Association Between Thyroid Dysfunction and Lipid Profiles Differs According to Age and Sex: Results from the Korean National Health and Nutrition Examination Survey

Abstract

Background:

Lipid profiles of men and women change differently during the aging process. Guidelines recommend that dyslipidemia patients should consider screening for hypothyroidism without consideration of age or sex.

Methods:

Data from the sixth Korean National Health and Nutrition Examination Survey were used. A total of 4275 participants without thyroid disease and without a past history of dyslipidemia or dyslipidemia medication were evaluated. The association between thyroid dysfunction and lipid profiles (total cholesterol [TC], low-density lipoprotein cholesterol [LDLC], and triglycerides [TG]) was analyzed by age and sex.

Results:

The prevalence of thyroid dysfunction was significantly different according to TC and LDLC levels (p = 0.003 and p = 0.021, respectively). In women, the weighted prevalence of thyroid dysfunction was significantly different according to levels of TC, LDLC, and TG (p = 0.007, p = 0.016, and p = 0.044, respectively). However, in men, no association was found in any of the lipid profiles. Female participants were divided into two groups using a cutoff age of 55 years. In younger women, the weighted prevalence of thyroid dysfunction was different according to the levels of TC, LDLC, and TG (p = 0.013, p = 0.007, and p = 0.007, respectively). However, in older women, no association was found for any of the lipid profiles.

Conclusions:

The prevalence of thyroid dysfunction was significantly different according to lipid profiles, and this association differed by age and sex.

Introduction

Many guidelines recommend that dyslipidemia patients should consider screening for hypothyroidism (1,2). The prevalence of hypothyroidism in patients with dyslipidemia is almost double that of those without dyslipidemia (3). In some studies, lipid profiles have different chronological changes in men and women (4,5). Men showed higher low-density lipoprotein cholesterol (LDLC) levels before midlife than women. After midlife, women had more increased LDLC levels than men, and the gap between men and women closed in higher age groups. The menopause is associated with the progression of dyslipidemia, as the sudden decrease in estrogen causes metabolic changes, which affects lipid profiles (6,7).

Depending on age and sex, the relationship between thyroid dysfunction and lipid profiles may differ. A few studies have tried to investigate this relationship according to age and sex, with conflicting results (8 –12). Furthermore, most of these studies evaluated lipid profiles according to changes in the thyrotropin (TSH) level (8 –10) without considering the free thyroxine (fT4) level and/or categorical thyroid dysfunction.

The Korean National Health and Nutrition Examination Survey (KNHANES) is an ongoing surveillance system that assesses the health and nutritional status of the Korean population, and it provides data on the prevalence of major chronic diseases in Korea as far back as 1998 (13). In the most recent KNHANES VI (2013–2015), thyroid function test was newly included.

The current analysis aimed to investigate the association between categorical thyroid dysfunction and lipid profiles according to age and sex using data from the KNHANES.

Materials and methods

Data source and study population

This study was performed using data from the KNHANES VI, which took place from 2013 to 2015. It is a nationwide, cross-sectional survey conducted by the Korea Centers for Disease Control and Prevention that uses stratified, multistage clustered probability sampling to select a representative sample of the civilian, non-institutionalized Korean population (13). Research participants were selected using two-stage stratified cluster sampling of the population and housing census data. The KNHANES obtained written informed consent from every participant prior to completing the survey, and the present study used secondary anonymized data for the analysis. The study protocol was approved by the Institutional Review Board of the Asan Medical Center, Seoul, Korea.

Lipid profile testing has been conducted in subjects since 1998. Thyroid function testing has been conducted in one third of all subjects, approximately 2400 persons aged ≥10 years, annually from 2013 to 2015. A total of three years represents the entire population of Korea. The population was selected by stratified subsampling considering the number of inhabitants in each year.

A total of 4709 participants >30 years underwent both thyroid function testing and lipid profile testing. Among them, 69 participants who had a prior history of thyroid disease and a history of taking medicines that could influence thyroid function were excluded. Therefore, 4640 participants were eligible for the study. To analyze the association between thyroid dysfunction and lipid profile, 4275 participants without a previous history of dyslipidemia or taking dyslipidemia medication were evaluated.

Laboratory measurements

Blood samples were obtained from each participant's antecubital vein in the morning after fasting for at least eight hours. The samples were properly processed, immediately refrigerated, and sent to the Central Testing Institute in Seoul, Korea, within 24 hours.

As previously reported, serum TSH and fT4 were measured with an electrochemiluminescence immunoassay (Roche Diagnostics, Mannheim, Germany) (14). Briefly, TSH was measured with an E-TSH kit (Roche Diagnostics), and the TSH reference interval was determined to be between the 2.5th and 97.5th percentile of the serum TSH levels of the reference population, as previously reported (15). Serum fT4 was measured using an E-Free T4 kit (Roche Diagnostics), and the reference range was 0.89–1.76 ng/mL. The measurements of TSH and fT4 met the criteria of the College of American Pathologists (16).

