Clinical hypothyroidism (CH) results from a lack of thyroid hormones or the inadequate actions of thyroid hormones at target tissues. Subclinical hypothyroidism (SCH) is defined as serum thyrotrophin (TSH) levels above a statistically defined upper limit of a reference range when serum free thyroxine (FT4) and free tri-iodothyronine (FT3) concentrations are at normal levels [
Angiopoietin-like proteins (Angptls) share common protein domain characteristics with angiopoietins, which are a family of secreted glycoproteins expressed in the liver. Angptls family members include eight subtypes, Angptl1–8. Angptl3, along with Angptl4 and 8, have emerged as important regulators in lipoprotein metabolism, through the inhibition of lipoprotein lipase (LPL) [
Currently, some studies have indicated that adipocyte- or hepatocyte-derived metabolic regulators, such as fibroblast growth factor 19 (FGF19), fibroblast growth factor21 (FGF21), and Irisin also induce changes in patient thyroid dysfunction [
Despite well-established associations between metabolic and thyroid function, few studies have investigated if the underlying associations between dyslipidemia and thyroid dysfunction are caused by changes in Angptls. In general, autoimmune thyroiditis has two clinical stages, which are classified according to the degree of thyroid dysfunction: subclinical hypothyroidism and hypothyroidism. In the present study, we investigate Angptl3, 4, and 8 levels in CH and SCH patients with different thyroid function statuses and interpret associations between Angptls, lipid profiles, and thyroid function indices, including thyroid hormone levels and thyroid autoantibodies.
This study was performed at the Endocrine Metabolic and Immune Diseases Center at the Beijing Lu-He Hospital, affiliated to Capital Medical University, from February 2017 to November 2018. All research protocols were approved by the ethics committee from the Institutional Review Board of Capital Medical University (number 2019LH-WZ-006). All participants were informed of the study and provided written informed consent prior to participation. The study complied with the Declaration of Helsinki.
A total of 88 subjects were enrolled including age and sex-matched subjects with hypothyroidism (CH, n=29), subclinical hypothyroidism (SCH, n=30), and healthy participants (Controls, n=29). Fasted blood samples were collected from all participants. Subjects were excluded based on acute or chronic disease status, for example, diabetes mellitus, heart failure, coronary heart disease, acute-chronic renal disease, cerebrovascular event, malignancy, liver diseases, rheumatic diseases, alcohol, and smoking.
Collected blood was allowed to coagulate at 4°C and serum was separated by centrifugation for 15 min at 3000 rpm. Serum TSH, FT4, FT3 levels, thyroid peroxidase antibody (TPOAb), thyroglobulin antibody (TgAb) levels, LDL-c, HDL-c, TC, and TG levels were measured using electrochemiluminescence immunoassay by Cobas Elecsys 601 (Roche Diagnostics, Switzerland). CH was diagnosed as serum TSH concentrations above the statistically defined upper limit of the reference range, TSH > 4.2 uIU/mL, along with a subnormal FT4 level, FT4 < 0.93 ng/dL. SCH was diagnosed as serum TSH concentrations above the statistically defined upper limit of the reference range, TSH > 4.2 uIU/mL, and serum FT4 within the normal reference range. Controls were defined by TSH, FT4, TPOAb, and TgAb within normal ranges.
Serum sample aliquots were stored at -80°C until ELISA was performed. Serum angiopoietin-like proteins—Angptl3, Angptl4, and Angptl8—were measured by ELISA (Immuno-Biological Laboratories International GmBH, Japan). Experiments were conducted in compliance with manufacturer’s instructions, with an intra-assay coefficient of variation (CV) of ≤ 4.8% and an interassay CV of ≤ 7.2%. All samples were performed in duplicate and repeated if the CV > 15%.
For normally distributed variables, the data were expressed as mean ± standard deviations (SD). For nonnormally distributed variables, the data were expressed as medians with interquartile range, and categorical variables were expressed as frequencies. The Mann-Whitney rank sum test or
The characteristics of patients with CH, SCH, and controls are described (Table
Study population information.
Group | Healthy control | Subclinical hypothyroidism | Overt hypothyroidism | |
---|---|---|---|---|
Sex (F/M) | 29 (16/13) | 30 (23/7) | 29 (24/5) | 0.05 |
Age (year) | 43±11 | 44±13 | 44±15 | 0.43 |
T3 (ng/mL) | 1.30 (1.20-1.41) | 0.95 (0.87-1.02) | 0.89 (0.82-0.97) | <0.001 |
T4 (ug/dL) | 9.78 (1.8-2.76) | 6.61 (6.26-6.97) | 4.13 (3.49-4.77) | <0.001 |
FT3 (pg/mL) | 3.60 (3.33-3.86) | 2.82 (2.64-3.00) | 2.16 (1.98-2.35) | <0.001 |
FT4 (ng/dL) | 1.39 (1.30-1.48) | 1.07 (1.01-1.13) | 0.70 (0.58-0.81) | <0.001 |
TSH (uIU/mL) | 2.46 (1.8-2.76) | 14.19 (10.36-19.07) | 16.48 (9.86-45.17) | <0.01 |
TPOAb (U/mL) | 18.71(15-22.79) | 177.45 (23.26-236.68) | 117.45 (57.69-352.23) | <0.01 |
TgAb (U/mL) | 16.08 (12.18-30.69) | 468.6 (206.2-557.5) | 437.15 (234.13-1326.2) | <0.01 |
Insulin (pmol/L) | 10.44±4.82 | 10.69±7.54 | 11.58±8.77 | >0.05 |
HDL-C (mmol/L) | 1.15 (1.07-1.22) | 1.41 (1.28-1.53) | 1.25 (1.15-1.35) | >0.05 |
LDL-C (mmol/L) | 2.33(1.77-2.59) | 2.58(2.21-2.91) | 2.73 (2.44-3.62) | <0.001 |
TG (mmol/L) | 1.01 (0.83-1.19) | 1.45 (1.16-1.73) | 1.37 (1.08-1.66) | >0.05 |
TC (mmol/L) | 4.00 (3.76-4.24) | 4.58 (4.10-5.06) | 4.87 (4.51-5.24) | <0.05 |
Angptl3 (ng/mL) | 4.94 (4.63-5.26) | 5.36 (5.03-5.69) | 5.49 (5.11-5.88) | <0.05 |
Angptl4 (pg/mL) | 144.80 (116.61–172.99) | 171.21 (138.54–203.89) | 146.89 (122.05–171.73) | >0.05 |
Angptl8 (pmol/L) | 0.63(0.57-0.69) | 0.69 (0.64-0.74) | 0.67(0.61-0.74) | <0.05 |
F: female; M, male.
