0021-972X/06/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 91(9):3389 –3393 Copyright © 2006 by The Endocrine Society doi: 10.1210/jc.2006-0414

Psychological Well-Being Correlates with Free Thyroxine But Not Free 3,5,3ⴕ-Triiodothyronine Levels in Patients on Thyroid Hormone Replacement Ponnusamy Saravanan, Theo J. Visser, and Colin M. Dayan Henry Wellcome Laboratories for Integrative Neuroscience and Endocrinology (P.S., C.M.D.), University of Bristol, Bristol BS1 3NY, United Kingdom; and Department of Internal Medicine (T.J.V.), Erasmus University Medical Centre, 3000 DR Rotterdam, The Netherlands Context and Objective: An association between mood disorders and overt thyroid dysfunction is well established, but there are few data on the potential for thyroid hormone levels closer to the reference range to correlate with psychological well-being. Design, Setting, and Patients: We analyzed the relationship between psychological well-being and free T4 (fT4), free T3 (fT3), TSH, and total rT3 in 697 patients on thyroid hormone replacement therapy at entry to a randomized, controlled trial of combined T4 and T3 replacement therapy. All patients were on 100 ␮g or more T4. Interventions and Main Outcome Measures: Psychological wellbeing was assessed with General Health Questionnaire-12 (GHQ-12), Thyroid Symptom Questionnaire, and Hospital Anxiety and Depression Scale.

T

HE OCCURRENCE OF various abnormalities in brain function including cognitive and memory impairment in patients with overt hypothyroidism is now well established. Reduced levels of thyroid hormone appear to slow serotonergic neurotransmission in the brain (1), an effect associated with low mood. In addition, thyroid hormones are widely used to augment antidepressant treatment (2), although the trial evidence underlying this is controversial (3). Evidence for lesser variation in thyroid hormone levels affecting mood and psychological well-being remains more controversial. Some cross-sectional studies suggested that subclinical thyroid dysfunction is associated with depression, cognitive impairment, and memory loss (reviewed in Ref. 4), and Carr et al. (5) reported that patients receiving thyroid hormone replacement appeared more content on higher doses of T4. The large HUNT (Nørd-Trondelag Health Study) community-based study failed to find an association, but the correlations were made with categories of TSH level rather than using T4 and TSH, as continuous variables. Interestingly, in the subgroup of patients on T4, a link with depression was reported (6). Recently several studies of thyFirst Published Online June 27, 2006 Abbreviations: b, Coefficient beta; fT3, free T3; fT4, free T4; GHQ, General Health Questionnaire; HADS, Hospital Anxiety and Depression Scale; NR, normal range; OR, odds ratio; TPO, thyroid peroxidase; TSQ, Thyroid Symptom Questionnaire; WATTS, Weston Area T3/T4 Study. JCEM is published monthly by The Endocrine Society (http://www. endo-society.org), the foremost professional society serving the endocrine community.

Results: fT4 and TSH showed a strong correlation with GHQ-12 scores (fT4 ⫺ b: ⫺0.16, P ⫽ 0.005; TSH ⫺ b: 0.663, P ⫽ 0.04). No correlations were seen between the GHQ scores and fT3 (b: 0.318, P ⫽ 0.275), rT3 (b: 0.095, P ⫽ 0.95), rT3 to fT4 ratio (b: 71.83, P ⫽ 0.09) or fT3 to rT3 ratio (b: 0.05, P ⫽ 0.32). The correlations remained when the data set was limited to patients with TSH in the range 0.3– 4.0 mIU/liter. Similar correlations were seen with the Thyroid Symptom Questionnaire, although not with the Hospital Anxiety and Depression Scale scores. Conclusions: Differences in fT4 and TSH concentration, even within the reference range, may be a determinant of psychological well-being in treated hypothyroid patients although not necessarily with symptoms typical of anxiety or depression. (J Clin Endocrinol Metab 91: 3389 –3393, 2006)

