General Hospital Psychiatry 28 (2006) 503 508

Self-reported thyroid disease and mental disorder prevalence in the general populationB
Scott B. Patten, M.D., F.R.C.P. (C), Ph.D.a,b,4, Jeanne V.A. Williams, B.A., M.Sc.a, Eleonora Esposito, M.D.c, Cynthia A. Beck, M.A.Sc., M.D., F.R.C.P. (C)b
a

Department of Community Health Sciences, University of Calgary, Calgary, Alberta, Canada T2N 4N1 b Department of Psychiatry, University of Calgary, Calgary, Alberta, Canada T2N 2T9 c University of Verona, 37100 Verona, Italy Received 2 May 2006; accepted 11 September 2006

Abstract Objective: Community studies have failed to confirm that biochemically assessed thyroid status is significantly associated with psychopathology. However, it has been reported that self-reported thyroid disease is associated with symptoms of depression and anxiety. The objective of the current study was to determine whether self-reported thyroid disease is associated with elevated mental disorder prevalence in the general population. Method: Data from the Canadian Community Health Survey (CCHS) 1.2: Mental Health and Well-being were used. The CCHS 1.2 included the World Mental Health version of the Composite International Diagnostic Interview and collected self-report data about professionally diagnosed chronic medical conditions, including thyroid disease. Results: Twelve-month and lifetime mental disorder prevalence was higher in subjects with thyroid disease than in subjects reporting no chronic conditions. For each condition examined (major depressive disorder, bipolar disorder, panic disorder/agoraphobia and social phobia), the 12-month and lifetime prevalence in subjects with thyroid disease resembled that of an aggregate category of subjects having other chronic conditions. After adjustment for age, sex and other chronic conditions, only social phobia was found to be associated with thyroid disease. Conclusions: People with thyroid disease are not a particularly high-need group for mental disorder screening or intervention, at least not in the community population. D 2006 Elsevier Inc. All rights reserved.
Keywords: Major depressive disorder; Panic disorder; Agoraphobia; Social phobia; Drug dependence; Thyroid disease

1. Background Traditionally, thyroid dysfunction has been regarded as a biological risk factor for mood and anxiety disorders. Case reports associating mental status changes with thyroid disease date back to some of the earliest clinical descriptions of thyroid disease (see review by Whybrow et al. [1]). Hyper- and hypothyroidism are listed among the causes of mood and anxiety disorders due to a general medical
The analyses presented in this article were derived from data collected by Statistics Canada. However, the analyses were conducted by the authors and do not reflect the interpretations or opinions of Statistics Canada. 4 Corresponding author. University of Calgary, Community Health Sciences, Calgary, Alberta, Canada T2N 4N1. Tel.: +1 403 220 8752; fax: +1 403 270 7307. E-mail address: patten@ucalgary.ca (S.B. Patten). 0163-8343/$ see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.genhosppsych.2006.09.001
B

condition in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV ) [2]. The contemporary literature provides some support to the notion that overt disturbances of thyroid function are associated with psychiatric disturbances. Bunevicius et al. [3] conducted a casecontrol study of women hospitalized with Graves disease, some of them with continued hyperthyroidism and some without. Upon assessment with the Mini International Neuropsychiatric Interview [4,5], a higher prevalence of major depression, bipolar disorder, panic disorder and generalized anxiety disorder was found in these patients than in a gynecological control group. Trzepacz et al. [6] used the Schedule for Affective Disorders and Schizophrenia [7] to assess a series of 13 patients admitted to a clinical research unit with untreated Graves disease. Each subject was found to have at least one mood or anxiety

