| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institute of Community Medicine (H.V., N.F., S.S., R.H.), Department of Gastroenterology, Endocrinology, and Nutrition (S.S., R.H., H.W.), Institute of Clinical Chemistry and Laboratory Medicine (J.L., M.N.), and Clinic of Internal Medicine B (M.D.), University of Greifswald, D-17487 Greifswald, Germany; and Department of Endocrinology (G.B.), Christie Hospital, University of Manchester, Manchester M20 4BX, United Kingdom
Address all correspondence and requests for reprints to: PD Dr. Med. Henry Völzke, Department of Epidemiology and Social Medicine, Ernst Moritz Arndt University, Walther Rathenau Str. 48, D-17487 Greifswald, Germany. E-mail: voelzke{at}uni-greifswald.de.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Design: This was a cross-sectional Study of Health in Pomerania.
Setting: The study was conducted in the general population of northeast Germany.
Subjects: The study population comprised 3662 subjects (1746 women) without history of thyroid disorders.
Interventions: No interventions have been performed.
Main Outcome Measures: Goiter and thyroid nodules were determined by ultrasound. Serum TSH levels less than 0.25 mIU/liter were considered decreased.
Results: Adjusted for major confounders and risk factors for thyroid disorders, subjects with serum IGF-I levels above the upper tertile had higher odds for goiter relative to subjects with serum IGF-I levels below the lower tertile [odds ratio (OR) 1.67; 95% confidence interval (CI) 1.24–2.26 in women; OR 2.04; 95% CI 1.55–2.68 in men]. A similar association was present for thyroid nodules in men (OR 1.64; 95% CI 1.17–2.32) and for decreased serum TSH levels in women (OR 1.65; 95% CI 1.00–2.69). Serum IGFBP-3 levels were not associated with thyroid disorders and did not represent effect modifiers for the association between serum IGF-I levels and the endpoints.
Conclusions: We conclude that high serum IGF-I levels are associated with goiter. Whereas high serum IGF-I levels are also related to thyroid nodules in men, they are related to decreased serum TSH levels in women. Serum IGFBP-3 and TSH levels did not modulate these associations.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
IGF-I mediates the effects of GH. The majority of circulating IGF-I is bound to IGF binding protein (IGFBP)-3, which prevents IGF-I from degradation, exhibits direct effects, or context-dependently modulates the effects of IGF-I (8). Thus, IGFBP-3 induces tissue-specific organomegaly when ubiquitously overexpressed in transgenic mice (9). IGFBP-3 might also modulate IGF-I actions on thyroid growth. Similarly, TSH might modify thyroid-specific effects of IGF-I. A clinical study (10) evaluating the associations between GH administration and thyroid size as a function of serum TSH levels in hypopituitary patients indicated that IGF-I does not independently stimulate thyroid growth but enhances proliferation of thyroid cells by potentiating mitogenic effects of TSH.
Whereas evidence is compelling that organ-specific IGF-I has substantial effects on thyroid morphology and function, there is current controversy on whether serum IGF-I levels are associated with thyroid disorders. For example, one study demonstrated an association between serum IGF-I levels and multinodular goiter in acromegaly patients (7), but other studies did not find such an association (11). Likewise, studies in children yielded conflicting results. In one study (12), children with goiter had higher serum IGF-I levels than children without goiter, but also this finding has not been confirmed in another children population (13). Whereas therapy with recombinant human GH leads, at least in women, to suppressed serum TSH levels (14), no association between serum IGF-I and TSH levels was found in obese patients (15).
Currently data on the possible association between serum IGF-I levels and goiter or thyroid nodules in unselected adult populations are not available. The aims of the present study were: 1) to investigate possible associations between serum IGF-I levels and thyroid disorders, including goiter, nodules, and decreased serum TSH levels; and 2) to analyze the role of serum IGFBP-3 and TSH levels for these associations. For this, we used data from the large-scale population-based Study of Health in Pomerania (SHIP).
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SHIP is a cross-sectional population-based survey in West Pomerania, a region in the northeastern part of Germany. The study region is a previously iodine-deficient area with a high prevalence of iodine deficiency-related disorders such as goiter, thyroid nodules, and decreased serum TSH levels (16). Study details are given elsewhere (17, 18). In brief, the total population comprised 212,157 inhabitants. A sample from the population aged 20–79 yr was drawn. The sample was selected using population registries, in which all German inhabitants are registered. Only individuals with German citizenship and principal residency in the study area were included. The net sample (after exclusion of migrated or deceased persons) comprised 6267 eligible subjects who received a maximum of three written invitations. In case of nonresponse, letters were followed by phone calls or home visits if contact by phone was not possible. The SHIP population finally comprised 4310 participants (2193 women), corresponding to a final response of 68.8%. The study was reviewed by a board of independent scientists and approved by the Ethics Committee of the University of Greifswald. All participants gave written informed consent.
