Abstract
Background
This study aimed to examine the associations of thyroid hormone sensitivity indices, including free triiodothyronine-to-free thyroxine (FT3/FT4) ratio, thyroid feedback quantile-based index by FT4 (TFQIFT4), thyroid-stimulating hormone index (TSHI), and thyrotrophic thyroxine resistance index (TT4RI) with all-cause mortality in euthyroid adults.
Methods
The study included 6243 euthyroid adults from the National Health and Nutrition Examination Survey (NHANES) 2007–2012. FT3/FT4 ratio, TFQIFT4, TSHI, and TT4RI were calculated. The multivariable Cox proportional hazard regression, restricted cubic spline (RCS), and subgroup analysis were conducted.
Results
Individuals in fourth quartile (Q4) had lower all-cause mortality than those in first quartile (Q1) of FT3/FT4 ratio (hazard ratio (HR): 0.70, 95% CI: 0.51, 0.94). Regarding TFQIFT4, individuals in Q4 of TFQIFT4 had a 43% higher all-cause mortality than those in Q1 (HR: 1.43, 95% CI: 1.05, 1.96) (P < 0.05, all). Compared with participants in Q1, no associations of TSHI and TT4RI with mortality were found. TFQIFT4 was linearly and positively associated with mortality. However, the FT3/FT4 ratio showed a U-shaped association with mortality.
Conclusions
Increased risk for all-cause mortality was positively associated with TFQIFT4, suggesting that increased risk for all-cause mortality was associated with decreased central sensitivity to thyroid hormones. Furthermore, the FT3/FT4 ratio showed a U-shaped association with mortality, with an inflection point at 0.5. However, more cohort studies are needed to validate the conclusions.
Introduction
The thyroid hormones play an essential role in regulating glucose and lipid metabolism and mitochondrial function, increasing energy expenditure and thermogenesis (1). Disordered thyroid hormones may result in detrimental health consequences, including, but not limited to, coronary heart disease (2) and cancer (3), and in intensive care unit (ICU) mortality (4). In addition, even slight fluctuations in thyroid function within the normal range could be associated with adverse health outcomes comparable to those observed in cases of overt or subclinical hypothyroidism (5, 6). There are abundant studies associating thyroid-stimulating hormone (TSH) and free thyroxine (FT4) with mortality (7, 8, 9, 10, 11, 12, 13, 14). A Chinese study of 264 sarcopenic patients aged 80 years and older found that in a euthyroid population, those with lower levels of free triiodothyronine (FT3) had a higher risk of mortality (15). An American survey also found the association of ‘low–normal’ thyroid function and subclinical hypothyroidism with increased all-cause mortality (16). However, there are still some studies that come to a different conclusion. A study conducted in Italy indicated that a normal–low TSH is an independent risk factor for all-cause mortality in euthyroid elderly adults, while neither FT3 nor FT4 exhibited any association with mortality (17). The direct correlation between TSH levels or thyroid hormones and all-cause mortality necessitates further clarification.
Given the inconsistent relationship between thyroid function and mortality, it is plausible to assume that these conflicting results are due to abnormal sensitivity to thyroid hormones in certain populations (18). Thyroid hormone sensitivity, evaluated by complex indices using T3, T4, and TSH, has received increasing attention recently. FT4 and FT3 exhibit a physiological and inverse correlation with TSH, regulated by a highly sensitive feedback loop known as the hypothalamus–pituitary–thyroid (HPT) axis (19). Laclaustra et al. proposed a novel approach for calculating the thyroid hormone central sensitivity index, the thyroid feedback quantile-based index (TFQIFT4), which utilizes FT4 and TSH levels (20). Additional thyroid hormone central sensitivity indices include the thyrotroph thyroxine resistance index (TT4RI) and TSH index (TSHI). The FT3/FT4 ratio reflects deiodinase activity and is a proxy for peripheral thyroid hormone sensitivity (21). Central resistance phenomena affect the feedback loop set point within the central nervous system, while peripheral resistance phenomena diminish the metabolic effects of hormones (20). Impaired thyroid sensitivity was associated with metabolic disorders, including osteoporosis (22), carotid plaque (23), and metabolic dysfunction-associated fatty liver disease (24). A study has indicated that elevated levels of resistance to thyroid hormone indices are associated with diabetes-related mortality in the general population of the United States (20). Additionally, a Spanish study that enrolled 3750 individuals showed that, in a euthyroid population, TFQI may be associated with a progressive increase in all-cause mortality (25). Nonetheless, investigations on the potential relationship of TFQIFT4 and all-cause mortality in euthyroid participants are scarce. Consequently, in this study, we aim to determine the predictive capacity of thyroid sensitivity indices for long-term outcomes in euthyroid participants who are representative of adults in the United States. Our hypothesis posits that this approach may facilitate the identification of crucial risk parameters, thereby enabling early risk stratification and future prognostication.
