Subclinical hypothyroidism (SCH) is defined by mildly

ORIGINAL Endocrine ARTICLE Care Subclinical Hypothyroidism and Indices for Metabolic Syndrome in Japanese Women: One-Year Follow-Up Study Yasuyo Nakajima,* Masanobu Yamada,* Masako Akuzawa, Sumiyasu Ishii, Yasuhiro Masamura, Tetsurou Satoh, Koshi Hashimoto, Mayumi Negishi, Yohnosuke Shimomura, Isao Kobayashi, Yoshitaka Andou, and Masatomo Mori Department of Medicine and Molecular Science (Y.N., M.Y., S.I., T.S., K.H., M.M.), Gunma University Graduate School of Medicine, Maebashi 371-8511, Japan; and Hidaka Hospital (M.A., Y.M., M.N., Y.S., I.K., Y.A.), Takasaki 301-0001, Japan Context: Subclinical hypothyroidism (SCH) and metabolic syndrome (MetS) increase with age; however, their relationship remains unclear. Objective: Our objective was to investigate the relationship between SCH and indices of metabolic syndrome and follow up subjects for 1 year. Design: Cross-sectional and longitudinal follow-up studies of cases were collected from Takasaki Hidaka Hospital between 2003 and 2007. Participants: Overall, 11 498 participants of health checkups were analyzed. The mean age was 48 9 years. Main Outcome Measures: The relationship between SCH and indices of MetS were examined. Results: Serum free T 4 levels were lower in women than men in most of the age groups, and the prevalence of SCH, 6.3% in women vs 3.4% in men, increased with age, reaching 14.6% in 70-year-old women. Multivariate logistic-regression analyses revealed that waist circumference and the serum triglyceride and low-density lipoprotein-cholesterol levels were significantly higher in subjects with SCHthanwithoutamongwomen.Reflectingthesefindings,theadjustedoddsratioofMetSinpatientswith SCH was higher than in the euthyroid subjects in women with an odds ratio of 2.7 (95% confidence interval 1.1 5.6;P.017)butnotinmen.Furthermore,progressionfromeuthyroidintoSCHresultedinasignificant increase in the serum triglyceride levels but not low-density lipoprotein-cholesterol in women. Conclusion: Japanese women exhibited a high prevalence of SCH associated with low free T 4 levels. There was a strong association between SCH and several indices of metabolic syndrome in women. SCH may affect serum triglyceride levels and be a risk factor for metabolic syndrome. (J Clin Endocrinol Metab 98: 3280 3287, 2013) Subclinical hypothyroidism (SCH) is defined by mildly elevated plasma TSH, with normal thyroid hormone levels (1, 2). It remains controversial whether SCH leads to metabolic disturbances including dyslipidemia and the metabolic syndrome (MetS). In the cross-sectional Rotterdam study reported by Hak et al (3), SCH appeared to be an independent risk factor for atherosclerosis and myocardial infarction in elderly women. Although several ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright 2013 by The Endocrine Society Received February 8, 2013. Accepted May 24, 2013. First Published Online June 4, 2013 * Y.N. and M.Y. contributed equally to this work. Abbreviations: AOR, adjusted odds ratio; BMI, body mass index; BP, blood pressure; CI, confidence interval; FT4, free T 4 ; HbA1C, glycosylated hemoglobin; HDL-C, high-density lipoprotein-cholesterol; H-H, change from SCH to SCH after 1 year of follow-up; H-N, change from SCH to euthyroid; LDL-C, low-density lipoprotein-cholesterol; MetS, metabolic syndrome; N-H, change from euthyroid to SCH; NHANES III, National Health and Nutrition Examination Survey III database; N-N, change from euthyroid to euthyroid; SCH, subclinical hypothyroidism; SREBP, sterol regulatory element-binding protein; TG, triglyceride. 3280 jcem.endojournals.org J Clin Endocrinol Metab, August 2013, 98(8):3280 3287 doi: 10.1210/jc.2013-1353

doi: 10.1210/jc.2013-1353 jcem.endojournals.org 3281 studies have since been carried out, it is still unclear whether SCH and overt hypothyroidism are associated with cardiovascular disease (2, 4). The cholesterol-lowering effect of thyroid hormone in patients with hypothyroidism was first described in 1930. Since then, elevated low-density lipoprotein-cholesterol (LDL-C) levels have been consistently reported in patients with SCH and TSH greater than 10 miu/l; however, in patients with TSH less than 10 miu/l, results remain controversial, possibly due to the small numbers studied, inconsistent exclusion criteria that may include patients treated with medications for dyslipidemia, and the inconsistencies in the timing of blood sampling [free T 4 (FT4) and lipid metabolism may have circadian rhythms] (2, 5). Furthermore, triglycerides (TGs) have been recently reported to be associated with the risk of atherosclerotic vascular disease and may be related to SCH (6). MetS is a cluster of diseases that include obesity, hyperglycemia, dyslipidemia, and hypertension. MetS is a risk factor for type 2 diabetes and cardiovascular disease; the risk for