roid hormone replacement therapy reported that the combination of T4 and T3 is not superior to T4 alone (7, 8). However, where thyroid hormone levels were raised, psychological well-being appeared to have improved (9). In view of the large body of circumstantial evidence linking thyroid hormone levels and mood and the relative stability of endogenous thyroid hormone levels within a given individual over time (10, 11), we hypothesized that variation in thyroid hormone levels, even within the laboratory reference range, might represent an independent risk factor for low mood and depression. To test this hypothesis, we examined the relationship between thyroid hormone parameters and psychological well-being across a large cohort of patients treated with T4 who were about to take part in a randomized, controlled trial of different forms of thyroid hormone treatment (12). Subjects and Methods Baseline data were obtained from subjects recruited to a randomized, controlled trial studying the effects of combined T3/T4 therapy vs. T4 alone for treating hypothyroidism [Weston Area T3/T4 Study (WATTS)] (12). The study was approved by the local research ethics committee. All patients provided written, informed consent. Data from 697 patients in WATTS were available for analysis. Entry criteria for the WATTS have previously been published (12) and included age between 18 and 75 yr and T4 replacement at a dose of 100 ␮g/d or more, unchanged for a minimum of 3 months and with a TSH level reported to be in the laboratory reference range in the preceding 15 months. At study entry to WATTS, blood was drawn (untimed in relation to dose) and stored for measurement of thyroid hormones at the end of the study: serum TSH [normal range (NR) 0.3– 4.0 mIU/ml], free

3389

3390

J Clin Endocrinol Metab, September 2006, 91(9):3389 –3393

Saravanan et al. • Correlation of Thyroid Hormones and Well-Being

T4 (fT4) [NR 0.78 –1.86 ng/dl (10.0 –24.0 pmol/liter)], and anti-thyroid peroxidase (TPO) antibodies (positive if titer is ⱖ 100) were measured by RIA (Diagnostic Product Corp., Los Angeles, CA). Serum free T3 (fT3) [NR 0.18 – 0.457 ng/dl (2.8 –7.1 pmol/liter)] was measured by chemiluminescence assay (Elecsys system 1010; Roche Diagnostics, Mannheim, Germany). Serum rT3 was measured by an in-house RIA [NR 9 –22 ng/dl (0.14 – 0.34 nmol/liter)] (13). All the samples were analyzed as a single batch at the end of the study. Serum from 637 patients was available for the rT3 assays.

Psychological assessment Patients’ well-being was assessed at study entry by the General Health Questionnaire (GHQ)-12, which is a well-validated tool in predicting psychiatric morbidity when compared with complex psychometric tests and detailed psychiatric interview (14). The GHQ-12 has four responses for each question: “better than usual,” “same as usual,” “less than usual,” and “much less than usual.” In addition to the GHQ12, all subjects completed the Hospital Anxiety and Depression Scale (HADS) (15) and an unvalidated questionnaire, the Thyroid Symptom Questionnaire (TSQ) based on symptoms frequently reported by patients on thyroid hormone (16). The TSQ responses were developed in a similar way to the GHQ-12. The GHQ-12 and TSQ were scored by both the Likert method [0 –3 per question, maximum score 36 (most dissatisfied), linear method] and the GHQ scoring method [0, 0, 1, 1, maximum score 12 (most dissatisfied), categorical method]. A score of 3 or greater by the GHQ scoring method is taken to indicate caseness, as when the GHQ was compared with complex psychiatric interview; such scores have been found to be strongly predictive of a psychiatric diagnosis being confirmed at interview (14). The HADS questionnaire was also scored by both the linear and categorical scoring method (15).