504

S.B. Patten et al. / General Hospital Psychiatry 28 (2006) 503 508

disorder, including nine subjects with major depression. Placidi et al. [8] used a modified version of the SCID [9] to assess a series of 93 inpatients with thyroid disease (predominantly hyperthyroidism). Psychiatric diagnoses were identified in 68 of these subjects, the most common conditions being panic disorders and generalized anxiety disorder. Kathol and Delahunt [10] examined a series of 30 patients presenting to an endocrinology clinic with untreated hyperthyroidism. Depressive disorders were identified in 10 of these patients, but this number was reduced to 3 when modified diagnostic criteria that de-emphasized thyroid symptoms were applied. The prevalence estimates reported in the aforementioned studies are high, suggesting that psychiatric syndromes are common in overt thyroid disease. However, since all but one of these studies were uncontrolled, they fail to adequately quantify the extent of association. For example, in the single controlled study by Bunevicius et al. [11], 60% of Graves disease patients had a current depressive episode; however, 40% of the control group also had a current depressive episode. In view of this, the range of estimates reported by other studies (10% to 100%) is difficult to interpret. Furthermore, as all of these were clinical studies, their results cannot necessarily be generalized to the population level. Hospitalized patients in particular are known to have a high prevalence of psychiatric disorders [12]. Treatment protocols for thyroid cancer may involve periods of withdrawal from thyroid supplementation to deliberately induce a hypothyroid state. Several studies have examined changes within individual subjects as they move from a euthyroid to a hypothyroid state during cancer treatment or posttreatment monitoring. Constant et al. [13] identified small increases in depression and anxiety ratings during this transition. However, the clinical significance of the increase was uncertain. The mean Beck Depression Inventory [14] rating in the hypothyroid state was 12, which is lower than the level expected in clinically significant depressive episodes. Increases in anxiety ratings were found to be somewhat more prominent than those associated with depression in this study. Burmeister et al. [15] reported a similar change in Beck Depression Inventory ratings (from a mean score of 7.3 to 15.3) in a series of 13 athyrotic cancer patients after they had discontinued their thyroid supplementation. The low frequency of depressive symptoms and the more prominent increase in anxiety symptoms during temporarily induced hypothyroid states have also been replicated by Tagay et al. [16]. These studies confirm that symptoms of depression and anxiety, particularly the latter, can change in intensity in association with large changes in thyroid status. However, the clinical significance of the reported changes is uncertain, and the relevance of observations made in this specific context is unknown. Most of the literature about mental status changes in thyroid diseases derives from severely ill clinical samples. The extent to which less severe disturbances of thyroid function contribute to psychopathology is more uncertain. Haggerty et al. [17] recruited a sample of 31 female

volunteers, 16 of whom had evidence of subclinical hypothyroidism. None of the subjects had current major depression, but the lifetime prevalence of major depression was found to be elevated in the subclinically hypothyroid group. The finding was interpreted as evidence that subclinical hypothyroidism was a possible risk factor for major depression. However, this study deliberately recruited volunteers with a past history of borderline thyroid function. This would have created a vulnerability to selection bias since subjects with a past history of major depression are probably more likely to have undergone thyroid testing and to be aware of their borderline thyroid status. Oomen et al. [18] compared the frequency of thyroid abnormalities (TSH and thyroid antibodies) in psychiatric inpatients to that of healthy controls. No differences were found. However, the authors reported that when psychiatric subjects with thyroid abnormalities were compared to those without such abnormalities, rapid cycling bipolar disorder was found to be more common in the former group. Several studies have sought to evaluate the importance of mild or subclinical thyroid abnormalities by assessing response to treatment. Bono et al. [19] described a series of patients with mild hypothyroidism in whom depression ratings improved following treatment. However, the scale used to assess depression was one that is more appropriate for evaluation of symptom severity in people with diagnosed depressive disorders: the Hamilton Depression Rating Scale [20]. Furthermore, a mean improvement of only 1 point on this scale was observed. This small change could easily be due to regression to the mean or to placebo effects. These same vulnerabilities to bias apply to a Turkish study that shows improvement in psychological symptoms in subjects being treated for overt and subclinical hypo- and hyperthyroidism [21]. In the absence of appropriate controls, these treatment studies do not provide strong evidence that mild or subclinical disturbances of thyroid function are associated with significant psychiatric morbidity. The literature of general population studies is smaller than that of clinical studies. One large-scale epidemiological study was reported by Engum et al. [22]. These investigators obtained symptom ratings for depression and anxiety from a large community sample who also answered questions about their thyroid history and provided a blood sample. Subjects with biochemical hypothyroidism (assessed with a single basal TSH and T-4 assay) had lower levels of depressive and anxious symptoms than euthyroid subjects. However, those self-reporting a history of thyroid disease of any type (current or past) had higher levels of both depressive and anxious symptoms [22]. In another analysis, no association between thyroid autoimmunity and symptoms of depression or anxiety was found [23]. Another population-based study to evaluate thyroid status biochemically was the Study of Health in Pomerania [24]. This study employed a scale to assess a variety of mental and physical complaints, including depression and anxiety (evaluated with single items in a questionnaire), in a