Methods
Study design and population enrollment
Data from the National Health and Nutrition Examination Survey (NHANES) conducted during the period of 2007–2012 were used for this study. The NHANES is a nationally representative survey of the civilian noninstitutionalized U.S. population, applying a stratified, multistage, clustered probability sampling design (26). The National Center for Health Statistics ethics review board approved this survey. All participants provided informed and written consent. Data were analyzed from January to April 2023. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guideline.
In total, 30,442 people participated in the investigation. We excluded participants who were less than 20 years old (n = 12,729), with missing thyroid profile data (n = 8943), had thyroid dysfunction (n = 1233), had a history of thyroid disease (n = 565), reported the use of medications that may alter serum thyroid function (amiodarone, thyroid hormone replacement, and/or antithyroid drugs (n = 20)), had a history of cancer (n = 614), were pregnant or lactating at the time of blood draw (n = 86), and lacked data for a death record (n = 9). Finally, 6243 participants were included in the current study for analysis (Fig. 1).
Flowchart of study participants.
Citation: European Thyroid Journal 13, 1; 10.1530/ETJ-23-0130
Definition of serum thyroid function
In the NHANES datasets, serum TSH levels were measured using a microparticle enzyme immunoassay, serum FT4 levels were measured using a two-step enzyme immunoassay, and serum FT3 levels were measured using a competitive binding immunoenzymatic assay (27, 28). The normal ranges for TSH, FT4, and FT3 were defined as 0.39–4.60 mIU/L, 0.6–1.6 ng/dL (7.8–20.8 pmol/L), and 2.5–3.9 pg/mL (3.85–6.006 pmol/L), respectively (27). Participants with serum TSH, FT4, and FT3 concentrations within the normal range were considered to be euthyroid.
The FT3/FT4 ratio was applied to reflect the converting activity of peripheral T4 to T3. Regarding the central sensitivity to thyroid hormones, three indices were evaluated. TFQIFT4 was achieved by the algorithm TFQI = cumulative distribution function (cdfFT4) – (1 – cdfTSH); TSHI was calculated as ln TSH (mIU/L) + 0.1345 × fT4 (pmol/L); and TT4RI was calculated as fT4 (pmol/L × TSH (mIU/L). The value of TFQIFT4 ranged from −1 to 1 (20). Negative values indicate higher thyroid hormone sensitivity, and positive values indicate lower sensitivity. For TSHI and TT4RI, the higher the values, the lower the central sensitivity to thyroid hormone (18).
Covariates
Directed acyclic graphs (29) were employed, representing the existing literature to select a minimally sufficient set of covariates to adjust for confounding (Supplementary Fig. 1, see the section on supplementary materials given at the end of this article).
Baseline demographics, anthropometric assessment, comprehensive laboratory data, and lifestyle factors were acquired using established protocols (30). Race/ethnicity was categorized as Mexican American, other Hispanic, non-Hispanic Black, non-Hispanic White, and other races. Education level was classified into high school education or lower, or education beyond high school (31). Smoking was classified as current smokers (those who have smoked ≥100 cigarettes and smoke currently), ex-smokers (who have smoked ≥100 cigarettes but do not smoke currently), and nonsmokers (24). Alcohol consumption was reported as standard drinks and converted to grams by multiplying by 14. It was considered an abusive drink if >30 g/day for men and >20 g/day for women (32). Body mass index (BMI) was calculated by the formula weight/height (2). Laboratory methods for measurements of thyroid peroxidase antibody (TPOAb), thyroglobulin antibody (TgAb), hemoglobin A1c (HbA1c), creatinine, total cholesterol (TC), and triglycerides (TG) were reported in detail elsewhere (33). To calculate the estimated glomerular filtration rate (eGFR), the Chronic Kidney Disease Epidemiology Collaboration equation was used: GFR = 141 × min(Scr/κ, 1)α × max(Scr/κ, 1)−1.209 × 0.993Age × 1.018 (if woman) × 1.159 (if black), where Scr is serum creatinine in mg/dL, κ is 0.7 for women and 0.9 for men, α is −0.329 for women and −0.411 for men, min indicates the minimum of Scr/κ or 1, and max indicates the maximum of Scr/κ or 1 (34).
Outcome ascertainment
All participants aged over 20 years in NHANES 2007–2012 underwent a passive mortality follow-up through December 31, 2019. The primary outcome was all-cause mortality using death certification information provided by the National Death Index after record matching by Social Security number, name, date of birth, race and ethnicity, sex, state of birth, and state of residence (28). The follow-up time for each participant was calculated from the day of TSH measurement to the date of death or the end of the follow-up period.