There were 423 subjects (341 women) with known self-reported thyroid disease or current thyroid-related medication according to the anatomical-therapeutical-chemical coding category H03. In addition, 11 pregnant women, 24 subjects (eight women) with missing data on thyroid volume, and 190 subjects (87 women) with no blood drawn or missing data on serum IGF-I for other reasons were excluded from analyses. Thus, the final study population comprised 3662 subjects (1746 women). Serum IGF-I levels did not differ between subjects with and without history of thyroid disease.
Measurements
Sociodemographic and medical characteristics and in females information regarding pregnancies, births, and lifetime use of oral contraceptives and menopausal hormone therapy were assessed by computer-assisted personal interviews. As to smoking habits, subjects were categorized into current, former, and never-smokers. Diabetes mellitus was defined by self-report. Height and weight were measured for the calculation of the body mass index [BMI = weight (kilograms)/height2 (square meters)]. Overweight was defined as BMI of 25 kg/m2 or greater and obesity as BMI of 30 kg/m2 or greater .
Nonfasting blood samples were drawn from the cubital vein in the supine position between 0700 and 1600 h. The two analytical laboratories involved in this study participated every 3 months in the official national German tests for quality assurance. In addition, control samples were daily analyzed for internal quality assurance. Serum TSH levels were measured by immunochemiluminescent procedures (Byk Sangtec Diagnostica GmbH, Frankfurt, Germany). The functional sensitivity of the TSH assay was 0.02 mIU/liter. Decreased serum TSH levels were defined using the lower TSH reference limit (0.25 mIU/liter) that was recently established for this study region (19). Serum IGF-I and IGFBP-3 levels were determined by automated two-site chemiluminescence immunoassays (Nichols Advantage; Nichols Institute Diagnostica GmbH, Bad Vilbel, Germany). All serum samples were acidified to separate IGF-I from IGFBPs. The analytical sensitivity of the IGF-I assay was 6 ng/ml. The IGF-I assay has been calibrated against the World Health Organization international reference reagent 1988, IGF-I 87/518. The analytical sensitivity of the IGFBP-3 assay was 20 ng/ml. The assay reference standard was analytically prepared with glycosylated recombinant human IGFBP-3. Only one lot of reagents was used for all IGFBP-3 measurements.
Thyroid ultrasonography was performed using an Ultrasound VST-Gateway with a 5-MHz linear array transducer (Diasonics, Santa Clara, CA). Nodular changes exceeding 10 mm in diameter are defined as nodules. Thyroid volume was calculated as length x width x depth x 0.479 (milliliters) for each lobe (20). The intra- and interobserver reliabilities were assessed before the start of the study and afterward semiannually during the study. All measurements of the thyroid volume showed Spearman correlation coefficients of greater than 0.85 and mean differences (± 2 SD) of the mean bias of less than 5% (<25%) (21). Goiter was defined as thyroid volume greater than 18 ml in women and greater than 25 ml in men (22). This definition is widely used in epidemiological research (16, 23).
Statistical analysis
Data on quantitative characteristics are expressed as median (25th and 75th percentiles). Data on qualitative characteristics are expressed as percent values or absolute numbers as indicated. Multivariable statistical analyses were performed using logistic regression analysis. Odds ratios (OR) and its 95% confidence interval (95% CI) are given. We incorporated two-way interaction effects of serum IGF-I with IGFBP-3 and TSH levels into the models adjusted for variables as indicated. A value of P < 0.05 was considered statistically significant. All statistical analyses were performed with SPSS software, version 14.0.1 (SPSS GmbH Software, Munich, Germany).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
). Women with high serum IGF-I levels more often used oral contraceptives but less often menopausal hormone therapy and reported less commonly at least one birth than women with low serum IGF-I levels. Regarding thyroid-related characteristics, groups did not differ with respect to urinary iodine concentrations and proportions of subjects with goiter. Whereas women with higher serum IGF-I levels had less commonly thyroid nodules than women with lower serum IGF-I levels, no such differences were observed in men. Subjects with high serum IGF-I levels had slightly higher serum TSH levels (borderline significant in women, P = 0.06), but proportions of subjects with decreased serum TSH levels were similar among groups (Table 1).