Statistical analyses All the analyses were conducted in Stata version 8.0 (17). The relationships between psychological questionnaire scores and serum thyroid hormone measurements were ascertained using linear and logistic regression analyses for continuous and binary versions of the questionnaires, respectively. Multiple regression analysis was used when adjusting for age, sex, and anti-TPO antibody positivity. The results of the linear regression analyses were reported as correlation coefficient beta (b) and results of logistic regression analyses as odds ratio (OR). All correlations were reported with the whole data set and again with the restricted data set of subjects with TSH levels in the range of 0.3– 4.0 mIU/liter at the time of psychological testing (n ⫽ 473). ␹2 test and ANOVA were used to ascertain the relationship between fT4 levels (high and normal) and the binary and continuous versions of the questionnaires, respectively.

Results

The mean age of the patients was 57.3 yr. Eighty-four percent of patients were women (n ⫽ 586). The original causes of hypothyroidism were autoimmune hypothyroidism (73.45%), Graves’ disease (17.07%), toxic multinodular goiter (2.58%), and nontoxic goiter (6.89%). The current diagnoses were autoimmune hypothyroidism (73.45%), postradioactive iodine (9.33%), postthyroidectomy (15.78%), and postthyroidectomy ⫹ postradioactive iodine (1.44%). Forty-four percent of the patients (n ⫽ 307) had a strongly positive titer for anti-TPO an-

tibodies (titer ⬎ 100). Baseline thyroid function at the time of initial psychological testing in all subjects is shown in Table 1, and the distribution of TSH, fT4, fT3, and rT3 values is shown in Fig. 1. Nearly 43% of patients scored 3 or more on the GHQ-12 categorical score, demonstrating increased psychiatric morbidity. This is approximately 18% higher than reported in the general population (18) and 10% higher than seen in our crosssectional study (16) and may reflect selection bias in subjects volunteering for an intervention trial. Ten of the subjects (1.4%) had undetectable levels of TSH (⬍0.01 mIU/liter) and 171 (24.5%) had TSH levels less than 0.3 mIU/liter. Fifty-three (7.6%) of the subjects had TSH levels more than 4.0 mIU/liter; the highest TSH level was 12.9 mIU/liter. Although all subjects had a TSH level within their local laboratory reference range in the 15 months before recruitment, when remeasured at study entry, 32% of subjects had TSH levels outside the range 0.3– 4.0 mIU/liter in the study laboratory. A more sensitive TSH assay and a narrower reference range used for the study might be a contributing factor. Baseline fT4 showed a strong negative correlation to the GHQ-12 scores (correlation coefficient b: ⫺0.155, P ⫽ 0.005). The relationship persisted even after correcting for age, sex, and anti-TPO antibody positivity (b: ⫺0.14, P ⫽ 0.015) and was also present in the subset of patients with TSH between 0.3 and 4.0 mIU/liter (b: ⫺0.159, P ⫽ 0.038, n ⫽ 473). The same correlations were observed when the GHQ was scored as a categorical parameter (GHQ scoring) (Table 2). The relationship was in the expected direction (higher fT4 associated with lower GHQ scores implying improved well-being), and the slope indicated an improvement of 1 GHQ point for a 0.51 ng/dl (6.5 pmol/ liter) rise in fT4. A positive correlation was seen with log TSH and GHQ [b: 0.66, P ⫽ 0.04; no change after controlling for age, sex, and anti-TPO antibody positivity (b: 0.68, P ⫽ 0.04)]. This correlation was preserved in the subset of patients with TSH in the range 0.3– 4.0 mIU/liter (b: 2.3, P ⫽ 0.006). In contrast, no correlation was seen among fT3, rT3, rT3 to T4 and T3 to rT3 ratios, and anti-TPO positivity and GHQ scores in either the full data set or the subset. A significant relationship was seen with fT3/fT4 ratio, but this was due to the contribution of fT4 and not an independent effect (Table 2). Similar results were observed with the TSQ. fT4 showed significant correlation with both the linear (correlation coefficient b: ⫺ 0.11, P ⫽ 0.03) and categorical scores of TSQ, and this persisted in the TSH 0.3– 4.0 mIU/liter subset (Table 3). Whereas no correlation was seen between linear TSQ and log TSH (b: 0.09, P ⫽ 0.41), a relationship was seen between the categorical TSQ and log TSH (OR 1.4, P ⫽ 0.007) but was lost in the TSH 0.3– 4.0 mIU/liter subset. Similar to GHQ, no other correlation was seen between TSQ and any other thyroid hormone parameters. No correlation was seen among any of the