a. Data are expressed as mean±SD.
b. Data are expressed as median with interquartile range.
Serum Angptl3 (a), Angptl4 (b), and Angptl8 (c) levels in SCH and CH subjects. (a) Compared with healthy controls, Angptl3 levels in SCH and CH were significantly elevated; (b) Angptl4 showed no significant differences when compared with controls; (c) Serum Angptl8 was increased in the CH group when compared to the control group and was also higher in the SCH group. Data are expressed as mean ± SEM.
To investigate the relationship between Angptl subtypes and thyroid function, we performed a correlation analysis between thyroid function, blood lipid indices, and Angptl3 and 8 plasma levels. Angptl3 was positively correlated with HDL (r = 0.431,
Angptl3 was positively correlated with HDL.
Angptl3 was negatively correlated with TT3 and FT3.
Angptl8 was positively correlated with TG and CHO.
Angptl8 was negatively correlated with TT3.
Both Angptl3 and Angptl8 levels appeared to be related to thyroid dysfunction, so we performed a receiver operating characteristic (ROC) curve analysis to evaluate the diagnostic performance of both Angptl subtypes in discriminating thyroid dysfunction. The area under curve (AUC) for detecting thyroid dysfunction based on Angptl3 was 0.635 (optimal cutoff value, 341.05
Receiver operating characteristic (ROC) analysis. (a) ROC curves for predicting SCH and CH for Angptl3. (b) ROC curves for predicting SCH and CH for Angptl8. (c) ROC curves for predicting SCH and CH for both Angptl3 and Angptl8. AUC, area under the receiver operating characteristic curve.
This study evaluated the associations between Angptl3, 4, and 8 and thyroid dysfunction. Here, we demonstrate for the first time that circulating Angptl3 and 8 levels were significantly increased in patients with CH and SCH, when compared with the control group. However, Angptl4 levels remained unchanged. Serum Angptl3 was positively correlated with HDL and negatively correlated with TT3. Due to increasing rates of metabolic disease, more and more studies have focused on their molecular elucidation. Some studies have demonstrated that multiple circulating cytokines are associated with metabolic disorders, particularly lipid and glucose diseases. Among these studies, several groups have demonstrated that Angptl3 and 8 serve as important lipid metabolism factors closely associated with metabolic disease. A study by Abu-Farha
The hepatic overexpression of Angptl8 causes hypertriglyceridemia and increased insulin secretion. In addition, Angptl8-knockout mice have reduced adipose tissue and lower triglyceride levels when compared to wild-type mice [
CH and SCH are associated with some metabolic disorders including insulin resistance or high insulin levels, dyslipidemia, obesity, and endothelial dysfunction [
Furthermore, by using ROC analysis, we confirmed that Angptl3 and Angptl8 could predict thyroid dysfunction. Overall, we have shown that Angptl3 and 8 could be promising biomarkers and therapeutic targets for thyroid dysfunction.
There are limitations to our study. Firstly, it was cross-sectional, and it only addressed associations between changes in circulating Angptl3 and 8 and thyroid dysfunction. Changes in serum Angptl 3 and 8 levels after chronic replacement of thyroid hormones were not assessed. Furthermore, other factors such as adipokines or hepatokines, hormones, or other parameters that may have affected Angptl 3 and 8 levels were not examined. In addition, the sample size was small. Based on these limitations, more large-scale population studies are needed to understand the relationship between Angptl and thyroid dysfunction.
In summary, our study demonstrated that circulating Angptl3 and Angptl8 levels are increased in patients with CH and SCH. Thus, it appears that thyroid insufficiency may impact on circulating Angptl3 and 8 levels. To highlight the significance of Angptl subtypes in metabolic disease pathogenesis, further studies are required. These studies could elucidate the role of Angptl subtypes in hypothyroidism and uncover mechanistic connections between thyroid dysfunction and metabolic disease.
All data generated or analyzed during this study are included in this published article.
This study was approved by the ethics committee of the Institutional Review Board of Capital Medical University. The study was complied with reference to the Declaration of Helsinki.
Informed consent was obtained from all participants in this study.
The authors declare they have no conflicts of interest.
Longyan Yang, Ruili Yin, and Zongwei Wang equally contributed to the study.
This study was funded by the Science and Technology Committee of Tongzhou District (number KJ2018CX008-17); Natural Science Foundation of Beijing (number 7194282); Natural Cultivation Fund of Capital Medical University (number PYZ2017051).