cardiovascular disease and cardiovascular death increased 3 times in subjects with metabolic syndrome, and the risk for diabetes increased 5 times (7). The prevalence of MetS in Western countries is 20% 30%, whereas it is lower in Japan (8). There are several studies regarding the correlation between thyroid function and indices of MetS; however, the results are not conclusive and may differ in iodine-sufficient Japan from that in Western countries (9). Because both SCH and dyslipidemia increase with age, we first evaluated the relation between SCH and lipid metabolism, adjusted by age, in healthy Japanese subjects participating in an annual health check. In addition, we evaluated these lipid parameters in subjects whose thyroid status progressed from normal to SCH within one year. Subjects and Methods Subjects Cross-sectional and longitudinal follow-up studies that included 11 498 Japanese who participated in health checkups at the Hidaka Hospital from April 2003 through March 2007 were used in the present study. For all participants, blood samples were collected in the early morning from 8:00 to 9:00 AM after fasting for at least 11 hours. Medical histories including history and medications were reported by the subjects themselves. Exclusion criteria were as follows: any history of diseases such as thyroid diseases, liver cirrhosis and renal failure; the taking of medications including levothyroxine, antithyroid drugs, insulin, hypoglycemic drugs, corticosteroid, hypolipidemic drugs, or antihypertensive drugs; and missing data. Using these criteria, 978 subjects were excluded. Furthermore, of 10 520 subjects, 4764 revisited our hospital for annual health checkups and could be followed up for 1 year. All subjects provided informed consent for this study, which was approved by the Ethics Committee of the Hidaka Hospital. Thyroid function testing and the definition of subclinical hypothyroidism Plasma TSH and FT4 were measured using Chemilumi ACS- TSHIII and ACS-FT4 kits (Siemens Healthcare Diagnostics, Inc, Tokyo, Japan). Similar to several other studies, we clarified SCH based on serum FT4 within the reference range and TSH above the upper limit of normal range, 4.0 miu/l with no upper limit (10 12). Criteria for metabolic syndrome and other risk factors MetS was diagnosed according to the modified criteria of the National Cholesterol Education Program Adult Treatment Panel III adapted for the Japanese population (13). The Japanese Obesity Society defined central obesity for Japanese people as a waist circumference of 85 cm or greater for men and of 90 cm or greater for women because the cutoff point of visceral fat at 100 cm 2 is indicative of the risk of obesity-related disorders and a waist circumferences of 85 cm in men and 90 cm in women approximates to this visceral fat mass (14). In addition to central obesity, subjects with MetS had to have 2 or more additional criteria: elevated blood pressure (BP), elevated fasting glucose, and/or dyslipidemia. Blood pressure was measured 3 times in the sitting position after 5-minute rest periods. Elevated BP was defined as a systolic BP of 130 mm Hg or greater and/or a diastolic BP of 85 mm Hg or greater. Elevated fasting glucose was defined as a fasting glucose level of 110 mg/dl or greater (6.1 mmol/l). Dyslipidemia was defined as a plasma TG level of 150 mg/dl or greater and/or a high-density lipoprotein-cholesterol (HDL-C) level of less than 40 mg/dl. In our questionnaire, we asked whether the patients are smoking at least once a day and whether the smoking is current, previous, or never. In the present study, we defined a subject who is currently smoking at least once a day as a smoker. All results are expressed as the mean SD for continuous variables and as absolute numbers and relative percentages for categorical variables. Group comparisons were performed with an ANOVA and Student s t test for normally distributed data or the Wilcoxon rank-sum test for nonnormally distributed data for continuous variables. Multiple group comparisons were performed with an ANOVA for a difference across multiple groups. Dunnett s method was used for multiple comparisons between different groups. The 2 test was used for categorical variables. Univariate and multivariate logistic-regression analyses were used to estimate odds ratios and corresponding 95% confidence intervals (CIs). The characteristics that affected the univariate analysis (P.1) were included in the multivariable models. All tests for significance and resulting P values were 2 sided, with a level of significance of 5%. Statistical analyses were performed using JMP 5.1.2 (SAS Institute Inc, Cary, North Carolina). The estimated sample size was based on the detection of a mean increase in the TG level by 30% between patients with SCH and controls, with a 2-sided type 1 error of less than 5% and a power of 70%.