TABLE 1. Baseline thyroid function tests Variable

Mean

SD

Range

NR

Age (yr) (n ⫽ 697) TSH (mIU/liter) (n ⫽ 697) (geometric mean/median) fT4 (ng/dl) (n ⫽ 697) fT3 (ng/dl) (n ⫽ 697) rT3 (ng/dl) (n ⫽ 637)

57.3 0.86 (0.948) 1.63 0.248 26

11.07 1.88 0.28 0.045 8

23.1–76.2 ⬍0.01–12.9 0.91–2.69 0.133– 0.523 9– 63

n/a 0.3– 4.0 0.78–1.86 0.18– 0.457 9–22

For SI units, multiply by 12.87 for T4, 15.55 for T3, and 0.0155 for rT3.

Saravanan et al. • Correlation of Thyroid Hormones and Well-Being

J Clin Endocrinol Metab, September 2006, 91(9):3389 –3393

3391

FIG. 1. Histograms of log TSH (A), fT4 (B), fT3 (C), and total rT3 (D).

thyroid hormone parameters and the anxiety and depression scales of the HADS (data not shown) with the exception of log TSH and HADS depression as a continuous variable (b: 0.562, P ⫽ 0.004). However, this relationship was not seen when HADS depression score was used as a categorical variable (OR 1.2, P ⫽ 0.54). Further post hoc analyses of the subgroup of patients with a

fT4 level above the reference range did not show any correlation between psychological well-being and the fT4 levels in this group using the regression model (GHQ Likert vs. fT4: b, 0.08, P ⫽ 0.70; GHQ Likert vs. log TSH: b, 0.08, P ⫽ 0.90; TSQ Likert vs. fT4: b, 0.21, P ⫽ 0.26; TSQ Likert vs. log TSH: b, ⫺0.99, P ⫽ 0.08; data with categorical scores were not shown). However, the mean GHQ scores are significantly lower (improved well-

TABLE 2. Correlations between GHQ scores and thyroid functions

Variable

fT4 fT3 fT3/fT4 Log TSH rT3 rT3 to fT4 ratio fT3 to rT3 ratio ⫹ Anti-TPO (titer ⬎ 100)

GHQ linear scores b (P value) All patients (n ⫽ 697)

TSH 0.3– 4.0 (n ⫽ 473)

⫺0.155 (0.005)a 0.318 (0.28) 13.38 (0.003)a 0.663 (0.038)a 0.095 (0.95) 71.83 (0.09) 0.05 (0.32) 0.006 (0.32)

⫺0.159 (0.038)a 0.065 (0.86) 9.09 (0.098) 2.3 (0.006)a 0.89 (0.69) 77.75 (0.12) 0.01 (0.84) 0.001 (0.21)

GHQ categorical scores OR (P value) All patients (n ⫽ 693)

0.95 (0.014)a 1.06 (0.58) 58.7 (0.016)a 1.31 (0.027)a 0.412 (0.17) 0.69 (0.98) 1.03 (0.07) 1.0 (0.37)

TSH 0.3– 4.0 (n ⫽ 470)

0.94 (0.037)a 1.03 (0.82) 29.67 (0.10) 2.47 (0.005)a 0.447 (0.09) 14.0 (0.81) 1.03 (0.28) 1.0 (0.41)

Regression coefficient (beta-linear scores) and ORs (categorical scores) for the relationship between GHQ scores and thyroid function parameters. a Significant values (P ⬍ 0.05).