S.B. Patten et al. / General Hospital Psychiatry 28 (2006) 503 508

505

community sample (n = 3790) having no previously known thyroid disorder. Biochemically confirmed hypothyroidism was not associated with increased symptomatic complaints, but hyperthyroidism was associated with a lower frequency of some symptoms. Data from the Maastricht Aging Study was used to evaluate possible associations between thyroid dysfunction and cognitive decline in an elderly community sample [25]. This study included a depressive symptom rating scale and found no significant correlation between TSH and depression scores. Some waves of the National Health and Nutrition Survey included diagnostic measures for major depression and also measured TSH and T4. These data were analyzed by Forman-Hoffman and Philibert [26], who reported that men with current depression had slightly lower TSH and higher T4 whereas women also had higher T4 but did not have lower TSH. These differences were interpreted as state-dependent changes in thyroid indices during depressive episodes. Two Danish register studies have examined possible associations between thyroid disease and psychiatric admissions. One of these studies reported that patients with an admission for bipolar disorder were more likely to be subsequently admitted to hospital with hyperthyroidism [27]. Another analysis of the registry found that patients admitted with a diagnosis of hypothyroidism had a higher frequency of a subsequent admission with an affective disorder, especially during the year following the index admission, than did patients with osteoarthritis or nontoxic goiter [28]. Despite the use of national registry data, the number of new admissions with bipolar disorder was too small to evaluate. Because hospital admission data were used as an indicator of psychiatric and thyroid status, these studies may have been vulnerable to selection bias (Berksons bias). In other words, people with comorbid thyroid disease and mental disorders may have been more likely to be admitted, creating the false impression of an association, even if the two conditions were not associated in the population. From the public health perspective, the possibility of an association between a history of thyroid disease and psychiatric morbidity, as reported by Engum et al. [22], is potentially significant. In distinction to the current emphasis on screening for biochemical thyroid abnormalities in people with mood and anxiety disorders, it is possible that the emphasis should be on case-finding for mental illness in subjects with a history of thyroid disease. However, the Engum study did not employ a validated measure of depressive disorders. For these reasons, we sought to evaluate the association between a variety of disorders and self-reported thyroid disease in a large community sample. 2. Methods The Canadian Community Health Survey (CCHS) 1.2 was a nationally representative community mental health survey conducted by Statistics Canada (Canadas

national statistical agency) between May 2002 and December 2002. Detailed information about the methods employed in this study may be found in a paper by Gravel and Be land [29]. The target population included persons aged 15 years and above living in private occupied dwellings (98% of the population). Excluded were individuals living in health care institutions, on Indian Reserves, on government-owned land, in one of the three northern territories or in some remote regions. Full-time members of the armed forces were sampled separately and were not included in the current analysis. One person aged 15 years or above was randomly selected from each sampled household. A significant effort was made to interview respondents in person at their place of residence (86% of cases). Interviews were conducted in English, French, Chinese or Punjabi (as required or requested by the interviewee). From an initial selection of 48,047 households, there was an 86.5% household-level response rate, and among responding households, there was an 89.0% person-level response rate. The overall response rate was thus 77.0%, resulting in a total sample size of 36,984 respondents. The CCHS 1.2 interview included the World Mental Health version of the Composite International Diagnostic Interview (WMH-CIDI) [30], adapted for use in Canada. A copy of the Canadian adaptation is available on the Statistics Canada web page [31]. Well-trained lay interviewers using computer-assisted interviewing procedures administered the survey. Five disorders were evaluated: major depressive disorder, bipolar disorder, social phobia, agoraphobia and panic disorder. Diagnostic algorithms followed DSM-IV criteria, with the exception of the duration requirement for a manic episode. The CCHS asked only whether manic symptoms had lasted bseveral days or longerQ whereas a duration of 7 days is required by the DSM-IV criteria when there has not been a need for hospitalization. In this analysis, we differentiated between major depressive disorder and bipolar disorder by identifying subjects with one or more lifetime manic episodes according to the WMH-CIDI. The WMH-CIDI assigns both lifetime and annual diagnoses. Subjects were also read a list of chronic medical conditions and asked whether they had been diagnosed with one of these conditions by a health professional. The exact wording of the item was: bNow Id like to ask about certain chronic health conditions which you may have. We are interested in long-term conditions which are expected to last, or have already lasted, 6 months or more and that have been diagnosed by a health professional.Q This was followed by bDo you have a thyroid condition?Q Annual and lifetime prevalence of the various mental disorders were cross-tabulated against the presence or absence of a thyroid condition and against the presence or absence of any chronic condition. Next, logistic regression was used to produce estimates adjusted for age, sex and chronic conditions. In these models, mental disorder prevalence was the dependent variable and thyroid status, age and sex were the independent (predictor) variables. The