Lipid profiles were measured with a Hitachi Automatic Analyzer 7600 (Hitachi, Tokyo, Japan) using commercially available kits (Sekisui, Osaka, Japan). Serum total cholesterol (TC) and triglycerides (TG) were measured by enzymatic methods, and serum LDLC and high-density lipoprotein cholesterol (HDLC) were measured by homogeneous enzymatic colorimetric methods. LDLC was calculated using Friedewald's formula in individuals with a TG of <200 mg/dL in 2013 and 2014 (17). LDLC was directly measured in selected participants in 2013 and 2014 (participants with a TG of ≥200 mg/dL) and all participants in 2015. HDLC was calibrated with suggested methods, as previously described (18).

Definitions

Participants were classified into four groups according to age (cutoff age of 55 years) and sex. Disease-free participants were defined as having no prior history of thyroid disease or thyroid cancer and no history of taking medications that could influence thyroid function. This definition was adopted from a previous study from a U.S. population using the National Health and Nutrition Examination Survey III (19).

Thyroid dysfunction was classified into seven groups as follows: (i) OHypo: overt hypothyroidism, elevated serum TSH levels >6.8 mIU/L with low fT4 levels; (ii) SCHypo (TSH ≥10): subclinical hypothyroidism, serum TSH levels ≥10 mIU/L with normal fT4 levels; (iii) SCHypo (TSH <10): subclinical hypothyroidism, serum TSH levels <10 mIU/L with normal fT4 levels; (iv) euthyroid: euthyroidism, within normal range of serum TSH and fT4 levels; (v) SCHyper (TSH >0.1): subclinical hyperthyroidism, serum TSH levels >0.1 mIU/L with normal fT4 levels; (vi) SCHyper (TSH ≤0.1): subclinical hyperthyroidism, serum TSH levels ≤0.1 mIU/L with normal fT4 levels; and (vii) OHyper: overt hyperthyroidism, serum TSH levels <0.6 mIU/L with high fT4 levels. Hypothyroidism includes OHypo, SCHypo (TSH ≥10), and SCHypo (TSH <10), and hyperthyroidism includes OHyper, SCHyper TSH (≤0.1), and SCHyper (TSH >0.1).

Lipid profiles were divided into several groups in order to analyze the weighted prevalence of thyroid dysfunction according to lipid values referenced in the criteria of the National Cholesterol Education Program Adult Treatment Panel III (1). TC was classified into three groups: TC <170 mg/dL, TC 170–199 mg/dL, and TC ≥200 mg/dL. LDLC was classified into three groups: LDLC <100 mg/dL, LDLC 100–129 mg/dL, and LDLC ≥130 mg/dL. TG was classified into three groups: TG <100 mg/dL, TG 100–149 mg/dL, and TG ≥150 mg/dL. HDLC was classified into two groups: high HDLC (>40 mg/dL in men and >50 mg/dL in women) and low HDLC (<40 mg/dL in men and <50 mg/dL in women). Dyslipidemia was defined as a TC ≥200 mg/dL, LDLC ≥130 mg/dL, TG ≥150 mg/dL, and low HDLC.

Statistical analysis

R v3.4.0, the R libraries survey, and Cario were used for statistical analysis (R Foundation for Statistical Computing; www.R-project.org). All statistics were calculated using sample weights assigned to sample participants by the KNHANES. The sample weights were constructed to attain unbiased estimates representing the general Korean population, with consideration for its stratified multistage probability sampling design of each survey year (13). Logarithmic transformation of TSH values was used to normalize the distribution. The TSH reference interval and the weighted prevalence of thyroid dysfunction according to lipid profiles were analyzed after calculating weighted values. Continuous variables are described as means with standard error of the mean (SEM) and were analyzed by Student's t-test. Categorical variables are presented as frequencies with percentages and were analyzed by the chi-square test. A binary logistic regression model was used to evaluate the risk factors for dyslipidemia. In the risk analysis, age was analyzed by five-year intervals. All p-values were two-sided, and values <0.05 were considered statistically significant.

Results

Lipid profiles according to age and sex

Among 4640 study subjects, the number of study participants taking dyslipidemia medication differed by age for both sexes (p < 0.001 for both men and women; Table 1). Among older women (aged ≥55 years), 20.5% were taking dyslipidemia medications. In study participants who were not taking dyslipidemia medication, the lipid profiles according to each age group showed discordant results by sex. In male participants, TC, LDLC, and HDLC did not differ significantly according to age (p = 0.052, p = 0.705, and p = 0.082, respectively). However, in female participants, all lipid profiles were significantly different according to age (p < 0.001).

Table 1.