3392

J Clin Endocrinol Metab, September 2006, 91(9):3389 –3393

Saravanan et al. • Correlation of Thyroid Hormones and Well-Being

TABLE 3. Correlations between TSQ scores and thyroid functions TSQ linear scores b (P value)

Variable

All patients (n ⫽ 696)

fT4 fT3 fT3/fT4 Log TSH rT3 rT3 to fT4 ratio fT3 to rT3 ratio ⫹ Anti-TPO (titer ⬎100)

⫺0.108 (0.03)a 0.437 (0.098) 11.85 (0.003)a 0.23 (0.44) 0.99 (0.52) 77.73 (0.04)a 0.02 (0.69) 0.0007 (0.19)

TSQ categorical scores OR (P value) TSH 0.3– 4.0 (n ⫽ 472)

⫺0.182 (0.007)a 0.21 (0.52) 11.24 (0.02)a 1.49 (0.04)a 1.26 (0.53) 89.39 (0.049)a ⫺0.02 (0.72) 0.0008 (0.21)

All patients (n ⫽ 691)

0.95 (0.028)a 0.88 (0.27) 4.86 (0.37) 1.4 (0.007)a 0.92 (0.90) 4.31 (0.24) 0.99 (0.50) 1.0 (0.06)

TSH 0.3– 4.0 (n ⫽ 469)

0.94 (0.058) 0.92 (0.58) 5.75 (0.42) 1.8 (0.07) 1.74 (0.53) 8.79e⫹13 (0.12) 0.97 (0.22) 1.0 (0.17)

Regression coefficient (beta-linear scores) and ORs (categorical scores) for the relationship between TSQ scores and thyroid function parameters. a Significant values (P ⬍ 0.05).

being) in this group as a whole, compared with the subgroup of patients with levels of thyroid hormones in the reference range (high fT4 vs. normal fT4: 12.33 ⫾ 4.79 vs. 13.72 ⫾ 5.45, P ⫽ 0.007). By the categorical scoring method, the percentage of caseness was also less in patients with high fT4 levels (35.1 vs. 45.3%, P ⫽ 0.03). Similar results were seen in TSQ scores (linear TSQ scores: high fT4 vs. normal fT4: 13.91 ⫾ 4.66 vs. 14.85 ⫾ 4.88, P ⫽ 0.04; percent caseness: high fT4 vs. normal fT4: 56.1 vs. 66.0%, P ⫽ 0.03) Similar post hoc analysis of patients according to anti-TPO antibody status did not show any significant difference in GHQ between anti-TPO-positive and negative patients (anti-TPO positive vs. anti-TPO negative: 13.71 ⫾ 5.42 vs. 13.12 ⫾ 5.26, P ⫽ 0.147). Similar results were obtained when antibody status was used as an interaction factor in the regression model. Discussion

This is the first large data set to explore the relationship between fT4, fT3, and rT3 and psychological well-being in subjects on thyroid hormone replacement. Improved psychological well-being was found to correlate with higher fT4 levels. The significance of the observation is supported by the finding of a relationship between psychological well-being and TSH with the opposite slope (higher TSH with reduced well-being as might be expected). A similar relationship with fT4 was found with an unvalidated score of symptoms that relates more directly to thyroid status (TSQ) making an artifactual association due to multiple testing less likely, although still possible. These relationships also persisted in the subset of patients with TSH values in the reference range (0.3– 4.0 mIU/liter). Interestingly, no clear association was seen with the HADS scale, which may suggest that the thyroid function influences parameters of psychological well-being not typical of anxiety or depression. The GHQ-12 refers more generally to psychological well-being comparing current status with how patients would usually expect to feel (14). A previous Norwegian study (the HUNT study) also showed no relationship between thyroid function and HADS ratings except in individuals who are already on T4 replacement (6). It should be noted, however, that the subjects in the HUNT study were grouped according to TSH, and T4 was measured only when the TSH was outside of the range 0.2– 4.0 mIU/liter. TSH is often considered the most sensitive measure of thyroid function. However, it appears that the relationship be-

