negative scores after stratification of the data showed that the association with negative life events did not differ between the age groups ( 45 vs > 45 years) or the sexes. The OR for the total same as

negative score, analysed as a continuous variable, was 1-3% (95% CI 1-1-1-5) for the younger group, 1-4 (1-1-1-6) for the older group, 1-4 (1-0-2-1) for men, and 1-3 (1-2-1-5) for women.

Fig 2-Mean positive score for cases and controls in the 12 months before diagnosis/pseudodiagnosis. Observation times, adjusted as described in methods, resulted in apparently low incidence of life events in months 12and 11

Neither the N score nor the S score differed significantly between cases and controls. A work mastery score, based on three questions, suggested that the controls were more able to relax after work and also felt more secure about their employment than did the cases (table III). However, a score for mastery in life, apart from work, did not differ between the groups. Finally, cases tended, though not significantly, towards more type A behaviour ( 25 points) than did controls (28% vs 18%).

1478

TABLE III-RISK OF GRAVES’ DISEASE ACCORDING TO WORK MASTERY, MARITAL STATUS, AND DRINKING HABITS

more

there

than five such episodes (OR 2-5; was no regular trend.

95% CI 1-0-6-4), and

Reproductive characteristics Among women, there were no significant differences

or

trends between the groups with respect to age at menarche or whether they were premenopausal or postmenopausal. More women among cases had been pregnant (1-7,0-9-3-1), had had an abortion (1 -4, 0-9-2-4), or had had a miscarriage (1 -4, 0-8-2-3) but none of these differences was significant. No differences were found with respect to whether they had children (1-2, 0-7-2-1) or to the number of children. 10 patients (6%) had given birth or had an abortion or a miscarriage during the past year compared with 9 controls

(3%) (data not shown). Marital status, job type, exercise, and alcohol OR =odds ratio, C

I internal

The risk of having Graves’ disease was slightly but increased among divorced individuals than among those who were married or cohabiting (table III). There were no differences between the groups with respect to category of occupation or current amount of physical exercise (data not shown). The controls consumed alcohol more often than did cases (table III).

significantly

Hereditary factors Familial occurrence of any thyroid disease was strongly associated with the development of Graves’ disease (table IV). For individuals with a relative with thyroid disease the risk was increased almost four-fold (three-fold for a first-degree relative and six-fold for a second-degree relative). 65 (31 %) of the cases and 43 (12%) of the controls had more than one relative with thyroid disease. In a separate analysis there was no association between hereditary factors and negative life events (data not shown). When the data were stratified by age ( 45 vs > 45 years) the association with familial occurrence of any thyroid disease was stronger in the younger than in the older group (OR 4-9 [95% CI 25-92] vs 2-9 [1-5-5-4]). Men and women were broadly similar with respect to hereditary factors (data not shown). The population attributable risk for any familial occurrence of thyroid disease was calculated as 41 % from the formula: (Pe [OR-1]) =(Pe[OR-l] + 1) where Pe is the proportion exposed in the population (ie, 81/320 for controls) or 74% for the exposed cases ([OR-1] — OR).16

Multivariate and interaction

analyses All variables that were important in the univariate analyses or regarded as of interest a priori were further analysed in a multivariate model. With the exception of divorce, the associations found in the univariate analysis remained largely unaltered indicating that no major confounding took place (data not shown). An analysis of a possible interaction between negative life-event score and familial occurrence of thyroid disease (any relative) did not give any indication of such an interaction as shown by ORs of 1 -31 and 1-27 per unit of the negative life-event score among individuals without and with familial occurrence of the disease, respectively. In respect of OR this implied a difference per unit of negative life-event score between the groups of 0-97 (95% CI 0-76-1-23).

Medical history A similar proportion of cases and controls had had other chronic diseases (OR 1-1; 95% CI 0-7-1-6). A history of frequent infections--eg, urinary tract infections and coldsduring the previous 12 months was associated with an increased risk of Graves’ disease but only when there were TABLE IV-RISK OF GRAVES’ DISEASE ACCORDING TO FAMILIAL OCCURRENCE OF ANY THYROID DISEASE AND GRAVES’

DISEASE

*Some categories add up to less than 208 (cases) and 372 (controls) Missing individuals had answered "don’t know" on this variable and were not included In the

analysis †Adjusted for all variables analysed

in multivariate analysis negative and positive life-event scores, alcohol mtake, divorced, smoking, social support, infections