Lipid Profiles of Study Subjects by Different Age and Sex Groups

	Younger men		Older men			Younger women		Older women
	n	%	n	%	p-Value	n	%	n	%	p-Value
Study subjects	1388		896			1467		889
Subjects with dyslipidemia medication	40	2.8	105	12.3	<0.001	42	3.3	178	20.5	<0.001
Subjects without dyslipidemia medication
TC					0.052					<0.001
<170	310	21.5	217	23.4		434	29.1	101	11.1
170–199	491	35.5	305	32.4		512	35.2	223	24.7
≥200	547	40.2	269	31.8		479	32.5	387	43.7
LDLC					0.705					<0.001
<100	387	27.2	243	25.2		496	32.8	128	14.0
100–129	530	38.0	316	35.5		575	40.0	252	27.6
≥130	431	31.9	232	27.0		354	23.9	331	37.9
TG					0.003					<0.001
<100	385	27.5	267	30.4		853	57.1	283	30.7
100–149	349	25.9	231	25.2		307	20.6	214	25.0
≥150	614	43.8	293	32.0		265	19.0	214	23.9
HDLC					0.082					<0.001
High	1006	72.3	560	61.6		871	59.3	352	38.4
Low	342	24.9	231	26.1		554	37.4	359	41.2

Participants were divided according to the age of 55 in each sex. Younger men/women are men/women younger than 55 years, and older men/women are men/women older than 55 years.

The percentages are shown considering sampling weights.

TC, total cholesterol; LDLC, low-density lipoprotein cholesterol; TG, triglyceride; HDLC, high-density lipoprotein cholesterol; High HDLC, >40 mg/dL in men and >50 mg/dL in women; Low HDLC, <40 mg/dL in men and <50 mg/dL in women.

Baseline characteristics of study participants not taking dyslipidemia medication

The baseline characteristics of the 4275 participants not taking dyslipidemia medication who were analyzed in the association between thyroid dysfunction and lipid profiles are described in Table 2. The mean age of the total study population was 48.8 years (SEM = 0.2 years), and no difference was observed between the mean ages of women and men (p = 0.130). The mean body mass index (BMI) of the study participants was significantly lower among women than men (p < 0.001). Overall, 5.4% of participants had a history of diabetes, and more men (6.5%) had diabetes than women (4.4%; p = 0.034). Women had more thyroid dysfunction than men (p < 0.001). TC and LDLC levels were not significantly different between women and men. However, TG and HDLC were significantly different between women and men (p < 0.001 for both TG and HDLC).

Table 2.

Baseline Characteristics of Study Participants Without Dyslipidemia Medication

	Total (n = 4275)	Men (n = 2139)	Women (n = 2136)	p-Value
Age, years	48.8 (0.2)	48.6 (0.2)	49.1 (0.2)	0.130
BMI, kg/m²	24.1 (0.1)	24.5 (0.1)	23.7 (0.1)	<0.001
History of diabetes	214 (5.4%)	131 (6.5%)	83 (4.4%)	0.034
Thyroid dysfunctions				<0.001
OHypo	38 (1.0%)	13 (0.6%)	25 (1.4%)
SCHypo	141 (3.3%)	48 (2.3%)	93 (4.3%)
SCHyper	115 (3.1%)	52 (2.9%)	63 (3.4%)
OHyper	33 (0.7%)	8 (0.3%)	25 (1.1%)
Lipid profiles
TC, mg/dL	194.2 (0.6)	193.6 (0.8)	194.7 (0.9)	0.343
LDLC, mg/dL	117.6 (0.5)	116.7 (0.7)	118.6 (0.8)	0.066
TG, mg/dL	148.5 (2.3)	177.2 (3.8)	118.6 (2.1)	<0.001
HDLC, mg/dL	50.1 (0.2)	47.1 (0.3)	53.1 (0.3)	<0.001

Continuous variables are presented as means (standard error of the mean), and categorical variables are presented as numbers (percentages). Percentages are shown considering sampling weights.

BMI, body mass index; OHypo, overt hypothyroidism; SCHypo, subclinical hypothyroidism; SCHyper, subclinical hyperthyroidism; OHyper, overt hyperthyroidism.

Association between the weighted prevalence of thyroid dysfunction and lipid profiles

Figure 1A and Supplementary Table S1 (Supplementary Data are available online at www.liebertpub.com/thy) show the weighted prevalence of each thyroid dysfunction according to lipid profiles in the total study population without dyslipidemia medication (n = 4275). The weighted prevalence of thyroid dysfunction was significantly different among those with different TC and LDLC levels (p = 0.003 and p = 0.021, respectively). The prevalence of hypothyroidism was greater in subjects with higher TC levels (3.4% for TC <170 mg/dL, 3.5% for TC 170–199 mg/dL, and 5.6% for TC ≥200 mg/dL). Subjects with higher TC levels had a lower prevalence of hyperthyroidism (4.9% for TC <170 mg/dL, 4.0% for TC 170–199 mg/dL, and 3.0% for TC ≥200 mg/dL). These patterns were similar for LDLC: the prevalence of hypothyroidism was higher in subjects with higher LDLC levels (3.3% for LDLC <100 mg/dL, 4.5% for LDLC 100–129 mg/dL, and 5.0% for LDLC ≥130 mg/dL), and the prevalence of hyperthyroidism was lower in subjects with higher LDLC levels (4.6% for LDLC <100 mg/dL, 4.4% for LDLC 100–129 mg/dL, and 2.5% for LDLC ≥130 mg/dL).