tween well-being and fT4 was as much if not more pronounced as with TSH (Table 2), especially when studied across the whole cohort. TSH levels reflect hypothalamopituitary sensing of circulating thyroid hormone levels, which may be different from thyroid hormone status in other tissues and the importance of fT4 measurement in addition to or distinct from serum TSH estimation to assess thyroid status has been emphasized in recent studies of pregnancy. Maternal hypothyroxinemia in the first trimester and not raised TSH was associated with impaired psychomotor development in offspring (19, 20), and a recent study from the northeast of England showed that fT4 but not TSH at 9 wk of pregnancy is directly proportional to the birth weight of the offspring (Vaidya B., personal communication). An association between fT4 but not TSH at booking visit (mean gestation 13.05 wk) and fetal birth length and head circumference has also been reported (21). Very recently Wekking et al. (22) failed to find a correlation between TSH and either cognitive function or psychological well-being in patients on T4, but this data set was relatively small (n ⫽ 141). TSH was only correlated as greater than or less than 2.0 mIU/liter (not as a continuous variable), and the relationship with fT4 was not reported. It should be noted that thyroid function testing was not timed with T4 dosing in our study. Although this is a limitation, if anything, this would be expected to reduce rather than augment the significance of any correlations with thyroid function. The failure to find a relationship between serum fT3 and GHQ/TSQ scores is also of interest. Many thyroidologists consider the T3 assay to be less technically reliable and less reflective of thyroid status, particularly in the hypothyroid range (23). Although T3 is the active hormone, free concentrations of T4 are five times higher, and many tissues obtain 30% or more of their intracellular T3 directly from circulating T4 (24). Hence, circulating T3 levels may not be directly reflective of intracellular levels. We measured serum rT3 levels as a possible measure of intracellular deiodinase activity (24, 25). The failure of rT3 levels or ratios with thyroid hormones to correlate with psychological well-being might relate to serum levels being more indicative of hepatic type 1 (D1) and 2 (D2) iodothyronine deiodinase activity, whereas intracellular levels are strongly influenced by local levels of membrane-bound deiodinases including type 2 and 3 (D3) iodothyronine deiodinase (24, 26). Our findings provide some support for the view that serum fT4 levels as well as TSH levels should be taken into account

Saravanan et al. • Correlation of Thyroid Hormones and Well-Being

when adjusting dosages and that TSH may not be a perfect indicator of the adequacy of replacement, particularly with regard to psychological well-being (27). In support of this, data from Carr et al. (5) as well as recent data from Appelhof et al. (9) suggest that patients prefer dosages of thyroid hormone that result in suppression of TSH. In addition, it is possible that the increased levels of psychological morbidity reported in patients on doses of T4 adjusted to normal TSH values might also relate to this (16). However, this is a complex area because there is increasing evidence that suppressive doses of T4 can be associated with adverse effects on both bone metabolism and the heart (reviewed in Ref. 25), and the current recommendation remains to titrate T4 dosages to TSH levels in the reference range where possible (23, 28). Our data relate only to patients on thyroid hormone replacement. Additional adequately powered independent studies with appropriate questionnaires and large populations of patients on T4 (and preferably not just patients selected to take part in an intervention trial) are required to confirm our findings. It is possible that in the general population without thyroid dysfunction, variation in thyroid hormone levels across the reference range is also a determinant of psychological well-being. Such a consequence of interindividual variation in normal levels of a hormone would be similar to the relationship observed between IGF-I levels and the risk of cancer across the reference range for IGF-I (29). Large population-based studies of thyroid function parameters including fT4 and psychological wellbeing will be required to explore this. Acknowledgments

J Clin Endocrinol Metab, September 2006, 91(9):3389 –3393

8.

9.

10. 11. 12.

13. 14. 15. 16.

17. 18. 19.

20.