Statistical analysis
R Version 4.1.0 was used for statistical analysis using suitable sampling weights for each analysis, as suggested by the NCHS. The mean (s.d.) or median (interquartile range) was calculated for continuous variables, and proportions were calculated for categorical variables (%). Differences between the groups were calculated using the Student’s t-test, the Mann–Whitney U test, or the chi-squared test. Bonferroni-adjusted P value was provided.
We used multivariable Cox proportional hazard regression models to estimate the hazard ratios (HRs) and 95% CIs of mortality for FT3/FT4, TSHI, TT4RI, or TFQIFT4 quartiles. The proportional hazards assumption was examined using Schoenfeld’s test. Survival analysis was assessed by a log-rank test for all identified predictive factors. Kaplan–Meier plots represented survival across FT3/FT4 or TFQIFT4 ratio quartiles. We used restricted cubic spline (RCS) models fitted for Cox proportional hazards models with three knots. The full model was adjusted for sex, age, race, the ratio of family income to poverty, BMI, TPOAb, TgAb, SBP, HbA1c, TC, TG, eGFR, alcohol consumption, education, and smoking status.
To determine whether associations between every one s.d. increase of thyroid sensitivity index and mortality within each polypharmacy strata differed for thyroid sensitivity index according to sex (men, women), age (<60, ≥60 years), and BMI (<25, 30 > BMI ≥ 25, ≥30 kg/m2), we examined for a possible interaction with these factors by adding cross-product terms of the stratified variables and continuous exposure variable to the final model and performing a Wald test (35). The full model was adjusted for sex, age, race, the ratio of family income to poverty, BMI, TPOAb, TgAb, SBP, HbA1c, TC, TG, eGFR, alcohol consumption, education, and smoking status. In the analysis stratified by sex, sex was excluded from the model. All statistical tests were two-sided and considered significant at a P < 0.05.
Results
General characteristics of the participants
The baseline characteristics of participants are presented in Table 1. Of 6243 participants, 3273 were men, and the mean (s.d.) age was 44.57 ± 15.57 years. Compared with participants who had deceased, those still alive were younger in baseline and had lower SBP, HbA1c, FT4, TFQIFT4, TSHI, and TT4RI. However, they exhibited a higher ratio of family income to poverty, DBP, FT3, FT3/FT4 ratio, and eGFR (P < 0.05, all). Additionally, more than half of the participants alive were nonsmokers in the baseline.
Baseline characteristics of participants. Data are expressed as proportions (%) for categorical variables and as mean (s.d.) or median (interquartile range) for continuous variables. Bonferroni-adjusted P value is provided.
| Overall | Alive | Deceased | P | |
|---|---|---|---|---|
| n | 6243 | 5506 | 737 | |
| Age (years) | 44.57 (15.57) | 42.77 (14.43) | 63.78 (14.23) | <0.001 |
| Men (%) | 52.3 | 51.4 | 62.3 | <0.001 |
| Race (%) | <0.001 | |||
| Mexican American | 9.0 | 9.5 | 3.6 | |
| Non-Hispanic Black | 11.5 | 11.5 | 12.4 | |
| Non-Hispanic White | 66.0 | 65.3 | 74.4 | |
| Other Hispanic | 6.0 | 6.2 | 3.5 | |
| Other | 7.5 | 7.6 | 6.1 | |
| Education (%) | <0.001 | |||
| Less than high school | 26.5 | 25.6 | 36.1 | |
| High school | 33.3 | 33.4 | 32.6 | |
| Beyond high school | 40.2 | 41.0 | 31.3 | |
| Abusive drink (%) | 0.4 | 0.4 | 0.4 | 0.893 |
| Smoking (%) | <0.001 | |||
| Nonsmokers | 55.0 | 56.7 | 37.7 | |
| Former smokers | 22.5 | 21.6 | 32.9 | |
| Current smokers | 22.4 | 21.8 | 29.4 | |
| Ratio of family income to poverty | 2.86 (1.41, 4.90) | 2.92 (1.43, 5.00) | 2.12 (1.25, 3.48) | <0.001 |
| BMI (kg/m2) | 28.47 (6.57) | 28.46 (6.51) | 28.61 (7.19) | 0.726 |