FIG. 1.

Weighted prevalence of thyroid dysfunction according to lipid profiles. Each bar graph indicates the weighted prevalence of each thyroid dysfunction according to the different levels of each lipid profile in (A) the total study participants without dyslipidemia medication, (B) in men, (C) in younger women (age <55 years), and (D) in older women (age ≥55 years). Euthyroidism participants are not shown in this bar graph. The p-value indicates the association between all thyroid dysfunctions and each lipid profile, and is shown next to the bar graph of each lipid profile. OHypo, overt hypothyroidism; SCHypo (TSH ≥10), subclinical hypothyroidism with TSH ≥10 mIU/L; SCHypo (TSH <10), subclinical hypothyroidism with TSH <10 mIU/L; SCHyper (TSH >0.1), subclinical hyperthyroidism with TSH >0.1 mIU/L; SCHyper (TSH ≤0.1), subclinical hyperthyroidism with TSH ≤0.1 mIU/L; OHyper, overt hyperthyroidism; TC, total cholesterol; LDLC, low-density lipoprotein cholesterol; TG, triglyceride; HDLC, high-density lipoprotein cholesterol; High HDLC, >40 mg/dL in men and >50 mg/dL in women; Low HDLC, <40 mg/dL in men and <50 mg/dL in women.

Figure 1B and Supplementary Table S2 show the relationship between thyroid dysfunction and lipid profiles in each sex. In male participants, there was no significant difference in the weighted prevalence of thyroid dysfunction among participants with different levels of all lipid profiles (p = 0.562 for TC, p = 0.452 for LDLC, p = 0.384 for TG, and p = 0.871 for HDLC). However, in female participants, the weighted prevalence of thyroid dysfunction was different among participants with different levels of TC, LDLC, and TG (p = 0.007 for TC, p = 0.016 for LDLC, and p = 0.044 for TG).

Female participants were divided into two groups using the age cutoff of 55 years. Figure 1C and D and Supplementary Table S3 show the association between thyroid dysfunction and lipid profiles in women by age. The weighted prevalence of hypothyroidism and hyperthyroidism was 5.9% and 5.1% in younger women and 5.7% and 3.8% in older women, respectively. In younger women, the weighted prevalence of thyroid dysfunction was significantly different among participants with different levels of TC, LDLC, and TG (p = 0.013 for TC, p = 0.007 for LDLC, and p = 0.007 for TG). In this group, the prevalence of hypothyroidism was 9.0% in participants with TC ≥200 mg/dL and 10.4% in participants with LDLC ≥130 mg/dL. However, no relationship was found between thyroid dysfunction and any lipid profile in older women (p = 0.221 for TC, p = 0.213 for LDLC, p = 0.885 for TG, and p = 0.975 for HDLC; Fig. 1D).

Risk factor analysis for dyslipidemia in male participants

Thyroid function (TSH or fT4) was not associated with any lipid profiles in male participants (Table 3). Age was associated with TG ≥150 mg/dL and low HDLC (p = 0.005 and p = 0.022, respectively). BMI was identified as an independent risk factor for dyslipidemia (p < 0.001 for TC ≥200 mg/dL, LDLC ≥130 mg/dL, TG ≥150 mg/dL, and low HDLC). HbA1c was an independent risk factor for TG ≥150 mg/dL (p < 0.001) and low HDCL (p < 0.001).

Table 3.