3282 Nakajima et al Subclinical Hypothyroidism and Metabolic Syndrome J Clin Endocrinol Metab, August 2013, 98(8):3280 3287 Table 1. Percentage of Patients With SCH Age Groups, y Results Sample Size Percentage Men Total 7120 100 3.4 21 30 99 1.4 3.0 31 40 1572 22.1 1.9 41 50 2606 36.6 2.5 51 60 2205 31.0 4.3 61 70 549 7.7 7.1 71 89 1.3 10.1 Women Total 4378 100 6.3 21 30 74 1.7 1.3 31 40 1025 23.4 4.2 41 50 1552 35.5 5.1 51 60 1394 31.8 8.4 61 70 285 6.5 9.8 71 48 1.1 14.6 Percentage of SCH Characteristic of thyroid status in middle-aged participants in Japan Table 1 shows the demographic and thyroid status of the participants. The total number of subjects was 11 498, and the percentage of men was 62%. The mean age was 48 9 years among men (range 22 88 years) and 48 9 years among women (range 24 87 years). The overall prevalence of SCH was 3.4% in men and 6.3% in women and increased with age, to 14.6% in 70-year-old women and 10.1% in 70-year-old men. Therefore, the prevalence was higher in women than men, which is similar to that in the United States and results of other reports (11, 12, 15 18). Serum FT4 levels in men was progressively decreased with age (Figure 1): the mean FT4 level was 1.4 0.2 ng/dl among men in their 30s and 1.2 0.2 ng/dl among those in their 70s (P.001). However, there was no significant change in women. In addition as shown in Figure 2A, the distribution of values was clearly shifted to the left for men in their 30s compared with men in their 60s. No change in the distribution was observed for women (Figure 2B). It is also worth noting that serum FT4 levels were significantly lower in women than in men at most ages. Figure 1. Relationship between age and the concentration of FT4 in men (A) and women (B). The line represents the linear regression fitted to the data points calculated using JMP software (SAS Institute Inc). Figure 2. The percentage of the serum FT4 levels in the groups 31 40 and 61 70 years old in men (A) and women (B). The distribution of serum FT4 levels in the elder group shifted only among men. Characteristics of subjects with and without subclinical hypothyroidism Table 2 shows the clinical characteristics of participants with and without SCH. The univariate analysis revealed that the mean age of patients with SCH was higher than that of euthyroid subjects in both sexes. It is of interest that the waist circumference was significantly higher in patients with SCH than normal subjects among women, but not among men, and the body mass index (BMI) was also significant in women. Serum LDL-C levels were significantly higher in SCH patients among women but not men. In addition, serum TG and HDL-C levels were significantly higher and lower, respectively, in patients with SCH only for women. Among women, all other factors including fasting blood glucose, glycosylated hemoglobin (HbA1C), and systolic and diastolic blood pressures were significantly higher in patients with SCH than in normal subjects. In contrast, among men, no factor was significantly changed in SCH patients compared with normal subjects. Furthermore, the ratio of SCH was significantly lower in smokers than that nonsmokers among men but not women. Because many factors that differed significantly between patients with SCH and normal subjects progress with age, we performed multivariate logistic-regression analyses. When adjusted with age and other factors, waist circumference and serum TG and LDL-C levels were significantly higher in subjects with SCH than without among women. However, systolic and diastolic BP (in men and women), HDL-C, and fasting blood glucose were not significant after adjusting for age, waist circumference, BMI, and smoking.

doi: 10.1210/jc.2013-1353 jcem.endojournals.org 3283 Table 2. Baseline Characteristics of Participants With and Without SCH Men (n 6375) Women (n 3975) SCH ( ) (n 6164) SCH ( ) (n 211, 3.3%) Univariate (P value) Multivariate (P value) SCH ( ) (n 3732) SCH ( ) (n 243, 6.4%) Univariate (P value) Multivariate (P value) Age, years 48 9 52 10.0001.0001 47 9 51 9.0001.0001 Waist circumference, cm (male: n 4290; 84 8 84 9.38 77 9 80 9.0003.009 a female: n 2537) BMI, kg/m 2 23.7 3.8 23.7 3.1.95 21.8 3.2 22.3 3.3.034.25 a LDL-C, mg/dl 122 31 125 31.14 113 31 122 34.0001.02 b TGs, mg/dl 128 92 130 93.77 79 64 92 56.0016.01 b HDL-C, mg/dl 52 13 53 12.61 64 14 61 13.0014.07 b Fasting blood sugar, mg/dl 104 15 104 14.69 96 11 98 11.0093 HbA1C, % 5.6 0.6 5.6 0.5.44 5.4 0.5 5.5 0.5.0028 Systolic BP, mm Hg 123 15 125 15.63 116 16 121 19.0001.21 b Diastolic BP, mm Hg 78 11 79 11.026.65 c 71 11 75 12.0001.98 b Smokers, % 41 16.0001.0001 c 8 5.089 MetS, % 15 15.92 2 5.001.017 c OR 2.7 (95% CI 1.1 5.6) Exclusion criteria were as follows: any history of diseases such as thyroid diseases, liver cirrhosis, and renal failure; the taking of medications including levothyroxine, antithyroid drugs, insulin, hypoglycemic drugs, corticosteroids, hypolipidemic drugs, or antihypertensive drugs; and missing data. a Adjusted for age, BMI, and smoking. b Adjusted for age, waist circumference, BMI, and smoking. c Adjusted for age. Reflecting these results, the adjusted odds ratio (AOR) of