| SBP (mm Hg) | 121.69 (16.79) | 120.50 (15.72) | 134.35 (21.82) | <0.001 |
| DBP (mm Hg) | 71.28 (12.10) | 71.46 (11.66) | 69.30 (15.90) | 0.003 |
| HbA1c (%) | 5.58 (0.88) | 5.53 (0.84) | 6.05 (1.19) | <0.001 |
| FT3 (pmol/L) | 4.91 (0.47) | 4.92 (0.46) | 4.72 (0.50) | <0.001 |
| FT4 (pmol/L) | 10.18 (1.56) | 10.14 (1.52) | 10.60 (1.86) | <0.001 |
| TSH (mIU/L) | 1.56 (1.09, 2.23) | 1.55 (1.09, 2.20) | 1.72 (1.17, 2.54) | <0.001 |
| FT3/FT4 ratio | 0.49 (0.08) | 0.50 (0.08) | 0.46 (0.09) | <0.001 |
| TFQIFT4 | -0.04 (0.38) | -0.05 (0.38) | 0.08 (0.41) | <0.001 |
| TSHI | 1.80 (0.53) | 1.78 (0.53) | 1.93 (0.56) | <0.001 |
| TT4RI | 15.61 (10.95, 22.24) | 15.42 (10.88, 21.96) | 17.92 (11.92, 26.11) | <0.001 |
| TPOAb (IU/mL) | 0.60 (0.30, 1.40) | 0.60 (0.30, 1.40) | 0.50 (0.30, 1.20) | 0.064 |
| TgAb (IU/mL) | 0.60 (0.60, 0.60) | 0.60 (0.60, 0.60) | 0.60 (0.60, 0.60) | 0.104 |
| eGFR )mL/min per 1.73 m2) | 97.59 (20.41) | 99.30 (19.29) | 79.32 (23.01) | <0.001 |
| TC (mmol/L) | 5.08 (1.05) | 5.09 (1.03) | 4.97 (1.17) | 0.083 |
| TG (mmol/L) | 2.41 (2.63) | 2.43 (2.63) | 2.21 (2.57) | 0.148 |
BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; FT3, free triiodothyronine; FT4, free thyroxine; HbA1c, glycated hemoglobin; SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; TFQIFT4, thyroid feedback quantile-based index by FT4; TgAb, thyroglobulin antibody; TPOAb, thyroid peroxidase antibody; TSH, thyroid-stimulating hormone; TSHI, thyroid-stimulating hormone index; TT4RI, thyrotrophic thyroxine resistance index.
Associations of thyroid sensitivity indices with mortality
The median duration of follow-up for mortality ascertainment was 133 (interquartile range, 106, 144) months, from which 737 all-cause deaths were identified. Figure 2 reveals the estimated HRs and CIs of sensitivity indices in relation to all-cause mortality. When known demographic variables and traditional risk factors, including sex, age, race, the ratio of family income to poverty, BMI, TPOAb, TgAb, SBP, HbA1c, TC, TG, eGFR, alcohol consumption, education, and smoking status, were taken into consideration, individuals in fourth quartile (Q4) had lower all-cause mortality than those within first quartile (Q1) of FT3/FT4 ratio (HR: 0.70, 95% CI: 0.51, 0.94). Regarding TFQIFT4, individuals in Q4 of TFQIFT4 had a 43% higher all-cause mortality than those in Q1 (HR: 1.43, 95% CI: 1.05, 1.96). However, no significant associations were found between TSHI or TT4RI and all-cause mortality (P > 0.05, all). We examined the association between low serum FT3 and mortality without finding a statistical difference (Q4 vs Q1, HR: 1.16, 95% CI: 0.89, 1.52, P > 0.05).
Associations between quartiles of thyroid sensitivity indices and all-cause mortality. The model was adjusted for sex, age, race, ratio of family income to poverty, BMI, TPOAb, TgAb, SBP, HbA1c, TC, TG, eGFR, alcohol consumption, education, and smoking status. FT3/FT4 ratio, free triiodothyronine-to-free thyroxine ratio; TFQIFT4, the thyroid feedback quantile-based index calculated by FT4; BMI, body mass index; SBP, systolic blood pressure; TG, triglycerides; TC, total cholesterol; TPOAb, thyroid peroxidase antibody; TgAb, thyroglobulin antibody; eGFR, estimated glomerular filtration rate; HbA1c, glycated hemoglobin.
Citation: European Thyroid Journal 13, 1; 10.1530/ETJ-23-0130
In addition, as showed in Kaplan–Meier survival curves (Fig. 3), participants in Q1 of FT3/FT4 ratio and Q4 of TFQIFT4 had increasing all-cause mortality.
Long-term survival of participants. Kaplan–Meier estimators for the total sample stratified by (A) FT3/FT4 ratio and (B) TFQIFT4. FT3/FT4 ratio, free triiodothyronine-to-free thyroxine ratio; TFQIFT4, the thyroid feedback quantile-based index calculated by FT4.