Discussion The principal findings of our study were that negative life and familial occurrence of thyroid disease were significantly associated with the occurrence of Graves’ disease. Also, prevalence of divorce was higher and job satisfaction was lower in cases than in controls. As in all epidemiological case-control studies, selection bias and confounding factors should be considered. In our study, there may have been a selection bias because patients consisted of most but not all incident cases in the four counties from which the controls were chosen. However, there is no reason to assume that the cases included were referred to the four hospitals because of a greater exposure to negative life events, or that geographical differences affected the risk estimates. The same assumption can be made about selection bias due to heredity. Recall bias is a more serious concern since it may lead to overestimation of the association between Graves’ disease and negative life events as well as the familial occurrence of thyroid disease. Patients may be more likely than healthy controls to report negative events, especially if such events are thought to be a consequence of the disease. However, this possibility is unlikely to account events

1479

for such a large increase in the relative risk and is also partly offset by the fact that both groups reported an equal number of positive events. 20% of the controls did not answer the questionnaire, and non-response can be associated with recent occurrence of negative life-events. However, when the 91 non-responders were included in the analysis and distributed with respect to negative life events in the same way as the cases, the risk estimates remained largely unchanged (data not shown). We chose date of diagnosis as onset of the disease. Since the latency period from the true onset to the date of diagnosis may vary, the difficulty is knowing whether the life events took place before or after onset. However, the temporal pattern that we found is evidence against recall bias as a principal explanation of the excess risk. Moreover, the multivariate analysis, which adjusted for possible confounding factors, did not greatly change the risk estimates. Numerous studies have pointed to a connection between stress and alterations in the immune system,17-20 but the mechanisms by which a stressful life event might influence the immune system are unknown. Our finding that divorce was associated with a slightly increased risk of having Graves’ disease has not been shown in previous studies, although divorced individuals have an increased morbidity, especially with respect to psychiatric disorders 21 On the other hand, that there was no difference in social support, or satisfaction with it, between our two groups contrasts with research suggesting that social support may be beneficial during stressful events 22 Moreover, since both our groups claimed to have had an equal number of reliable supporters as well as satisfaction with these social relationships, there was probably no recall bias. Hereditary factors are known to have a role in autoimmune thyroid diseases.23 The influence of sex is also well known, the female-to-male ratio in Graves’ disease being about 5/1.3 Although the incidence of both Graves’ disease and autoimmune thyroiditis is raised among close relatives,24 the link is weak and more than half the patients with Graves’ disease have no positive family history. Similarly, the associations between autoimmune thyroid disease and MHC class II antigens25 and between Graves’ disease and HLA-B8/DR3 in caucasians26,27 and HLA-B35, DRwl2 in Japanese28 are weak, and up to half the patients have no associated antigen. We have confirmed that heredity has a role in Graves’ disease since patients with a family history of thyroid disease had about a four-fold higher risk; that the association was stronger for second-degree relatives might be due to chance or to better recall among cases than among controls of thyroid disease in more distant relatives. The effect of negative life events was similar in individuals with or without familial occurrence of thyroid disease, which implies that these two risk factors influence the occurrence of Graves’ disease through different causal mechanisms. Viral infections have been proposed, though not proven, as one of the causes of many autoimmune diseases. In our study, the weak association between Graves’ disease and frequent episodes of infections (both viral and bacterial) in the year preceding onset of the disease may support this hypothesis, but it may also suggest that a patient with Graves’ disease is more susceptible to microbial infections. Graves’ disease often recurs and several attempts have been made to identify individuals with an increased risk of relapse. Age, hormone levels, and goitre size are important prognostic factors.29 Our results support the hypothesis that stress can be an important factor in the cause of the disease.

We thank Dr Matthew Zack for valuable comments on the manuscript and Mrs Ing-Marie Carlsson for excellent secretarial assistance. This work was supported by a grant from Planning and Coordination of Research.