MetS in patients with SCH compared with those without was 2.7 (1.1 5.6; P.017) among women. Furthermore, a similar significantly higher prevalence of MetS in women with SCH than in those without SCH was observed before the age of 50 years [3 of 61 (4.92%) vs 0.87% (13 of 1492) (P.002], and several other studies demonstrated the menopausal status to be a risk factor for metabolic syndrome (19 21). Therefore, both SCH and the menopausal status may be risk factors for MetS in women. Progression to SCH increased the serum TG levels in women We were able to follow up 4764 of the 10 520 subjects for a period of 1 year. Among these subjects (3074 men and 1690 women), 4.2% [105 men (3.5%) and 94 women (5.7%)] had SCH. Only 2 subjects (1 male and 1 female) had overt hypothyroidism. As shown in Figure 3, euthyroid progressed to SCH in 1.7% of the men (n 51) and 2.2% of the women (n 37). Fifty-four percent of patients with SCH (105 of 195) still had SCH after 1 year (56% of men and 52% of women). Of note, 46% of SCH patients returned to normal after 1 year (44% of men and 48% of women), and only 1 patient with SCH developed to overt hypothyroidism in men and none in women. Although the precise mechanism underlying the normalization of serum TSH levels in some patients with SCH after 1 year remains unclear, a similar phenomenon has been reported by many other studies (22 24); the percentages of normalization over a few years ranged from 9% to 46%. No particular mechanisms have been reported for this phenomenon; however, we found that all patients with SCH who returned to euthyroid showed a milder elevation in serum TSH levels (a median of 4.7 miu/l with a maximum of 8.1 miu/l among men and a median of 4.6 miu/l with a maximum of 9.3 miu/l among women) than that of the patients with SCH who remained SCH (a median of 5.5 miu/l with a maximum of 15.9 miu/l among men and a median of 5.5 miu/l with a maximum of 12.0 miu/l among women) (adjusted by age and smoking; P.0068 for men, P.001 for women). We expected that the Figure 3. Changes of serum TSH levels over 1 year in each participant among men (A) and women (B). We designated changes from euthyroid to euthyroid after the 1-year follow-up as N-N; SCH to SCH, H-H; euthyroid to SCH, N-H, and SCH to euthyroid, H-N. The gray lines represent individuals who belonged to the N-N group; black, N-H; red, H-N; and blue, H-H in Table 3. In men (A), H-H accounted for 1.9% of participants, which was 56% of SCH at baseline; N-H, 1.7%; H-N, 1.5% (44% of SCH), and N-N, 95.0%. In women (B), H-H accounted for 2.9% of participants (52% of SCH at baseline); N-H, 2.2%; H-N, 2.7% (48% of SCH); and N-N, 92.1%.

3284 Nakajima et al Subclinical Hypothyroidism and Metabolic Syndrome J Clin Endocrinol Metab, August 2013, 98(8):3280 3287 Table 3. Changes in the Indices of MetS in Men After One Year in Normal Subjects and Subjects N-H, N-N, H-N, and H-H Baseline After 1 y (Percent Increase) N-N (n 2850) N-H (n 51) H-N (n 45) H-H (n 57) P Value N-N (n 2850) N-H (n 51) H-N (n 45) H-H (n 57) P Value Age, y 47 8 50 8 46 7 51 9.01 a BMI, kg/m 2 23.5 2.9 24.2 3.1 23.5 2.4 23.4 2.7 0.1 3.5 0.6 5.0 0.8 4.1 0.2 3.0 TG, mg/dl 125 78 123 70 140 124 121 69 7.9 52.4 14.4 42.2 9.9 44.6 19.2 47.0 HDL-C, mg/dl 51 12 49 10 51 13 53 15 2.2 12.3 4.8 10.0 1.4 14.9 2.8 12.1 LDL-C, mg/dl 123 31 127 25 119 35 122 31 0.5 15.2 0.2 15.5 0.6 15.2 1.2 14.8 FBS, mg/dl 103 14 110 24 101 10 105 15 0.2 7.6 2.0 10.0 0.1 7.2 0.7 7.8 HbA1C, % 5.6 0.6 5.9 0.9 5.5 0.3 5.6 0.6 0.2 5.0 1.2 7.3 0.2 4.0 0.3 4.6 SBP, mm Hg 121 14 123 15 125 12 125 15 0.1 9.4 0.2 9.5 2.8 8.7 0.1 9.9 DBP, mm Hg 77 10 78 10 79 7 80 10 0.2 10.6 0.2 11.0 1.6 9.8 0.7 10.4 FT4, ng/dl 1.3 0.2 1.2 0.2 1.3 0.2 1.2 0.2.001 b 0.8 13.9 2.4 13.2 1.2 11.2 1.35 13.6 TSH, miu/l 1.5 0.7 2.9 0.7 5.0 1.0 6.2 2.2.05 b 3.0 38.6 84.7 66.7 35.1 17.5 3.1 30.2.0001 c Smoking, % 40 20 7 23.0001 c Abbreviations: DBP, diastolic BP; FBS, fasting blood sugar; SBP, diastolic BP. Data are expressed as mean SD. Exclusion criteria were as follows: any history of diseases such as thyroid diseases, liver cirrhosis and renal failure; the taking of medications including levothyroxine, antithyroid drugs, insulin, hypoglycemic drugs, corticosteroid, hypolipidemic drugs, or antihypertensive drugs; and missing data. Values are as follows: N, TSH 0.4 4.0 miu/l and FT4 0.7 1.9 ng/dl (normal range); H, TSH greater than 4.0 miu/l and FT4 normal. a N-H and H-H were significantly different from N-N. b Significantly different from N-N, and P values were adjusted for age and smoking. c P values were adjusted for age and smoking. N-H and H-N were statistically significantly different from N-N. amount of iodine intake or nonthyroidal disease may have contributed to the transient rise in TSH in some participants. We designated the changes from euthyroid to euthyroid after 1 year of follow-up as N-N; SCH to SCH, H-H; euthyroid to SCH, N-H, and SCH to euthyroid, H-N. We next examined changes in the indices of MetS after 1 year in the above 4 groups. The upper column of Table 3 shows the baseline characteristics of the male participants. There was no significant difference in any of the factors examined except