|
). Both women and men with serum IGF-I levels above the upper tertile exhibited the highest odds for goiter, but this was more pronounced in men, compared with women. The relation between serum IGF-I levels and goiter was not substantially affected by further adjustment for potential confounders and important determinants of goiter. According to the full model, women with high serum IGF-I levels had an OR for goiter of 1.64 (95% CI 1.22; 2.21), compared with women with low serum IGF-I levels. The respective OR in men was 2.03 (95% CI 1.55; 2.67) (Table 2). Analyses were repeated using thyroid volume exceeding the sex-specific 75th percentile (>20 ml in women, >29 ml in men) as an alternative cutoff for thyroid enlargement (Table 3). These analyses yielded results similar to those reported in Table 2.
|
|
).
|
). Consistent overall statistical models, women with high serum IGF-I levels had the highest odds for decreased serum TSH levels, compared with women with low serum IGF-I levels [OR in the final model 1.75 (95% CI 1.02; 3.00)]. In men, no such differences were present [P for trend = 0.273, OR in the final model 1.09 (95% CI 0.69; 1.72), Table 5].
|
were further adjusted for diabetes and urinary iodine concentrations. All aforementioned associations between serum IGF-I levels and thyroid disorders were also stable over these analyses (Fig. 1). Thus, subjects with serum IGF-I levels above the upper tertile had higher odds for goiter relative to subjects with serum IGF-I levels below the lower tertile (OR 1.67; 95% CI 1.24–2.26 in women; OR 2.04; 95% CI 1.55–2.68 in men), and a similar association was present for thyroid nodules in men (OR 1.64; 95% CI 1.17–2.32) and for decreased serum TSH levels in women (OR 1.65; 95% CI 1.00–2.69).
|
). As expected, serum TSH levels were inversely related to goiter and nodules, but there was no association between serum IGFBP-3 levels and both dependent variables and decreased serum TSH levels. We further analyzed possible effect modifications between serum IGF-I, IGFBP-3, and TSH levels. For this, the three variables were dichotomized using the respective median. No statistically significant interactions were found.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The association between serum IGF-I levels and goiter and, at least in men, thyroid nodules is in line with previous findings reported from children (12) and acromegaly patients (8). Our study, however, is not only confirmatory but adds two novel aspects to current knowledge. First, we were able to demonstrate a dose-response relation between serum IGF-I levels and the risk for thyroid enlargement and nodules. Second, all associations were studied in unselected adults. Our findings are partly in contrast to a study in adult acromegaly patients (6). Although the patients had larger thyroid volumes and more commonly nodules than the controls and also the duration of the disease was associated with thyroid size thereby supporting the findings of our study, there was no association between serum IGF-I levels and goiter within this patient group (6). Conflicting results to our study might be explained by different treatment effects on serum IGF-I levels. Thus, increased thyroid vascularization has only been demonstrated in acromegaly patients with active disease but not in those with stable disease (5). To control for treatment effects in our study, all subjects with previous history of thyroid disease were excluded from the present analyses, and there was no subject with GH treatment or known acromegaly in our study population.
Two other studies (11, 13) are also in contrast to our investigation and did not find an association between serum IGF-I levels and thyroid volume. The first was conducted in 28 acromegaly patients (11) and the second in 64 children (13). Small study populations and selection procedures may explain the differences in the findings between our study and the latter studies. Selection might also explain that one study did not find an association between serum IGF-I and TSH levels in severely obese patients (15).
Our study was performed in a region with previous iodine deficiency. Although adjustment for urinary iodine did not affect the major results, we cannot fully rule out that the specific history of iodine fortification in our study area has contributed to the present findings. Observational studies suggested that insufficient iodine supply decreases serum IGF-I levels in children explaining the negative impact of iodine deficiency on children’s growth (13, 24). A recent population-based, controlled intervention trial demonstrated that treatment of iodine deficiency increases serum IGF-I and IGFBP-3 levels in school-age children (25). The findings presented herein should therefore be confirmed by independent research preferably conducted in regions with stable iodine sufficiency.
Serum IGF-I levels decline, whereas the risk of goiter, nodules, and decreased serum TSH levels in iodine-deficient regions increases with age. Therefore, age represents a major confounder for the investigated associations. Whereas, for example, descriptive statistics revealed subjects with different serum IGF-I levels to have similar frequencies of goiter, multivariable analyses (Table 2) detected the association between serum IGF-I levels and goiter. Furthermore, in women, a phenomenon was observed that is usually referred to as Yule-Simpson paradox (26, 27), i.e. whereas subjects with high serum IGF-I levels had even a tendency toward a lower frequency of goiter than subjects with low serum IGF-I levels in descriptive statistics, the opposite was found after appropriate adjustments had been made for age and further confounders. In other words, appropriate statistical adjustment reveals that high serum IGF-I levels were actually associated with increased rather than decreased risk of goiter.