Citation: European Thyroid Journal 13, 1; 10.1530/ETJ-23-0130
Nonlinear relationships of thyroid sensitivity indices with mortality
In Fig. 4, we used RCS to flexibly model and visualize the relationship between thyroid sensitivity indices and all-cause mortality. We found that the FT3/FT4 ratio showed a U-shaped association with mortality. Figure 4A showed a substantial reduction of the risk within the lower range of the FT3/FT4 ratio, which reached the lowest risk around 0.5 and then increased after that (P for nonlinearity <0.001). However, TFQIFT4 were linearly and positively associated with mortality (Fig. 4B) after adjusting for sex, age, race, the ratio of family income to poverty, BMI, TPOAb, TgAb, SBP, HbA1c, TC, TG, eGFR, alcohol consumption, education, and smoking status.
Nonlinear relationships of thyroid sensitivity indices with all-cause mortality. Restricted cubic spline (RCS) analysis was conducted. A. FT3/FT4 ratio and mortality. B. TFQIFT4 and mortality. The model was adjusted for sex, age, race, ratio of family income to poverty, BMI, TPOAb, TgAb, SBP, HbA1c, TC, TG, eGFR, alcohol consumption, education, and smoking status. FT3/FT4 ratio, free triiodothyronine-to-free thyroxine ratio; TFQIFT4, the thyroid feedback quantile-based index calculated by FT4; BMI, body mass index; SBP, systolic blood pressure; TG, triglycerides; TC, total cholesterol; TPOAb, thyroid peroxidase antibody; TgAb, thyroglobulin antibody; eGFR, estimated glomerular filtration rate; HbA1c, glycated hemoglobin.
Citation: European Thyroid Journal 13, 1; 10.1530/ETJ-23-0130
Subgroup analyses
To examine the robustness and potential variations in different subgroups, we performed subgroup analysis stratified by sex, age, and BMI groups. The results are shown in Fig. 5.
Subgroup analysis stratified by sex, age, and BMI. A. FT3/FT4 ratio and mortality. B. TFQIFT4 and mortality. The associations were adjusted for age, race, BMI, TPOAb, TgAb, SBP, HbA1c, TC, TG, eGFR, alcohol consumption, education, and smoking status in subgroup analysis stratified by sex. In subgroup analysis stratified by age and BMI groups, the associations were further adjusted for sex and the variables mentioned above. FT3/FT4 ratio, free triiodothyronine-to-free thyroxine ratio; TFQIFT4, the thyroid feedback quantile-based index calculated by FT4; BMI, body mass index; SBP, systolic blood pressure; TG, triglycerides; TC, total cholesterol; TPOAb, thyroid peroxidase antibody; TgAb, thyroglobulin antibody; eGFR, estimated glomerular filtration rate; HbA1c, glycated hemoglobin.
Citation: European Thyroid Journal 13, 1; 10.1530/ETJ-23-0130
The associations were adjusted for age, race, the ratio of family income to poverty, BMI, TPOAb, TgAb, SBP, HbA1c, TC, TG, eGFR, alcohol consumption, education, and smoking status in subgroup analysis stratified by sex. In subgroup analysis stratified by age and BMI groups, the associations were further adjusted for sex and the variables mentioned above. A significant relationship between the FT3/FT4 ratio and mortality was detected in men and elder subjects (P < 0.05, all) (Fig. 5A). A significant relationship between TFQIFT4 and mortality was found in men, elderly subjects, and normal and overweight subjects (P < 0.05, all) (Fig. 5B). No interaction on FT3/FT4 ratio and TFQIFT4 was found in sex, age, and BMI stratification (P > 0.05, all).
Discussion
In this study, we investigated the associations of various thyroid sensitivity indices with all-cause mortality among euthyroid individuals residing in communities across the United States. We found that increased risk for all-cause mortality was positively associated with TFQIFT4, indicating that decreased central sensitivity to thyroid hormones is associated with an increased risk for all-cause mortality. Furthermore, the FT3/FT4 ratio demonstrated a robust U-shaped association with mortality, with an inflection point at 0.5.
Given the intricate interplay within the HPT axis, combined indicators offer a more comprehensive depiction of the association between thyroid hormones and mortality than individual indices. Previous studies indicated that high-normal FT4 is associated with an increased mortality risk (36, 37). However, an inverse relationship between TSH and outcomes was not found, which deviates from the anticipated negative feedback loop. Our findings are congruent with previous research highlighting the potential impact of thyroid hormone sensitivity (25).