Risk Factor Analysis for Dyslipidemia in Male Participants

	Univariate analysis			Multivariate analysis
	OR	CI	p-Value	OR	CI	p-Value
TC ≥200 mg/dL
Log TSH	1.05	0.79–1.40	0.744	1.01	0.73–1.40	0.932
fT4	0.90	0.52–1.57	0.713	0.92	0.49–1.71	0.793
Age^a	0.96	0.92–1.00	0.039	0.97	0.92–1.01	0.152
BMI	1.09	1.05–1.12	<0.001	1.08	1.05–1.12	<0.001
HbA1c	1.03	0.92–1.16	0.574	1.02	0.90–1.15	0.775
LDLC ≥130 mg/dL
Log TSH	1.06	0.78–1.44	0.697	1.12	0.79–1.58	0.521
fT4	1.25	0.71–2.20	0.448	1.45	0.76–2.75	0.252
Age^a	0.97	0.93–1.01	0.140	0.99	0.94–1.03	0.561
BMI	1.10	1.06–1.14	<0.001	1.10	1.06–1.14	<0.001
HbA1c	1.04	0.92–1.16	0.566	1.01	0.88–1.14	0.937
TG ≥150 mg/dL
Log TSH	1.45	1.08–1.96	0.014	1.25	0.89–1.76	0.193
fT4	0.44	0.25–0.78	0.006	0.51	0.26–1.02	0.057
Age^a	0.94	0.91–0.98	0.005	0.94	0.90–0.98	0.005
BMI	1.19	1.15–1.24	<0.001	1.18	1.13–1.22	<0.001
HbA1c	1.32	1.15–1.52	<0.001	1.30	1.13–1.49	<0.001
Low HDLC
Log TSH	0.93	0.68–1.28	0.649	0.78	0.56–1.10	0.156
fT4	0.56	0.28–1.10	0.094	0.64	0.30–1.32	0.227
Age^a	1.05	1.00–1.10	0.049	1.06	1.01–1.11	0.022
BMI	1.16	1.12–1.20	<0.001	1.16	1.12–1.20	<0.001
HbA1c	1.30	1.15–1.47	<0.001	1.22	1.09–1.36	<0.001

Age was analyzed by five-year intervals.

OR, odds ratio; CI, confidence interval; TSH, thyrotropin; fT4, free thyroxine.

Risk factor analysis for dyslipidemia in female participants

In female participants, TSH was found to be an independent risk factor for TC ≥200 mg/dL and LDLC ≥130 mg/dL (odds ratio [OR] = 1.59 [confidence interval (CI) 1.21–2.09], p = 0.001; and OR = 1.84 [CI 1.35–2.49], p < 0.001, respectively; Table 4). Meanwhile, only TG ≥150 mg/dL was independently associated with fT4 (OR = 0.30 [CI 0.13–0.67], p = 0.003). Age and BMI were identified as independent risk factors for dyslipidemia in all lipid profiles. HbA1c was an independent risk factor for TG ≥150 mg/dL (p = 0.004) and low HDCL (p < 0.001).

Table 4.

Risk Factor Analysis for Dyslipidemia in Female Participants

	Univariate analysis			Multivariate analysis
	OR	CI	p-Value	OR	CI	p-Value
TC ≥200 mg/dL
Log TSH	1.79	1.43–2.25	<0.001	1.59	1.21–2.09	0.001
fT4	0.37	0.22–0.61	<0.001	0.66	0.36–1.19	0.168
Age^a	1.27	1.22–1.34	<0.001	1.25	1.19–1.32	<0.001
BMI	1.09	1.06–1.12	<0.001	1.05	1.02–1.09	0.003
HbA1c	1.34	1.11–1.61	0.002	1.01	0.86–1.20	0.862
LDLC ≥130 mg/dL
Log TSH	1.87	1.45–2.40	<0.001	1.84	1.35–2.49	<0.001
fT4	0.48	0.29–0.79	0.004	0.96	0.58–1.61	0.883
Age^a	1.31	1.25–1.38	<0.001	1.28	1.22–1.35	<0.001
BMI	1.10	1.07–1.14	<0.001	1.06	1.03–1.10	0.001
HbA1c	1.42	1.18–1.71	<0.001	1.07	0.90–1.25	0.450
TG ≥150 mg/dL
Log TSH	1.56	1.14–2.14	0.006	1.13	0.79–1.61	0.510
fT4	0.21	0.10–0.45	<0.001	0.30	0.13–0.67	0.003
Age^a	1.18	1.12–1.24	<0.001	1.11	1.05–1.17	<0.001
BMI	1.21	1.17–1.26	<0.001	1.17	1.13–1.22	<0.001
HbA1c	1.72	1.37–2.15	<0.001	1.32	1.09–1.58	0.004
Low HDLC
Log TSH	0.95	0.76–1.19	0.658	0.95	0.72–1.25	0.699
fT4	1.08	0.76–1.52	0.681	1.24	0.77–2.01	0.376
Age^a	1.14	1.09–1.19	<0.001	1.06	1.01–1.11	0.010
BMI	1.16	1.13–1.20	<0.001	1.13	1.10–1.17	<0.001
HbA1c	2.01	1.60–2.51	<0.001	1.54	1.25–1.90	<0.001

Age was analyzed by five-year intervals.

Discussion

This was a nationwide cohort study to evaluate the association between thyroid dysfunction and lipid profiles according to age and sex across all age groups. Most previous studies have evaluated lipid profiles according to TSH level (8 –10) without considering fT4 level and/or categorical thyroid dysfunction. The participants were classified by categorical thyroid dysfunction considering both TSH and fT4.