The authors thank Dr. B. Vaidya for the use of his unpublished data. 21.

Received February 22, 2006. Accepted June 19, 2006. Address all correspondence and requests for reprints to: Dr. Colin M. Dayan, Henry Wellcome Laboratories for Integrative Neuroscience and Endocrinology, Dorothy Hodgkin Building, University of Bristol, Whitson Street, Bristol BS1 3NY, United Kingdom. E-mail: [email protected]. This work was supported by Southwest National Health Service Research and Development, United Kingdom, and Goldshield Pharmaceuticals PLC, United Kingdom.

References 1. Bauer M, Heinz A, Whybrow PC 2002 Thyroid hormones, serotonin and mood: of synergy and significance in the adult brain. Mol Psychiatry 7:140 –156 2. Altshuler LL, Bauer M, Frye MA, Gitlin MJ, Mintz J, Szuba MP, Leight KL, Whybrow PC 2001 Does thyroid supplementation accelerate tricyclic antidepressant response? A review and meta-analysis of the literature. Am J Psychiatry 158:1617–1622 3. Appelhof BC, Brouwer JP, van Dyck R, Fliers E, Hoogendijk WJG, Huyser J, Schene AH, Tijssen JGP, Wiersinga WM 2004 Triiodothyronine addition to paroxetine in the treatment of major depressive disorder. J Clin Endocrinol Metab 89:6271– 6276 4. McDermott MT, Ridgway EC 2001 Subclinical hypothyroidism is mild thyroid failure and should be treated. J Clin Endocrinol Metab 86:4585– 4590 5. Carr D, McLeod DT, Parry G, Thornes HM 1988 Fine adjustment of thyroxine replacement dosage: comparison of the thyrotrophin releasing hormone test using a sensitive thyrotrophin assay with measurement of free thyroid hormones and clinical assessment. Clin Endocrinol (Oxf) 28:325–333 6. Engum A, Bjoro T, Mykletun A, Dahl AA 2002 An association between depression, anxiety and thyroid function—a clinical fact or an artefact? Acta Psychiatr Scand 106:27–34 7. Escobar-Morreale HF, Botella-Carretero JI, Escobar del Rey F, Morreale de

22.

23.

24. 25. 26. 27.

28.

29.