for age and the serum TSH level between these 4 groups. The mean age was older for N-H and H-H than for N-N. It is important to note that patients with higher serum TSH levels were susceptible to SCH after 1 year. Furthermore, FT4 levels were significantly higher in the N-N group than the H-H, N-H, and H-N groups, and again the rate of smoking was higher in the N-N group. As in men, significant changes in women were observed in serum TSH and FT4 levels when compared N-N with others and between N-N and H-H (Table 4). However, in contrast to men, the progression from euthyroidism to SCH (N-H) induced a significant increase (about 30%) in Table 4. Changes in the Indices of MetS in Women After One Year in Normal Subjects and Subjects N-H, N-N, H-N, and H-H Baseline After 1 y (Percent Increase) N-N (n 1527) N-H (n 37) H-N (n 45) H-H (n 48) P Value N-N (n 1527) N-H (n 37) H-N (n 45) H-H (n 48) P Value Age, y 47 8 49 8 48 9 47 9 BMI, kg/m 2 21.8 3.0 22.1 3.8 21.9 3.2 20.8 2.3 0.1 3.4 1.2 4.4 0.4 3.0 0.6 3.8 TG, mg/dl 75 43 74 44 79 37 81 44 9.3 42.6 27.5 45.2 2.5 30.6 0.5 34.2.028 a HDL-C, mg/dl 64 14 61 12 65 14 64 14 1.3 11.7 1.7 11.5 0.7 12.3 2.3 13.0 LDL-C, mg/dl 113 30 116 25 115 31 118 33 0.6 15.6 1.3 14.4 0.7 16.6 2.4 1.7.6 FBS, mg/dl 96 10 95 6 95 9 95 8 0.6 6.8 0.8 7.0 0.2 7.2 0.3 7.5 HbA1C, % 5.4 0.4 5.3 0.4 5.5 0.4 5.4 0.4 0.9 4.3 2.0 0.5 0.5 4.3 1.6 4.8 SBP, mm Hg 116 15 115 17 115 15 120 16 0.4 9.6 2.5 10.48 0.0 9.1 1.5 8.8 DBP, mm Hg 71 10 69 11 72 10 74 11 0.1 11.8 1.6 10.9 1.2 14.6 0.9 9.1 FT4, ng/dl 1.2 0.2 1.2 0.2 1.2 0.2 1.1 0.1.0001 b 0.2 14.6 3.6 16.8 1.8 19.4 1.0 13.3 TSH, miu/l 1.7 0.8 2.7 0.6 4.8 1.0 5.8 1.7.0001 c 3.2 38.7 132.7 215.0 34.2 16.6 7.7 25.0.0001 d Smoking, % 7 14 2 2 Abbreviations: DBP, diastolic BP; FBS, fasting blood sugar; SBP, diastolic BP. Data expressed as mean SD. Values are as follows: N, subjects with normal TSH and FT4 levels; H, subjects with high TSH and normal FT4 levels. a Comparison of N-N with N-H was significant. b Comparison of N-N with H-H was significant. c Comparison of N-N with other groups was significant. d Comparisons among N-N, N-H, and H-N were significant.

doi: 10.1210/jc.2013-1353 jcem.endojournals.org 3285 the TG level in women. Unexpectedly, there was no significant change in LDL-C levels among these groups. Discussion Prevalence of SCH in the Japanese population SCH has been reported to occur in 4% 20% of the adult population (2). The prevalence of SCH in the United States was 4.3% among 17 353 people over age 12 years in the National Health and Nutrition Examination Survey III database (NHANES III) (10), 10.8% among menopausal women in the Rotterdam study (3), and 9.5% among 25 862 participants in the Colorado study (12). In an examination of atomic bomb survivors in Nagasaki, Japan, SCH was observed in 9.8% of men and 10.5% of women (25). We found that the prevalence of SCH was 3.4% in middle-aged men and 6.3% in women. Because the value increased with age, the high prevalence in the Nagasaki study was likely due to the average age; 58.5 years vs 48 years in our study. Although previous studies suggested that aging was associated with reduced TSH secretion, several more recent studies including NHANES III have shown that serum TSH concentrations increase with age (10, 12). However, changes in either serum FT4 or total T 4 levels associated with age are controversial. The NHANES III study reported a decrease in total T 4 levels with age, and a negative association of FT4 with age was also reported by Corsonello et al (26). However, no changes in FT4 levels with age were reported in the European Prospective Investigation in Cancer and Nutrition-Norfolk study (11) and by Bremner et al (27). This discrepancy might be due to differences in age, race, the amount of iodine, body mass index, and the cutoff concentrations for the serum TSH level used to define SCH. In the present study, we clearly demonstrated a decrease in FT4 levels with age specific to men. Interestingly, 2 other Japanese studies have reported a decrease in FT4 levels with age among men, although the first former study was relatively small (approximately 500 participants) and the second was performed with patients who received medications at hospitals (28, 29). Considering these observations together with our results, it appears that there is a male-specific decrease of FT4 levels with age in Japan. With respect to the gender difference in thyroid function, the present study showed that the prevalence of SCH was higher in women than in men. The Colorado study also demonstrated that the percentage of subjects with an elevated TSH concentration was greater for women than men in each decade of life (12), and the distribution of the serum TSH level progressively shifted toward higher concentrations with age (30). We also found that serum FT4 levels were significantly lower in women than men in most decades of life. These results suggest that lower thyroid hormone levels might induce the higher prevalence of SCH in women. Although the mechanisms responsible for the lower T 4 levels in women remain unclear, several studies have suggested that the action of estrogen and thyroid