It was found that the relationship between thyroid dysfunction and lipid profiles was obvious in women but not in men. Some studies have shown similar results. Women had a stronger association between thyroid dysfunction and lipid profiles than men did (11,12,20 –22). According to a study conducted in middle-aged men and women, increased plasma TSH was accompanied by increased TC in women only (20). These findings suggest that the effect of thyroid dysfunction on lipid profiles may differ by sex. The inconsistent results by sex could be explained by the fact that the prevalence of thyroid dysfunction is relatively lower in men than in women. Another explanation could be the high prevalence of diabetes in men, which could obscure the association between thyroid dysfunction and dyslipidemia. Participants taking dyslipidemia medication were excluded, and so many diabetic participants would have been excluded. However, in a study based on four population-based cohorts in Germany, the relationship between thyroid dysfunction and dyslipidemia was significant in both men and women. In that study, the association was found in TC, LDLC, and TG for women and in TG for men (8). In the multivariate analysis of the present study, the association between dyslipidemia and TG disappeared in both men and women. This result suggests that TG is more affected by other factors such as BMI or age than by the change in TSH level. Other previous studies, the EPIC-Norfolk Study and the Hunt Study, demonstrated that increased TSH was associated with less favorable lipid profiles in both sexes (9,10). However, both studies were conducted only in participants with TSH levels within the reference range.

In the present study, all lipid profiles were significantly different according to age (older or younger than 55 years of age) in women. This can be explained by postmenopausal changes in lipid profiles and by the progression of dyslipidemia with the aging process in women (5 –7). In women, the menopause is associated with an unfavorable effect on lipid metabolism because of the sudden decline in the protective effect of estrogen (7). Thus, the menopause might have a more powerful effect on lipid metabolism than on thyroid dysfunction.

The association between hypothyroidism and dyslipidemia is well documented (3,23,24). The prevalence of hypothyroidism in patients with dyslipidemia was reported to range from 3.7% to 5.2% (3,25,26). The results of the present study are similar. In this study, the prevalence of hypothyroidism ranged from 4.6% to 5.6% in the highest stratum of TC, LDLC, and TG, and participants with dyslipidemia had a higher prevalence of hypothyroidism (5.6% in TC ≥200 mg/dL vs. 3.4% in TC <200 mg/dL, and 5.1% in LDLC 130 mg/dL vs. 3.9% in LDLC <130 mg/dL). This can be explained by the decrease in cholesterol excretion and a marked increase in apo B lipoproteins due to the decrease in both catabolism and turnover by a reduced number of LDLC receptors on the liver surface in hypothyroidism (27,28). In the hyperthyroid state, cholesterol excretion is augmented and LDLC turnover is increased (27,29), which leads to a decrease in TC and LDLC.

The association between thyroid dysfunction and HDLC is still controversial (11,12,27,30). Thyroid hormone is believed to increase the activity of hepatic lipase (HL), which leads to the decline of HDLC level and the effect of thyroid hormone on the reverse transport of cholesterol by influencing the activity of HL and cholesteryl ester transfer protein (CETP), which modulates the distribution of HDLC (27,31,32). This study did not find any association between the prevalence of thyroid dysfunction and HDLC levels.

The upper limit of TSH in this study was 6.8 mIU/L, which was higher than those from other countries (19,33,34). The National Academy of Clinical Biochemistry guidelines did not consider geographical differences in iodine intake. Korea is a country with a high iodine intake, and high urine iodine levels are related to high serum TSH levels (14). Similar results were also reported from China (35). Because of the high TSH upper limit, the prevalence of SCH in this study is low compared with previous studies (36). Future guidelines for TSH reference ranges should consider the nutritional iodine intake status to define reference populations.

This study has some limitations. First, the association between thyroid dysfunction and lipid profile was found only in younger women. This result should be cautiously applied to the general population. This study could have been statistically underpowered because the prevalence of thyroid dysfunction was higher in women than in men. Second, a selected registry population was used (i.e., the Korean population >30 years). Thus, the results are inadequate to generalize to other populations. More research with larger populations and diverse ethnic groups is needed to elucidate sex differences in the association between thyroid dysfunction and lipid profiles. Third, only single points of thyroid hormone and lipid levels were used, meaning that the effects of thyroid function changes on lipid profiles could not be determined. Fourth, the benefits or risks of the treatment of thyroid dysfunction on lipid profiles could not be determined. Fifth, this study only evaluated the association between thyroid dysfunction and lipid profiles. Considering the fact that dyslipidemia may aggravate early atherosclerotic lesions and increase the risk of cardiovascular disease (CVD), further studies on the association between thyroid dysfunction and CVD risk according to age and sex are warranted. The recent Rotterdam Study suggested that the optimal health ranges for thyroid function based on CVD could differ according to age and sex (37). Lastly, a relatively large proportion of older women (age ≥55 years; 20%) who were taking dyslipidemia medication were excluded in the analysis of thyroid dysfunction and lipid profiles. This may have affected the results of the association between thyroid dysfunction and lipid profiles in the older group.