the Swedish Council for

REFERENCES H, Uno H, Hirota Y, et al. Immunogenetics of Hashimoto’s and Graves’ diseases. J Clin Endocrinol Metab 1985; 60: 62-66. 2. Uno H, Sasazuki T, Tamai H, Matsumoto H. Two major genes, linked to HLA and Gm, control susceptibility to Graves’ disease. Nature 1981; 292: 768-70. 3. Volpé R. Autoimmune thyroid disease. In: Volpé R, ed. Autoimmunity and endocrine disease. New York: Marcel Dekker, 1985; 109-286. 4. Weetman AP, McGregor AM. Autoimmune thyroid disease: developments in our understanding. Endocrine Rev 1984; 5: 309-55. 5. Weisman SA. Incidence of thyrotoxicosis among refugees from nazi prison camps. Ann Intern Med 1958; 48: 747-52. 6. Bennett AW, Cambor CG. Clinical study of hyperthyroidism. Arch Gen Psychiatry 1961; 4: 160-65. 7. Lidz T. Emotional factors in the etiology of hyperthyroidism occurring in relation to pregnancy. Psychosom Med 1955; 17: 420-27. 8. Grelland R. Thyrotoxicosis at Ullevål hospital in the years 1934-44 with a special view to frequency of the disease. Acta Med Scand 1946; 125: 108-38. 9. Meulengracht E. Epidemiological aspects of thyrotoxicosis. Arch Intern Med 1949; 83: 119-34. 10. Hadden DR, McDevitt DG. Environmental stress and thyrotoxicosis. Lancet 1974; ii: 577-78. 11. Gray J, Hoffenberg R. Thyrotoxicosis and stress. Q J Med 1985; 54: 153-60. 12. Sarason IG, Johnson JH, Seigel JM. Assessing the impact of life changes: development of the Life Experience Survey. J Consulting Clin Psychol 1978; 46: 932-46. 13. Sarason IG, Levine HM, Basham RB, Sarason BR. Assessing social support: the social support questionnaire. J Personality Soc Psychol 1983; 44: 127-39. 14. Jenkins C, Rosenman RH, Zyzanski SJ. The Jenkins Activity Survey for Health Prediction. Chapel Hill, N.C: C David Jenkins, 1965. 15. Breslow NE, Day NE. Statistical Methods in Cancer Research. The Analysis of Case-Control Studies. Lyon: International Agency for Research on Cancer, 1980. 16. Kuritz SJ, Landis JR. Attributable risk estimation from matched case-control data. Biometrics 1988; 44: 355-67. 17. Vialettes B, Ozanon JP, Kaplansky S, et al. Stress antecedents and immune status in recently diagnosed type 1 (insulin-dependent) diabetes mellitus. Diabetes Metab 1989; 15: 45-50. 18. Robinson N, Lloyd CE, Fuller JH, Yateman NA. Psychosocial factors and the onset of type 1 diabetes. Diabetic Med 1989; 6: 53-58. 19. Stein M, Keller S, Schleifer S. Stress and immunomodulation: the role of depression and neuroendocrine function. J Immunol 1985; 135: 827-33. 20. Croiset G, Heijnen C, Veldhuis HD, de Wied D, Ballieux RE. Modulation of the immune response by emotional stress. Life Sci 1987; 1. Tamai

40: 775-82.

21. Bloom BL, Asher SJ, White SW. Marital distruption as a stressor: a review and analysis. Psychological Bull 1978; 85: 867-94. 22. Hirsch BJ. Natural support systems and coping with major life changes. Am J Community Psychol 1980; 8: 63-69. 23. Bartels ED. Heredity in Graves’ disease. Copenhagen: Ejnar Munksgaards Forlag, 1941. 24. Skillern PG. Genetics of Graves’ disease. Mayo Clin Proc 1972; 47: 848-49. 25. Piccini LA, Roman SH, Davies TF. Autoimmune thyroid disease and thyroid cell class II major histocompatibility complex antigens. Clin Endocrinol 1987; 26: 253-72. 26. Farid NR, Sampson L, Noel EP, et al. A study of human D locus-related antigens in Graves’ disease. J Clin Invest 1979; 63: 108-13. 27. Allanic H, Fauchet R, Larcy Y, et al. HLA and Graves’ disease: an association with HLA-DRw3. J Clin Endocrinol Metab 1980; 51: 863-70. 28. Sasazuki T, Kobno Y, Iwanesto L, Tanimura M. HLA-B-D haplotypes associated with autoimmune disease in Japanese populations. Tissue Antigens 1977; 10: 21-28. 29. Winsa B, Dahlberg PA, Jansson R, Ågren H, Karlsson FA. Factors influencing the outcome of thyrostatic drug therapy in Graves’ disease. Acta Endocrinol (Copenh) 1990; 122: 722-28.

Stressful life events and Graves' disease.

The role of stressful life events in the onset of Graves' disease (toxic diffuse goitre) is controversial. However, the numerous early clinical report...
768KB Sizes 0 Downloads 0 Views