3393

Escobar G 2005 Review: treatment of hypothyroidism with combinations of levothyroxine plus liothyronine. J Clin Endocrinol Metab 90:4946 – 4954 Grozinsky-Glasberg S, Fraser A, Nahshoni E, Weizman A, Leibovici L 2006 Thyroxine-triiodothyronine combination therapy versus thyroxine monotherapy for clinical hypothyroidism: meta-analysis of randomized controlled trials. J Clin Endocrinol Metab 91:2592–2599 Appelhof BC, Fliers E, Wekking EM, Schene AH, Huyser J, Tijssen JG, Endert E, van Weert HC, Wiersinga WM 2005 Combined therapy with levothyroxine and liothyronine in two ratios, compared with levothyroxine monotherapy in primary hypothyroidism: a double-blind, randomized, controlled clinical trial. J Clin Endocrinol Metab 90:2666 –2674 Andersen S, Pedersen KM, Bruun NH, Laurberg P 2002 Narrow individual variations in serum T(4) and T(3) in normal subjects: a clue to the understanding of subclinical thyroid disease. J Clin Endocrinol Metab 87:1068 –1072 Dayan CM, Saravanan P, Bayly G 2002 Whose normal thyroid function is better—yours or mine? Lancet 360:353 Saravanan P, Simmons DJ, Greenwood R, Peters TJ, Dayan CM 2005 Partial substitution of thyroxine (T4) with tri-iodothyronine in patients on T4 replacement therapy: results of a large community-based randomized controlled trial. J Clin Endocrinol Metab 90:805– 812 van den Beld AW, Visser TJ, Feelders RA, Grobbee DE, Lamberts SWJ 2005 Thyroid hormone concentrations, disease, physical function, and mortality in elderly men. J Clin Endocrinol Metab 90:6403– 6409 Goldberg D, Williams P 1988 A user’s guide to the General Health Questionnaire. Windsor, UK: Nfer-Nelson Bjelland I, Dahl AA, Haug TT, Neckelmann D 2002 The validity of the Hospital Anxiety and Depression Scale. An updated literature review. J Psychosom Res 52:69 –77 Saravanan P, Chau WF, Roberts N, Vedhara K, Greenwood R, Dayan CM 2002 Psychological well-being in patients on ‘adequate’ doses of l-thyroxine: results of a large, controlled community-based questionnaire study. Clin Endocrinol (Oxf) 57:577–585 StatCorp 2003 Stata statistical software. release 8.0. College Station, TX: StatCorp. Pevalin DJ 2000 Multiple applications of the GHQ-12 in a general population sample: an investigation of long-term retest effects. Soc Psychiatry Psychiatr Epidemiol 35:508 –512 Pop VJ, Kuijpens JL, van Baar AL, Verkerk G, van Son MM, de Vijlder JJ, Vulsma T, Wiersinga WM, Drexhage HA, Vader HL 1999 Low maternal free thyroxine concentrations during early pregnancy are associated with impaired psychomotor development in infancy. Clin Endocrinol (Oxf) 50:149 –155 Morreale de Escobar G, Obregon MJ, Escobar del Rey F 2000 Is neuropsychological development related to maternal hypothyroidism or to maternal hypothyroxinemia? J Clin Endocrinol Metab 85:3975–3987 Hindmarsh P, Franklyn JA, Clark P, Geary M, Rodeck C, Kilby MD 2004 The relationship between maternal thyroid status in the antenatal period and newborn growth measurements: a cohort study. Endocrine Abstracts, p 133 (http://www.endocrine-abstracts.org) Wekking EM, Appelhof BC, Fliers E, Schene AH, Huyser J, Tijssen JG, Wiersinga WM 2005 Cognitive functioning and well-being in euthyroid patients on thyroxine replacement therapy for primary hypothyroidism. Eur J Endocrinol 153:747–753 Spencer CA, Demers LM 2002 Laboratory support for the diagnosis and monitoring of thyroid disease. National Academy of Clinical Biochemistry– laboratory medicine practice guidelines. http://www.nacb.org/lmpg/ thyroid_lmpg_pub.stm Bianco AC, Salvatore D, Gereben B, Berry MJ, Larsen PR 2002 Biochemistry, cellular and molecular biology, and physiological roles of the iodothyronine selenodeiodinases. Endocr Rev 23:38 – 89 Saravanan P, Dayan CM October 2004 Understanding thyroid hormone action and the effects of thyroid hormone replacement—just the beginning not the end. www.hotthyroidology.com Luiza Maia A, Kim BW, Huang SA, Harney JW, Larsen PR 2005 Type 2 iodothyronine deiodinase is the major source of plasma T3 in euthyroid humans. J Clin Invest 115:2524 –2533 Meier C, Trittibach P, Guglielmetti M, Staub JJ, Muller B 2003 Serum thyroid stimulating hormone in assessment of severity of tissue hypothyroidism in patients with overt primary thyroid failure: cross sectional survey. BMJ 326: 311–312 Vanderpump MP, Ahlquist JA, Franklyn JA, Clayton RN 1996 Consensus statement for good practice and audit measures in the management of hypothyroidism and hyperthyroidism. The Research Unit of the Royal College of Physicians of London, the Endocrinology and Diabetes Committee of the Royal College of Physicians of London, and the Society for Endocrinology. BMJ 313:539 –544 Holly JM, Gunnell DJ, Davey Smith G 1999 Growth hormone, IGF-I and cancer. Less intervention to avoid cancer? More intervention to prevent cancer? J Endocrinol 162:321–330

JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.