hormone may overlap. Further studies now of the gender difference in thyroid function and its change associated with age are needed. Prevalence of MetS in SCH Despite several studies, the clinical relevance of SCH as a risk factor for MetS remains controversial (31 33). The present study clearly demonstrated that the prevalence of MetS was significantly higher in patients with SCH than in euthyroid subjects among women (Table 2). SCH was found in 15.0% of subjects with MetS and 6.0% of those without (non-mets) (P.001). Conversely, MetS was found in 5.0% of subjects with and 2.0% of those without SCH, and the AOR of MetS in women with SCH was 2.7 [95% CI 1.1 5.6, P.017] (Table 2). Furthermore, although the criteria of waist circumference of 85 cm for Japanese men and 90 cm for women have been controversial, even when we used 87 cm for Japanese men and 80 cm for women (34), we observed no difference among the men and the AOR with 2.3 (95% CI 1.3 3.9, P.0046) among the women. A recent study by Roos et al (9) showed a significant correlation between FT4 levels within the normal range and the index of MetS, and low-normal FT4 levels were significantly associated with increased insulin resistance. Our study also demonstrated that the serum FT4 level was significantly lower in subjects with SCH than normal subjects in both sexes (men, 1.3 0.2 vs 1.2 0.2 ng/dl, P.0001; women, 1.2 0.2 vs 1.1 0.2 ng/dl, P.0001). The present study also demonstrated that almost all indexes of MetS including TGs, waist circumference, BP, and fasting blood glucose levels were significantly higher than those in normal subjects. However, these differences were observed only in women. Because we found FT4 levels to be significantly lower in women than those men, lower FT4 levels might have a significant influence on indexes of MetS only in women. We measured waist circumferences in some participants at the 1-year follow-up, but the number was too small to analyze. Lipid profiles in patients with SCH More than 90% of patients with overt hypothyroidism have hyperlipidemia with serum total cholesterol and LDL-C levels increased by approximately 30%. In addition, several studies have reported that serum TG levels to be increased in patients with overt hypothyroidism (5).

3286 Nakajima et al Subclinical Hypothyroidism and Metabolic Syndrome J Clin Endocrinol Metab, August 2013, 98(8):3280 3287 However, changes in serum lipid levels in patients with SCH have been inconsistent (5, 12). The present study clearly demonstrated that among women, both LDL-C and TG levels were higher in patients with SCH, even when adjusted by age, BMI, and waist circumstances. Similar changes and sex differences in lipid profiles of patients with SCH have been reported in Western countries; the Colorado study demonstrated that among 25 862 participants, mean total cholesterol and LDL-C levels were significantly greater in subjects with TSH values of between 5.1 and 10 miu/l than in euthyroid subjects (12). The recent European Prospective Investigation in Cancer and Nutrition-Norfolk prospective population study also reported that among 11 554 subjects aged 45 79 years, those with subclinical hypothyroidism (defined as a serum TSH 4.0 miu/l) had significantly higher levels of total cholesterol, LDL-C, and TGs and lower HDL-C among women but not men (11). Fabbrini et al (35) recently reported that TG concentrations were associated with subclinical thyroid dysfunction in women, and the very low-density lipoprotein-tg secretion rate was greater in women with subclinical hypothyroidism than euthyroidism and subclinical hyperthyroidism. In addition, Waterhouse et al (36) reported that among 728 women aged 45 60 years, fasting glucose and TG level were independently associated with TSH concentrations within the reference range. They also indicated that gender, age, obesity, and diet influence the correlation between serum TSH levels and fasting glucose and serum TG levels (36). Furthermore, Lai et al (37) reported that among 1534 adults in a Chinese cohort, there was a positive correlation between serum TSH levels within the normal range and TG levels after adjusting for gender, age, and homeostasis model assessment insulin resistance index. We have recently reported that mouse sterol regulatory element-binding protein (SREBP)-1c mrna expression was down-regulated by the administration of T 3 in vivo and that incubation with T 3 negatively regulated the promoter activity of the mouse SREBP-1c gene in vitro (38). Therefore, the decrease in FT4 levels in patients with SCH may induce an increase in serum TG levels by inducing hepatic production of very low-density lipoprotein through the activation of SREBP1c (35, 38). An important finding of this study is that we confirmed the effect of SCH on the serum TG level in women who progressed within 1 year from normal thyroid status to SCH. Unexpectedly, no significant change was observed in serum LDL-C levels. Several papers have reported that elevated levels of serum LDL-C and TG in SCH subjects were reduced within a few weeks by treatment with a thyroid hormone (5). In addition, it was recently reported that serum LDL-C