The findings of our study observed U-shaped associations between FT3/FT4 ratio level and mortality, which could be interpreted as indicating an optimal FT3/FT4 ratio level that balances the lowest mortality risk. An Italian study focusing on older adults discovered an inverse correlation between FT3/FT4 ratio value and frailty degree and mortality risk using logistic analysis (38). However, nonlinear relationships of FT3/FT4 ratio values with frailty degree and mortality were not examined. An earlier study using NHANES 2007–2012 data found that the risk of all-cause mortality dropped as the FT3/FT4 ratio increased in a reversed J-shaped pattern (39). The different results may be attributed to the different follow-up periods, inclusion and exclusion criteria (participants with abnormal thyroid function were also excluded in our study), and adjustments for different confounding factors. Thus, further cohort studies with a larger sample size and a longer follow-up period are needed. In the current study, an association between mortality and TFQIFT4 was found, but not other thyroid hormone sensitivity indexes (TSHI and TT4RI). Similar results were also found in the relationship between thyroid hormone sensitivity indexes and diabetes, where TFQIFT4 was associated with diabetes. However, TSHI and TT4RI were not (20). One possible reason may be that TFQIFT4 is more stable than TT4RI and TSHI, as it does not reach extreme values (20), indicating that the mechanisms of resistance to thyroid hormones were complex and challenging (40). Previous studies have made it clear that low T3 syndrome is associated with death. Low T3 syndrome is defined as T3 levels below the lower limit of normal. However, we included in this study a population with FT3, FT4, and TSH, all within the reference range. Therefore, FT3/FT4 is superior to FT3 in a population with normal thyroid function.
Some plausible biological mechanisms may underlie the association between thyroid sensitivity indices and mortality. TFQIFT4, reflecting the sensitivity to thyroid hormones in the pituitary, may be different from thyroid status in other peripheral tissues. Changes in the set point of the HPT axis can regulate the transformation of T4 into an appropriate state for pituitary perception. However, alterations in T4 levels may not be suitable for peripheral organs or tissues if the sensitivity of peripheral organs is maintained, causing disorders in the cardiovascular and endocrine systems (25). Taking RTHβ, for instance, it is a clinical syndrome characterized by impaired sensitivity to thyroid hormone (TH). It is caused by mutations in the thyroid hormone receptor beta (THRB) gene, which modulates the negative regulation of the HPT axis (41). RTH (40) is characterized by elevated serum TH levels without TSH suppression. In contrast, these individuals exhibit symptoms of thyrotoxicosis in other tissues that typically express TRα1 predominantly, such as the heart, bone, muscle, and adipose tissue, leading to various metabolic disorders (25, 42). In addition, excessive TH levels can produce reactive oxygen species (43), which might trigger and maintain cell damage, induce physiological abnormalities, and facilitate disease progression (44, 45). Furthermore, participants with a risk of mortality might be in a frailty status, such as starvation, chronic inflammation, impaired liver function, and sarcopenia, which may play a role in the impairment of peripheral T4 deiodination (38).
Our study has several limitations. First, the absence of serial thyroid function tests and ultrasonography data precluded the consideration of any trends or changes in thyroid function during the follow-up period. Secondly, the relatively small sample size and short follow-up period, coupled with a limited number of events available for analysis, restricted the exploration of specific analyses by causes of mortality, which should be assessed in studies with larger sample sizes. Thirdly, despite controlling for key personal characteristics, we cannot establish causal associations or rule out the risk bias due to the observational cohort nature of the design. Finally, the biological mechanism of thyroid sensitivity and mortality could not be determined in this study.
Conclusion
In conclusion, the TFQIFT4, reflecting thyroid hormone central sensitivity, is an independent predictor of all-cause mortality in euthyroid adults, while the FT3/FT4 ratio, reflecting thyroid hormone peripheral sensitivity, exhibits a U-shaped relationship with mortality. This information should be validated in further cohort studies with larger populations and longer follow-up periods.
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/ETJ-23-0130.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the study reported.
Funding
This work was supported by the National Natural Science Foundation of China (82200960, 82170800).
Ethics approval and consent to participate
The original survey was approved by the NCHS Research Ethics Review Board, and informed consent for data collection and storage was obtained from all participants.
Data availability
The data supporting the study findings are available from the corresponding authors upon reasonable request.
Author contribution statement
HW and GY performed the conceptualization; GY, SL, and CS conducted the data analysis; XC, YJ, HD, QM, D W, and YH conducted the data acquisition; GY drafted the manuscript; HW and JS revised the manuscript and served as scientific advisors. The final manuscript was read and approved by all authors.
Acknowledgements
The authors thank all team members and participants in the NHANES study and support from the National Natural Science Foundation of China (82200960, 82170800).