In conclusion, the prevalence of thyroid dysfunction was significantly different according to lipid profiles. This association was obvious in women, especially in younger women (<55 years). The relationship between thyroid dysfunction and lipid profiles differed by age and sex.

Acknowledgments

This work was supported by the Korean Association of Internal Medicine Research Grant 2014.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. 2001. Executive summary of the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA, 285:2486–2497.

Committee for the Korean Guidelines for the Management of Dyslipidemia. 2016. 2015 Korean guidelines for the management of dyslipidemia: executive summary (English translation). Korean Circ J, 46:275–306.

Diekman

, Lansberg

, Kastelein

, Wiersinga

. 1995. Prevalence and correction of hypothyroidism in a large cohort of patients referred for dyslipidemia. Arch Intern Med, 155:1490–1495.

Godsland

, Wynn

, Crook

, Miller

. 1987. Sex, plasma lipoproteins, and atherosclerosis: prevailing assumptions and outstanding questions. Am Heart J, 114:1467–1503.

Swiger

, Martin

, Blaha

, Toth

, Nasir

, Michos

, Gerstenblith

, Blumenthal

, Jones

. 2014. Narrowing sex differences in lipoprotein cholesterol subclasses following mid‐life: the very large database of lipids (VLDL‐10B). J Am Heart Assoc, 3:e000851.

Fonseca

MIH

, da Silva

, Ferreira

SRG

. 2017. Impact of menopause and diabetes on atherogenic lipid profile: is it worth to analyse lipoprotein subfractions to assess cardiovascular risk in women?. Diabetol Metab Syndr, 9:22.

Schaefer

, Lamon-Fava

, Cohn

, Schaefer

, Ordovas

, Castelli

, Wilson

. 1994. Effects of age, gender, and menopausal status on plasma low density lipoprotein cholesterol and apolipoprotein B levels in the Framingham Offspring Study. J Lipid Res, 35:779–792.

Meisinger

, Ittermann

, Tiller

, Agger

, Nauck

, Schipf

, Wallaschofski

, Jorgensen

, Linneberg

, Thiery

, Kluttig

, Greiser

, Werdan

, Burkhardt

, Volzke

. 2014. Sex-specific associations between thyrotropin and serum lipid profiles. Thyroid, 24:424–432.

Boekholdt

, Titan

, Wiersinga

, Chatterjee

, Basart

, Luben

, Wareham

, Khaw

. 2010. Initial thyroid status and cardiovascular risk factors: the EPIC-Norfolk prospective population study. Clin Endocrinol (Oxf), 72:404–410.

10.

Asvold

, Vatten

, Nilsen

, Bjoro

. 2007. The association between TSH within the reference range and serum lipid concentrations in a population-based study. The HUNT Study. Eur J Endocrinol, 156:181–186.

11.

Tognini

, Polini

, Pasqualetti

, Ursino

, Caraccio

, Ferdeghini

, Monzani

. 2012. Age and gender substantially influence the relationship between thyroid status and the lipoprotein profile: results from a large cross-sectional study. Thyroid, 22:1096–1103.

12.

Walsh

, Bremner

, Bulsara

, O'Leary

, Leedman

, Feddema

, Michelangeli

. 2005. Thyroid dysfunction and serum lipids: a community-based study. Clin Endocrinol (Oxf), 63:670–675.

13.

Kweon

, Kim

, Jang

M-j

, Kim

, Choi

, Chun

, Khang

Y-H

, Oh

. 2014. Data resource profile: the Korea National Health and Nutrition Examination Survey (KNHANES). Int J Epidemiol, 43:69–77.

14.

Jeon

, Kim

, Kwon

, Kim

, Park

, Oh

, Han

, Kim

, Shong

, Kim

. 2017. Excessive iodine intake and thyrotropin reference interval: data from the Korean National Health and Nutrition Examination Survey. Thyroid, 27:967–972.

15.

Kim

, Kim

, Woo

, Kim

, Cho

, Kim

, Shin

M-H

, Park

H-L

, Oh

, Chung

. 2017. Thyroid stimulating hormone reference range and prevalence of thyroid dysfunction in the Korean population: Korea National Health and Nutrition Examination Survey 2013 to 2015. Endocrinol Metab, 32:106–114.

16.

Ministry of the Interior. Seegene Medical Foundation, Laboratory Medicine of the Korea National Health and Nutrition Examination Survey VI (2013–2015). Available at: https://knhanes.cdc.go.kr/knhanes/eng/index.do (accessed March 6, 2017 ).

17.

Friedewald

, Levy

, Fredrickson

. 1972. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem, 18:499–502.

18.