There were considerable gender differences in our study. Although women and men had similar serum IGF-I levels, the extent of the association between serum IGF-I levels and goiter was more pronounced in men than women, the association between serum IGF-I levels and thyroid nodules was only statistically significant in men, and only women with high serum IGF-I levels more commonly had decreased serum TSH levels than women with low serum IGF-I levels. These findings are in good agreement with other studies that also have demonstrated gender differences in the effects of GH and IGF-I. Thus, GH treatment leads to beneficial effects on lean body mass and insulin sensitivity that are more pronounced in men than women (28, 29). Against this background it might be tempting to hypothesize that, with respect to hypertrophy and hyperplasia, male thyroids are more susceptible for IGF-I effects, compared with female thyroids, but this needs to be proven. Whereas the increase in serum IGF-I levels after recombinant human GH administration is stronger in men than women, the therapy gives rise to suppressed serum TSH in women only (14). This is in line with our finding that an association between serum IGF-I levels and decreased serum TSH levels was present in women but not men. Altogether, these findings indicate that female thyroids are more susceptible for IGF-I actions on thyroid hormone production than male thyroids.
From our study, we are able to hypothesize only about the mechanisms underlying these gender differences with estrogens offering some explanation. Using SHIP data, we recently demonstrated an inverse association between use of oral contraceptives and goiter, whereas parity was a determinant of goiter in premenopausal women (21). Inclusion of these women-specific factors in our multivariable models did not substantially affect the mean OR estimates of interest, indicating that these factors did not represent major confounders for the investigated associations. However, we cannot fully exclude residual confounding because measurements of serum estrogen levels were not available from our study.
Our study does not corroborate previous findings indicating potential modification of IGF-I effects by IGFBP-3 and TSH (9, 10). In mice overexpressing IGF-I, IGFBP-3 expression in the thyroid was reduced to 50% of the level observed in control mice (2), and enlarged, nodular thyroids in subjects with high serum IGF-I levels are associated with an increased risk for decreased serum TSH levels. Thus, direct and indirect down-regulation of IGFBP-3 and TSH in the presence of high IGF-I might provide a plausible explanation for lacking interactions in the present study.
Although a relatively large study population was investigated, we cannot fully exclude that some weak associations that were not present in our study might have escaped from detection due to low statistical power. For example, women with high serum IGF-I levels had a by a factor of 1.26 increased odds for thyroid nodules, compared with women with low serum IGF-I levels. The 95% CI for the mean OR estimate was 0.90–1.78, thus indicating a nonsignificant association. Post hoc power analyses revealed that at least 4600 subjects (instead of 3662 subjects included in the present study) would have had to be recruited to detect such difference as statistically significant (two sided 
2 test, 
= 5%, power = 80%). In particular, our study population might have been too small to detect the aforementioned interactions as statistically significant.
We conclude that high serum IGF-I levels are associated with goiter. Whereas high serum IGF-I levels are also related to thyroid nodules in men, they are related to decreased serum TSH levels in women. There are no interactions among serum IGF-I, IGFBP-3, and TSH levels with respect to thyroid disorders.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to declare.
First Published Online July 31, 2007
Abbreviations: BMI, Body mass index; CI, confidence interval; IGFBP, IGF binding protein; OR, odds ratio; SHIP, Study of Health in Pomerania.
Received April 11, 2007.
Accepted July 19, 2007.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. M. Gaglione, A. S. Can, H. J. Schneider, H.-U. Wittchen, H. Wallaschofski, T. Pischon, H. Boeing, and E. Riboli Obesity and Risk of Death N. Engl. J. Med., March 5, 2009; 360(10): 1042 - 1044. [Full Text] [PDF]
|
|
|
|
 |
|
 J. Moock, C. Albrecht, N. Friedrich, H. Volzke, M. Nauck, M. Koltowska-Haggstrom, T. Kohlmann, and H. Wallaschofski Health-related quality of life and IGF-1 in GH-deficient adult patients on GH replacement therapy: analysis of the German KIMS data and the Study of Health in Pomerania Eur. J. Endocrinol., January 1, 2009; 160(1): 17 - 24. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