506

S.B. Patten et al. / General Hospital Psychiatry 28 (2006) 503 508

Table 1 Mental disorder prevalence, by self-reported thyroid status Major depressive disorder, % (95% CI) 12-month Lifetime Bipolar disorder, % (95% CI) 12-month Lifetime Panic disorder/agoraphobia, % (95% CI) 12-month Lifetime Social phobia, % (95% CI) 12-month Lifetime

Thyroid disease 4.2 (3.15.3) 11.8 (10.013.5) 0.8 (0.41.2) Any chronic condition 4.7 (4.35.1) 12.4 (11.813.0) 1.2 (1.01.3) other than thyroid disease No chronic condition 2.4 (2.12.8) 7.5 (6.78.3) 0.5 (0.41.2)
a

3.1a (2.04.2) 2.4 (1.53.4) 5.0 (3.66.3) 3.5 (2.44.6) 8.7 (7.210.2) 2.8 (2.53.1) 2.6 (2.32.9) 4.3 (4.04.7) 3.4 (3.03.7) 8.9 (7.210.2) 1.4 (1.11.7) 1.0 (0.71.3) 2.3 (1.82.8) 2.2 (1.92.6) 6.4 (5.77.1)

Estimate flagged to highlight its imprecision; coefficient of variation is between 13% and 33%. The 12-month prevalence of substance dependence could not be reported because the coefficient of variation exceeded 33%. Lifetime prevalence of substance dependence was not assessed in the CCHS 1.2.

logistic regression analysis began with models including interaction terms. If the interaction terms did not achieve statistical significance, they were removed in order to produce simplified models for descriptive purposes. The CCHS 1.2 used a multistage, stratified cluster design to select eligible households and selected individual respondents from households. To account for these design effects, Statistics Canada recommends a bootstrap procedure that uses a set of replicate weights that they supply. All estimates and models presented here were produced using this approach. The standard errors associated with specific estimates, P values and confidence intervals (CIs) are also adjusted for survey design effects (most notably, cluster sampling, which could otherwise cause an underestimation of variance) by the bootstrap procedure. The bootstrap analysis used a SAS (8) macro developed by Statistics Canada for this purpose. Statistics Canada prohibits release of estimates that are associated with coefficients of variation greater than 33% because these are considered too imprecise to be informative. The estimate of substance use disorder prevalence in subjects with thyroid disease was associated with a coefficient of variation exceeding this level and, consequently, could not be reported. 3. Results Of 36,984 survey subjects, 2414 reported having a thyroid condition. The weighted prevalence was 5.6% (95% CI = 5.35.9). As expected, the prevalence was higher in women (9.0%; 95% CI= 8.49.5) than in men (2.2%; 95% CI= 1.92.5). The prevalence of self-reported thyroid conditions increased with age. The lowest prevalence was seen in the 15- to 24-year-old age group (1.0%; 95% CI = 0.61.4%), and the highest prevalence was in the 65+ age group (11.8%; 95% CI= 10.912.8%). When major depression prevalence was cross-tabulated against thyroid status, the 12-month and lifetime prevalence in subjects with thyroid disease was found to be greater than that of subjects reporting no chronic conditions. However, prevalence in subjects reporting thyroid disease was comparable to that of subjects reporting other chronic conditions but not thyroid disease (see Table 1). For other disorders, the pattern was similar. The prevalence of each