References
- 1↑
Kim B. Thyroid hormone as a determinant of energy expenditure and the basal metabolic rate. Thyroid 2008 18 141–144. (https://doi.org/10.1089/thy.2007.0266)
- 2↑
Rodondi N, den Elzen WP, Bauer DC, Cappola AR, Razvi S, Walsh JP, Asvold BO, Iervasi G, Imaizumi M, Collet TH, et al.Subclinical hypothyroidism and the risk of coronary heart disease and mortality. JAMA 2010 304 1365–1374. (https://doi.org/10.1001/jama.2010.1361)
- 3↑
Tran TV, Kitahara CM, Leenhardt L, de Vathaire F, Boutron-Ruault MC, & Journy N. The effect of thyroid dysfunction on breast cancer risk: an updated meta-analysis. Endocrine-Related Cancer 2023 30. (https://doi.org/10.1530/ERC-22-0155)
- 4↑
Bourcier S, Coutrot M, Ferre A, Van Grunderbeeck N, Charpentier J, Hraiech S, Azoulay E, Nseir S, Aissaoui N, Messika J, et al.Critically ill severe hypothyroidism: a retrospective multicenter cohort study. Annals of Intensive Care 2023 13 15. (https://doi.org/10.1186/s13613-023-01112-1)
- 5↑
Taylor PN, Razvi S, Pearce SH, & Dayan CM. Clinical review: a review of the clinical consequences of variation in thyroid function within the reference range. Journal of Clinical Endocrinology and Metabolism 2013 98 3562–3571. (https://doi.org/10.1210/jc.2013-1315)
- 6↑
Walsh JP. Setpoints and susceptibility: do small differences in thyroid function really matter? Clinical Endocrinology 2011 75 158–159. (https://doi.org/10.1111/j.1365-2265.2011.04036.x)
- 7↑
Gussekloo J, van Exel E, de Craen AJM, Meinders AE, Frölich M, & Westendorp RGJ. Thyroid status, disability and cognitive function, and survival in old age. JAMA 2004 292 2591–2599. (https://doi.org/10.1001/jama.292.21.2591)
- 8↑
Bano A, Chaker L, Mattace-Raso FUS, van der Lugt A, Ikram MA, Franco OH, Peeters RP, & Kavousi M. Thyroid function and the risk of atherosclerotic cardiovascular morbidity and mortality: the Rotterdam study. Circulation Research 2017 121 1392–1400. (https://doi.org/10.1161/CIRCRESAHA.117.311603)
- 9↑
Du Puy RS, Poortvliet RKE, Mooijaart SP, den Elzen WPJ, Jagger C, Pearce SHS, Arai Y, Hirose N, Teh R, Menzies O, et al.Outcomes of thyroid dysfunction in people aged eighty years and older: an individual patient data meta-analysis of four prospective studies (towards understanding longitudinal international older people studies consortium). Thyroid 2021 31 552–562. (https://doi.org/10.1089/thy.2020.0567)
- 10↑
Pearce SHS, Razvi S, Yadegarfar ME, Martin-Ruiz C, Kingston A, Collerton J, Visser TJ, Kirkwood TB, & Jagger C. Serum thyroid function, mortality and disability in advanced old age: the Newcastle 85+ study. Journal of Clinical Endocrinology and Metabolism 2016 101 4385–4394. (https://doi.org/10.1210/jc.2016-1935)
- 11↑
Xu Y, Derakhshan A, Hysaj O, Wildisen L, Ittermann T, Pingitore A, Abolhassani N, Medici M, Kiemeney LALM, Riksen NP, et al.The optimal healthy ranges of thyroid function defined by the risk of cardiovascular disease and mortality: systematic review and individual participant data meta-analysis. The Lancet Diabetes and Endocrinology 2023 11 743–754. (https://doi.org/10.1016/S2213-8587(2300227-9)
- 12↑
Cappola AR, Arnold AM, Wulczyn K, Carlson M, Robbins J, & Psaty BM. Thyroid function in the euthyroid range and adverse outcomes in older adults. Journal of Clinical Endocrinology and Metabolism 2015 100 1088–1096. (https://doi.org/10.1210/jc.2014-3586)
- 13↑
Waring AC, Arnold AM, Newman AB, Bùzková P, Hirsch C, & Cappola AR. Longitudinal changes in thyroid function in the oldest old and survival: the cardiovascular health study all-stars study. Journal of Clinical Endocrinology and Metabolism 2012 97 3944–3950. (https://doi.org/10.1210/jc.2012-2481)
- 14↑
Cappola AR, Fried LP, Arnold AM, Danese MD, Kuller LH, Burke GL, Tracy RP, & Ladenson PW. Thyroid status, cardiovascular risk, and mortality in older adults. JAMA 2006 295 1033–1041. (https://doi.org/10.1001/jama.295.9.1033)
- 15↑
Zhang L, Tu YY, Zhao Z, Jin J, Tao J, & Zhang XY. Lower serum FT3 within the reference range is associated with mortality for older adults over 80 years of age with sarcopenia. BMC Geriatrics 2023 23 77. (https://doi.org/10.1186/s12877-023-03783-8)