Yun

, Song

, Ji

, Kim

, Park

, Song

, Park

, Kim

, Nho

, Oh

. 2017. Calibration of high-density lipoprotein cholesterol values from the Korea National Health and Nutrition Examination Survey data, 2008 to 2015. Ann Lab Med, 37:1–8.

19.

Hollowell

, Staehling

, Flanders

, Hannon

, Gunter

, Spencer

, Braverman

. 2002. 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, 87:489–499.

20.

Bindels

, Westendorp

, Frolich

, Seidell

, Blokstra

, Smelt

. 1999. 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), 50:217–220.

21.

Meng

, Liu

, Zhang

, Liu

, Song

, Tan

, Jia

, Zhang

, Wang

, He

, Ren

, Zhu

, He

, Wang

, Li

, Zheng

, Hu

, Liu

, Upadhyaya

, Zhou

, Zhang

. 2015. Gender and age impact on the association between thyroid-stimulating hormone and serum lipids. Medicine, 94:e2186.

22.

Santos-Palacios

, Brugos-Larumbe

, Guillen-Grima

, Galofre

. 2013. A cross-sectional study of the association between circulating TSH level and lipid profile in a large Spanish population. Clin Endocrinol (Oxf), 79:874–881.

23.

Canaris

, Manowitz

, Mayor

, Ridgway

. 2000. The Colorado thyroid disease prevalence study. Arch Intern Med, 160:526–534.

24.

Jung

, Sung

, Shin

, Rhee

, Lee

, Kim

, Kang

, Kim

, Lee

, Park

, Kim

. 2003. Thyroid dysfunction and their relation to cardiovascular risk factors such as lipid profile, hsCRP, and waist hip ratio in Korea. Korean J Intern Med, 18:146–153.

25.

Tagami

, Kimura

, Ohtani

, Tanaka

, Hata

, Saito

, Miyazaki

, Araki

, Tanaka

, Yonezawa

, Sawamura

, Ise

, Ogo

, Shimbo

, Shimatsu

, Naruse

. 2011. Multi-center study on the prevalence of hypothyroidism in patients with hypercholesterolemia. Endocr J, 58:449–457.

26.

Willard

, Leung

, Pearce

. 2014. Thyroid function testing in patients with newly diagnosed hyperlipidemia. JAMA Intern Med, 174:287–289.

27.

Duntas

. 2002. Thyroid disease and lipids. Thyroid, 12:287–293.

28.

Duntas

, Brenta

. 2012. The effect of thyroid disorders on lipid levels and metabolism. Med Clin North Am, 96:269–281.

29.

Muller

, Seitz

. 1984. Thyroid hormone action on intermediary metabolism. Part III. Protein metabolism in hyper- and hypothyroidism. Klin Wochenschr, 62:97–102.

30.

Valdemarsson

, Hansson

, Hedner

, Nilsson-Ehle

. 1983. Relations between thyroid function, hepatic and lipoprotein lipase activities, and plasma lipoprotein concentrations. Acta Endocrinol (Copenh), 104:50–56.

31.

Tan

, Shiu

, Kung

. 1998. Plasma cholesteryl ester transfer protein activity in hyper- and hypothyroidism. J Clin Endocrinol Metab, 83:140–143.

32.

Lam

, Chan

, Yeung

. 1986. High-density lipoprotein cholesterol, hepatic lipase and lipoprotein lipase activities in thyroid dysfunction—effects of treatment. Q J Med, 59:513–521.

33.

National Committee for Clinical Laboratory Standards. 2000. How to Define, Determine and Utilize Reference Intervals in the Clinical Laboratory: Proposed Guidelines. National Committee for Clinical Laboratory Standards, Villanova, PA.

34.

Vadiveloo

, Donnan

, Murphy

, Leese

. 2013. Age- and gender-specific TSH reference intervals in people with no obvious thyroid disease in Tayside, Scotland: the Thyroid Epidemiology, Audit, and Research Study (TEARS). J Clin Endocrinol Metab, 98:1147–1153.

35.

Guan

, Shan

, Teng

, Li

, Teng

, Jin

, Yu

, Fan

, Chong

, Yang

, Dai

, Yu

, Li

, Chen

, Zhao

, Shi

, Hu

, Mao

, Gu

, Yang

, Chen

, Tong

, Wang

, Gao

, Li

, Teng

. 2008. Influence of iodine on the reference interval of TSH and the optimal interval of TSH: results of a follow-up study in areas with different iodine intakes. Clin Endocrinol, 69:136–141.

36.

Hennessey

, Espaillat

. 2015. Subclinical hypothyroidism: a historical view and shifting prevalence. Int J Clin Pract, 69:771–782.

37.

Chaker

, Korevaar

TIM

, Rizopoulos

, Collet

, Volzke

, Hofman

, Rodondi

, Cappola

, Peeters

, Franco

. 2017. Defining optimal health range for thyroid function based on the risk of cardiovascular disease. J Clin Endocrinol Metab, 102:2853–2861.