condition was higher in subjects with thyroid disease than in subjects with no chronic conditions but resembled the prevalence in subjects reporting any one or more chronic conditions (see Table 1). After adjustment for age, sex and other chronic conditions, only social phobia had a statistically significant adjusted odds ratio (OR; see Table 2). For social phobia, the adjusted OR was 1.5 (95% CI = 1.02.1). Although this model had an AgeSex interaction (a declining female preponderance with advancing age), there were no interactions that involved thyroid conditions. The proximity of the crude ORs in Table 2 to the null value of 1.0 for most disorders may seem surprising since Table 1 indicates that disorder prevalence in subjects with thyroid disease is comparable to that of other chronic conditions and higher than that of subjects without chronic conditions. This is because subjects with other chronic conditions are included in the denominator of the crude OR in Table 2. As such, the crude ORs presented in Table 2 compare the odds of various disorders in people with thyroid disease to that of those without thyroid disease, many of whom have other disorders that are associated with mood and anxiety disorder prevalence. The adjusted ORs presented in Table 2 do not provide evidence of an association between thyroid conditions and mood disorders. Unlike the association for chronic conditions generally (see Section 4), the associations between thyroid disorders and mood disorders became weaker in the logistic regression analysis. This is probably due to confounding by age. Since female gender is strongly
Table 2 Association of thyroid disease with annual mental disorder prevalence: crude and adjusteda ORs Crude OR Major depression Bipolar disorder Panic disorder/agoraphobia Social phobia 1.1 (0.81.4) 0.8 (0.51.4) 1.2 (0.81.9) 1.2 (0.81.7) Adjusted OR (95% CI) 1.1 (0.81.4) 1.2 (0.72.1) 1.5 (1.02.1) P value (Wald test v 2) .62 (0.24) .34 (0.92) .04 (4.4)

An adjusted OR is not reported for panic disorder/agoraphobia because of a significant ThyroidAgeSex interaction. a Adjusted for age as a continuous variable, for sex and for any chronic condition; some models include AgeSex interaction terms.

S.B. Patten et al. / General Hospital Psychiatry 28 (2006) 503 508

507

associated with both thyroid disease and major depression prevalence, it is not surprising that sex-adjusted estimates of the strength of association become considerably weaker. The findings for panic disorder/agoraphobia were complex. The results were not readily interpretable because of a Thyroid ConditionAgeSex interaction (Wald = 12.8, P b.01). The occurrence of this three-way interaction may have been a spurious result. However, the interaction precluded clear interpretation of the results for this condition, and an adjusted OR is not presented for panic disorder/agoraphobia in Table 2. 4. Discussion The lack of a statistically significant association between self-reported thyroid conditions and the 12-month prevalence of major depression and bipolar disorder after adjustment for age, sex and chronic conditions in the logistic regression analysis should not receive undue interpretive emphasis. The adjusted ORs in each case were slightly greater than 1 and may reflect weak associations that did not achieve statistical significance in this analysis. In a previous study employing a larger sample, an age- and sex-adjusted OR of 1.4 was reported for the association between major depression and self-reported thyroid disease [32]. Although the current study was large and population based, the estimates were not highly precise because of sample-size limitations. Overall, the results presented here seem consistent with the existence of weak associations between self-reported thyroid conditions and a variety of disorders. The 12-month prevalence and the lifetime prevalence of mood and anxiety disorders in people with thyroid disease resemble those associated with other chronic conditions. Logistic regression suggests that after adjustment for demographic variables, the association between thyroid conditions and mood and anxiety disorders is weaker than for most other chronic conditions. This is also the impression gained from previous studies in this literature. Special screening for mental disorders in subjects with a history of thyroid disease is probably not warranted. This conclusion can be illustrated by calculating fitted proportions from the logistic regression models underlying the adjusted estimates presented in Table 2. According to the model that predicts the 12-month prevalence of major depression [log odds of depression = 3.57+0.072 (thyroid disease)0.013 (age)+1.04 (female)+0.81 (other chronic condition)], a 30-year-old man with no thyroid disease and no other chronic condition has a predicted (12-month) major depression prevalence of 2.0%. Having thyroid disease would increase this to 2.1%. A 30-year-old man with no thyroid disease but with another chronic medical condition has a predicted prevalence of 4.3%, and with the addition of thyroid disease, the predicted value becomes 4.6%. Exponentiation of the regression coefficients to yield ORs