and TG levels significantly increased in patients with overt hypothyroidism induced by total thyroidectomy in patients with differentiated thyroid cancer within 2 3 weeks (39). These findings suggest that in the short term, a few months, the thyroid hormone affects serum lipid levels, but in the long term, a year for example, factors other than mild thyroid hormone status predominantly affect serum LDL-C. Further study is required for understanding this phenomenon. Limitations of the study Several limitations need to be considered when interpreting the present findings. First, we could not obtain data on thyroid autoimmunity, which represents an unreported thyroid disease and may influence TSH concentrations. NHANES III and other studies reported that serum thyroperoxidase antibody and thyroglobulin antibody concentrations increased with age, which suggested that age-dependent increases in the prevalence of hypothyroidism are due to the development of Hashimoto s thyroiditis; however, Kasagi et al (40) reported, using similar data to us from general health checkup systems in Japan, that the prevalence of thyroglobulin antibody and thyroperoxidase antibody tests were not associated with age-dependent increases in the prevalence of hypothyroidism. Therefore, the effect of thyroid autoimmunity on developing SCH in iodine-sufficient Japan still remains controversial. Second, because a part of this study was a cross-sectional study, a cause-and-effect relationship could not be discerned; however, the present study also included a longitudinal follow-up study for a year. Third, to examine the absolute value of lipid and glucose metabolism and blood pressures, we excluded patients taking medications. Therefore, the prevalence of MetS in this study was lower than expected. Fourth, compared with a population-based cohort study, there was some bias in the background of the subjects in terms of their economic status because only 17% of the participants received financial support from mutual aid associations; the other subjects paid all expenses for the checkup by themselves. Acknowledgments Address all correspondence and requests for reprints to: Masanobu Yamada, MD, PhD, Department of Medicine and Molecular Science, Gunma University Graduate School of Medicine, 3-39-15 Showa-machi, Maebashi, Gunma 371-8511, Japan. E-mail: myamada@gunma-u.ac.jp. Disclosure Summary: The authors have nothing to disclose. References 1. Surks MI, Ortiz E, Daniels GH, et al. Subclinical thyroid disease: scientific review and guidelines for diagnosis and management. JAMA. 2004;291:228 238.

doi: 10.1210/jc.2013-1353 jcem.endojournals.org 3287 2. Cooper DS, Biondi B. Subclinical thyroid disease. Lancet. 2012; 379:1142 1154. 3. Hak AE, Pols HA, Visser TJ, Drexhage HA, Hofman A, Witteman JC. Subclinical hypothyroidism is an independent risk factor for atherosclerosis and myocardial infarction in elderly women: the Rotterdam Study. Ann Intern Med. 2000;132:270 278. 4. Razvi S, Pearce SH. Subclinical hypothyroidism and cardiovascular disease-reply. Arch Intern Med. 2012;172:1523 1524. 5. Pearce EN. Update in lipid alterations in subclinical hypothyroidism. J Clin Endocrinol Metab. 2012;97:326 333. 6. Tognini S, Polini A, Pasqualetti G, et al. Age and gender substantially influence the relationship between thyroid status and the lipoprotein profile: results from a large cross-sectional study. Thyroid. 2012; 22:1096 1103. 7. Stern MP, Williams K, González-Villalpando C, Hunt KJ, Haffner SM. Does the metabolic syndrome improve identification of individuals at risk of type 2 diabetes and/or cardiovascular disease? Diabetes Care. 2004;27:2676 2681. 8. Shiwaku K, Nogi A, Kitajima K, et al. Prevalence of the metabolic syndrome using the modified ATP III definitions for workers in Japan, Korea and Mongolia. J Occup Health. 2005;47:126 135. 9. Roos A, Bakker SJ, Links TP, Gans RO, Wolffenbuttel BH. Thyroid function is associated with components of the metabolic syndrome in euthyroid subjects. J Clin Endocrinol Metab. 2007;92:491 496. 10. Hollowell JG, Staehling NW, Flanders WD, et al. Serum TSH, T(4), and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Examination Survey (NHANES III). J Clin Endocrinol Metab. 2002;87:489 499. 11. Boekholdt SM, Titan SM, Wiersinga WM, et al. Initial thyroid status and cardiovascular risk factors: the EPIC-Norfolk prospective population study. Clin Endocrinol (Oxf). 2010;72:404 410. 12. Canaris GJ, Manowitz NR, Mayor G, Ridgway EC. The Colorado thyroid disease prevalence study. Arch Intern Med. 2000;160:526 534. 13. [Definition and the diagnostic standard for metabolic syndrome Committee to Evaluate Diagnostic Standards for Metabolic Syndrome]. Nihon Naika Gakkai Zasshi. 2005;94:794 809. 14. Japan Society for the Study of Obesity. New criteria for obesity disease in Japan. Circ J. 2002;66:987 992. 15. Kanaya AM, Harris F, Volpato S, Pérez-Stable EJ, Harris T, Bauer DC. Association between thyroid dysfunction and total cholesterol level in an older biracial population: the health, aging and body composition study. Arch Intern Med. 2002;162:773 779. 16. Bindels AJ, Westendorp RG, Frölich M, Seidell JC, Blokstra A, Smelt AH. The prevalence of subclinical hypothyroidism at different total plasma cholesterol levels in middle aged men and women: a need for case-finding? Clin Endocrinol (Oxf). 1999;50:217 220. 17. Vanderpump MP, Tunbridge WM, French JM, et al. The incidence of thyroid disorders in the community: a twenty-year follow-up of the Whickham Survey. Clin Endocrinol (Oxf). 1995;43:55 68. 18. Parle JV, Franklyn JA, Cross KW, Jones SC, Sheppard MC. Prevalence and follow-up of abnormal thyrotrophin (TSH) concentrations in the elderly in the United Kingdom. Clin Endocrinol (Oxf). 1991;34:77 83. 19. Carr MC. The emergence of the metabolic syndrome with menopause. J Clin Endocrinol Metab. 2003;88:2404 2411. 20. Lin JW, Caffrey JL, Chang MH, Lin YS. Sex, menopause, metabolic syndrome, and all-cause and cause-specific mortality cohort analysis from the Third National Health and Nutrition Examination Survey. J Clin Endocrinol Metab. 2010;95:4258 4267. 21. Park YW, Zhu S, Palaniappan L, Heshka S, Carnethon MR, Heymsfield SB. The metabolic syndrome: prevalence and associated risk factor findings in the US population from the Third National Health and Nutrition Examination Survey, 1988 1994. Arch Intern Med. 2003;163:427 436. 22. Somwaru LL, Rariy CM, Arnold AM, Cappola AR. The natural history of subclinical hypothyroidism in the elderly: the cardiovascular health study. J Clin Endocrinol Metab. 2012;97:1962 1969. 23. Díez JJ, Iglesias P. Spontaneous subclinical hypothyroidism in patients older than 55 years: an analysis of natural course and risk factors for the development of overt thyroid failure. J Clin Endocrinol Metab. 2004;89:4890 4897. 24. Gussekloo J, van Exel E, de Craen AJ, Meinders AE, Frölich M, Westendorp RG. Thyroid status, disability and cognitive function, and survival in old age. JAMA. 2004;292:2591 2599. 25. Imaizumi M, Akahoshi M, Ichimaru S, et al. Risk for ischemic heart disease and all-cause mortality in subclinical hypothyroidism. J Clin Endocrinol Metab. 2004;89:3365 3370. 26. Corsonello A, Montesanto A, Berardelli M, et al. A cross-section analysis of FT3 age-related changes in a group of old and oldest-old subjects, including centenarians relatives, shows that a down-regulated thyroid function has a familial component and is related to longevity. Age Ageing. 2010;39:723 727. 27. Bremner AP, Feddema P, Leedman PJ, et al. Age-related changes in thyroid function: a longitudinal study of a community-based cohort. J Clin Endocrinol Metab. 2012;97:1554 1562. 28. Takeda K, Mishiba M, Sugiura H, Nakajima A, Kohama M, Hiramatsu S. Evaluated reference intervals for serum free thyroxine and thyrotropin using the conventional outliner rejection test without regard to presence of thyroid antibodies and prevalence of thyroid dysfunction in Japanese subjects. Endocr J. 2009;56:1059 1066. 29. Suzuki S, Nishio S, Takeda T, Komatsu M. Gender-specific regulation of response to thyroid hormone in aging. Thyroid Res. 2012; 5:1. 30. Surks M, Hollowell J. Age-specific distribution of serum thyrotropin and antithyroid antibodies in the US population: implications for the prevalence of subclinical hypothyroidism. J Clin Endocrinol Metab. 2007;92:4575 4582. 31. Wang CY, Chang TC, Chen MF. Associations between subclinical thyroid disease and metabolic syndrome. Endocr J. 2012;59:911 917. 32. Garduño-Garcia JeJ, Alvirde-Garcia U, López-Carrasco G, et al. TSH and free thyroxine concentrations are associated with differing metabolic markers in euthyroid subjects. Eur J Endocrinol. 2010; 163:273 278. 33. Uzunlulu M, Yorulmaz E, Oguz A. Prevalence of subclinical hypothyroidism in patients with metabolic syndrome. Endocr J. 2007; 54:71 76. 34. Matoba Y, Inoguchi T, Nasu S, Suzuki S, Yanase T, Nawata H, Takayanagi R. Optimal cut points of waist circumference for the clinical diagnosis of metabolic syndrome in the Japanese population. Diabetes Care. 2008;31:590 592. 35. Fabbrini E, Magkos F, Patterson BW, Mittendorfer B, Klein S. Subclinical hypothyroidism and hyperthyroidism have opposite effects on hepatic very-low-density lipoprotein-triglyceride kinetics. J Clin Endocrinol Metab. 2012;97:E414 E418. 36. Waterhouse DF, McLaughlin AM, Walsh CD, Sheehan F, O Shea D. An examination of the relationship between normal range thyrotropin and cardiovascular risk parameters: a study in healthy women. Thyroid. 2007;17:243 248. 37. Lai Y, Wang J, Jiang F, et al. The relationship between serum thyrotropin and components of metabolic syndrome. Endocr J. 2011; 58:23 30. 38. Hashimoto K, Yamada M, Matsumoto S, Monden T, Satoh T, Mori M. Mouse sterol response element binding protein-1c gene expression is negatively regulated by thyroid hormone. Endocrinology. 2006;147:4292 4302. 39. Pearce EN, Wilson PW, Yang Q, Vasan RS, Braverman LE. Thyroid function and lipid subparticle sizes in patients with short-term hypothyroidism and a population-based cohort. J Clin Endocrinol Metab. 2008;93:888 894. 40. Kasagi K, Takahashi N, Inoue G, Honda T, Kawachi Y, Izumi Y. Thyroid function in Japanese adults as assessed by a general health checkup system in relation with thyroid-related antibodies and other clinical parameters. Thyroid. 2009;19:937 944.