confirms that the adjusted OR of 2.2 for other chronic conditions (95% CI= 1.92.7) was higher than that for thyroid conditions (1.1; 95% CI = 0.81.4). This was also the case for each of the logistic regression models, including that for social phobia, where the OR for chronic conditions other than thyroid disease was 1.9 (95% CI = 1.52.3). Much emphasis has been placed on screening of psychiatric patients for occult thyroid disease. This emphasis probably relates to the idea that thyroid disease is a risk factor for psychiatric conditions bdue to a general medical condition.Q The failure of recent epidemiological studies to clearly identify a link between thyroid status and psychiatric symptoms in community-dwelling subjects may mean that thyroid dysfunction is not as important a risk factor as was believed previously, at least not in the context of current health care delivery standards. Thyroid disturbances that are sufficiently severe to cause psychiatric disturbances may now be uncommon in community populations. As such, the evidence reported here and in other recent studies that have failed to find associations between thyroid disorders and mental disorders and symptoms should not be interpreted as a challenge to the reality of biological connections between thyroid dysfunction and mental status but rather as an indication that the degree of control and management of thyroid disease in community populations now means that these conditions are not strongly associated with psychiatric disorder status in those populations. The association of self-reported thyroid disease with symptoms of mood and anxiety, as reported recently [22], might theoretically suggest an altered emphasis on the screening or monitoring of subjects with a history of thyroid disease irrespective of their biochemical thyroid status. This study, however, did not find support for this idea. Acknowledgment Dr. Patten is a Health Scholar with the Alberta Heritage Foundation for Medical Research and a Fellow with the Institute for Health Economics. This study was funded by an operating grant from the Research Coordinating Committee of the Institute of Health Economics. Dr. Beck is a Fellow with the Canadian Institutes for Health Research and a Clinical Fellow with the Alberta Heritage Foundation for Medical Research.

References
[1] Whybrow PC, Prange AJ, Treadway CR. Mental changes accompanying thyroid gland dysfunction. Arch Gen Psychiatry 1969;20: 48 63. [2] American Psychiatric Association. Diagnostic and statistical manual of mental disorders (DSM-IV-TR). Washington7 American Psychiatric Association; 2000. [3] Bunevicius R, Velickiene D, Prange Jr AJ. Mood and anxiety disorders in women with treated hyperthyroidism and ophthalmopathy caused by Graves disease. Gen Hosp Psychiatry 2005;27:133 9.

508

S.B. Patten et al. / General Hospital Psychiatry 28 (2006) 503 508 disorder in patients with abnormal thyroid function tests. Clin Endocrinol 1996;45:215 23. Bono G, Fancellu R, Blandini F, Santoro G, Maun M. Cognitive and affective status in mild hypothyroidism and interactions with l-thyroxine treatment. Acta Neurol Scand 2004;110:59 66. Hamilton M. Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 1967;6:278 96. Gulseren S, Gulseren L, Hekimsoy Z, Cetinay P, Ozen C, Tokatlioglu B. Depression, anxiety, health-related quality of life and disability in patients with overt and subclinical thyroid dysfunction. Arch Med Res 2006;37:133 9. Engum A, Bjbro T, Mykletun A, Dahl AA. An association between depression, anxiety and thyroid function a clinical fact or an artifact? Acta Psychiatr Scand 2002;106:27 34. Engum A, Bjbro T, Mykletun A, Dahl AA. Thyroid autoimmunity, depression and anxiety; are there any connections? An epidemiological study of a large population. J Psychosom Res 2005;59:263 8. Grabe HJ, Vflzke H, Lqdemann J, Wolff B, Schwahn C, John U, et al. Mental and physical complaints in thyroid disorders in the general population. Acta Psychiatr Scand 2005;112:286 93. van Boxtel MPJ, Menheere PPCA, Bekers O, Hogervorst E, Jolles J. Thyroid function, depressed mood, and cognitive performance in older individuals: the Maastricht Aging Study. Psychoneuroendocrinology 2004;29:891 8. Forman-Hoffman V, Philibert RA. Lower TSH and higher T4 levels are associated with current depressive syndrome in young adults. Acta Psychiatr Scand (On Line Early) 2006;114:132 9. Thomsen AF, Kessing LV. Increased risk of hyperthyroidism among patients hospitalized with bipolar disorder. Bipolar Disord 2005;7: 351 7. Thomsen AF, Kvist TK, Andersen PK, Kessing LV. Increased risk of developing affective disorder in patients with hypothyroidism: a register-based study. Thyroid 2005;15:700 7. Gravel R, Be land Y. The Canadian community health survey: mental health and well-being. Can J Psychiatry 2005;50:573 9. Kessler RC, Ustun TB. The World Mental Health (WMH) Survey Initiative Version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI). Int J Methods Psychiatr Res 2004;13:83 121. Statistics Canada. Canadian Community Health Survey (CCHS) Cycle 1.2 Questionnaire. http://www.statcan.ca/english/concepts/ health/cycle1_2/content.htm. Patten SB, Beck CA, Kassam A, Williams JVA, Barbui C, Metz LM. Long-term medical conditions and major depression: strength of association for specific conditions in the general population. Can J Psychiatry 2005;50:195 202.

[4] LeCrubier Y, Sheehan DV, Weiller E, Amorim P, Bonora I, Sheehan KH, et al. The Mini International Neuropsychiatric Interview (MINI). A short diagnostic structured interview: reliability and validity according to the CIDI. Eur Psychiatry 1997;12(5):224 31. [5] Sheehan DV, Lecubier Y, Sheehan KH, Janavs J, Weiller E, Keskiner A, et al. The validity of the Mini International Neuropsychiatric Interview (MINI) according to the SCID-P and its validity. Eur Psychiatry 1997; 12:232 41. [6] Trzepacz PT, McCue M, Klein I, Levey GS, Greenhouse J. A psychiatric and neuropsychological study of patients with untreated Graves disease. Gen Hosp Psychiatry 2006;10:49 55. [7] Endicott J, Spitzer RL. A diagnostic interview: the schedule for affective disorders and schizophrenia. Arch Gen Psychiatry 1978;35: 837 44. [8] Placidi GPA, Boldrini M, Patronelli A, Fiore E, Chiovato L, Perugi G, et al. Prevalence of psychiatric disorders in thyroid diseased patients. Neuropsychobiology 1998;38:222 5. [9] First M, Spitzer R, Gibbon M, Williams J. Structured clinical interview for DSM-IV Axis I disorderspatient edition (SCID-I/P, Version 2.0). New York7 New York State Psychiatric Institute; 1995. [10] Kathol RG, Delahunt JW. The relationship of anxiety and depression to symptoms of hyperthyroidism using operational criteria. Gen Hosp Psychiatry 1986;8:23 8. [11] Bunevicius R, Velickiene D, Prasad A. Mood and anxiety disorders in women with treated hyperthyroidism and ophthalmopathy caused by Graves disease. Gen Hosp Psychiatry 2005;27:133 9. [12] Silverstone PH. Prevalence of psychiatric disorders in medical inpatients. J Neurol Neurosurg Psychiatry 1996;184:43 51. [13] Constant EL, Adam S, Bruyer SR, Seghers A, Daumerie C. Anxiety an depression, attention and executive functions in hypothyroidism. J Int Neuropsychol Soc 2005;11:535 44. [14] Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:53 63. [15] Burmeister LA, Ganguli M, Dodge HH, Toczek T, DeKosky ST, Nebes RD. Hypothyroidism and cognition: preliminary evidence for a specific defect in memory. Thyroid 2001;11:1177 85. [16] Tagay S, Herpertz S, Langkafel M, Erim Y, Freudenberg L, Schfpper N, et al. Health-related quality of life, anxiety and depression in thyroid cancer patients under short-term hypothyroidism and TSH-suppressive levothyroxine treatment. Eur J Epidemiol 2005; 153:755 63. [17] Haggerty JJ, Stern RA, Mason GA, Beckwith J, Morey CE, Prange AJ. Subclinical hypothyroidism: a modifiable risk factor for depression? Am J Psychiatry 1993;150:508 10. [18] Oomen HAPC, Schipperijn AJM, Drexhage HA. The prevalence of affective disorder and in particular of a rapid cycling of bipolar

[19]

[20] [21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29] [30]

[31]

[32]