- 16↑
Kim D, Vazquez-Montesino LM, Escober JA, Fernandes CT, Cholankeril G, Loomba R, Harrison SA, Younossi ZM, & Ahmed A. Low thyroid function in nonalcoholic fatty liver disease is an independent predictor of all-cause and cardiovascular mortality. American Journal of Gastroenterology 2020 115 1496–1504. (https://doi.org/10.14309/ajg.0000000000000654)
- 17↑
Ceresini G, Marina M, Lauretani F, Maggio M, Bandinelli S, Ceda GP, & Ferrucci L. Relationship between circulating thyroid-stimulating hormone, free thyroxine, and free triiodothyronine concentrations and 9-year mortality in euthyroid elderly adults. Journal of the American Geriatrics Society 2016 64 553–560. (https://doi.org/10.1111/jgs.14029)
- 18↑
Sun Y, Teng D, Zhao L, Shi X, Li Y, Shan Z, & Teng W. Impaired sensitivity to thyroid hormones is associated with hyperuricemia, obesity, and cardiovascular disease risk in subjects with subclinical hypothyroidism. Thyroid 2022 32 376–384. (https://doi.org/10.1089/thy.2021.0500)
- 19↑
Chaker L, Bianco AC, Jonklaas J, & Peeters RP. Hypothyroidism. Lancet 2017 390 1550–1562. (https://doi.org/10.1016/S0140-6736(1730703-1)
- 20↑
Laclaustra M, Moreno-Franco B, Lou-Bonafonte JM, Mateo-Gallego R, Casasnovas JA, Guallar-Castillon P, Cenarro A, & Civeira F. Impaired sensitivity to thyroid hormones is associated with diabetes and metabolic syndrome. Diabetes Care 2019 42 303–310. (https://doi.org/10.2337/dc18-1410)
- 21↑
Pasqualetti G, Schirripa M, Dochy E, Fassan M, Ziranu P, Puzzoni M, Scartozzi M, Alberti G, Lonardi S, Zagonel V, et al.Thyroid hormones ratio is a major prognostic marker in advanced metastatic colorectal cancer: results from the Phase III randomised CORRECT trial. European Journal of Cancer 2020 133 66–73. (https://doi.org/10.1016/j.ejca.2020.04.023)
- 22↑
Liu C, Hua L, Liu K, & Xin Z. Impaired sensitivity to thyroid hormone correlates to osteoporosis and fractures in euthyroid individuals. Journal of Endocrinological Investigation 2023 46 2017–2029. (https://doi.org/10.1007/s40618-023-02035-1)
- 23↑
Liu Y, Li Z, Yang T, Li L, Yu L, Liu F, Ni T, Gao S, Li C, Yang R, et al.Impaired sensitivity to thyroid hormones and carotid plaque in patients with coronary heart disease: a RCSCD-TCM study in China. Frontiers in Endocrinology 2022 13 940633. (https://doi.org/10.3389/fendo.2022.940633)
- 24↑
Wan H, Yu G, Xu S, Chen X, Jiang Y, Duan H, Lin X, Ma Q, Wang D, Liang Y, et al.Central sensitivity to free triiodothyronine with MAFLD and its progression to liver fibrosis in euthyroid adults. Journal of Clinical Endocrinology and Metabolism 2023 108 e687–e697. (https://doi.org/10.1210/clinem/dgad186)
- 25↑
Alonso SP, Valdes S, Maldonado-Araque C, Lago A, Ocon P, Calle A, Castano L, Delgado E, Menendez E, Franch-Nadal J, et al.Thyroid hormone resistance index and mortality in euthyroid subjects: Di@bet.Es study. European Journal of Endocrinology 2021 186 95–103. (https://doi.org/10.1530/EJE-21-0640)
- 26↑
Yu G, Liu L, Qin T, Luo Y, Song C, Chen X, Duan H, Jiang Y, Zeng H, Wan H, et al.Associations of serum iron status with MAFLD and liver fibrosis in the USA: a nationwide cross-section study. Biological Trace Element Research 2024 202 87–98. (https://doi.org/10.1007/s12011-023-03666-4)
- 27↑
Neves JS, Leitao L, Baeta Baptista R, Bigotte Vieira M, Magrico R, Viegas Dias C, Oliveira A, Falcao-Pires I, Lourenco A, Carvalho D, et al.Lower free triiodothyronine levels within the reference range are associated with higher cardiovascular mortality: an analysis of the NHANES. International Journal of Cardiology 2019 285 115–120. (https://doi.org/10.1016/j.ijcard.2019.03.009)
- 28↑
Inoue K, Ritz B, Brent GA, Ebrahimi R, Rhee CM, & Leung AM. Association of subclinical hypothyroidism and cardiovascular disease with mortality. JAMA Network Open 2020 3 e1920745. (https://doi.org/10.1001/jamanetworkopen.2019.20745)
- 29↑
Textor J, van der Zander B, Gilthorpe MS, Liskiewicz M, & Ellison GT. Robust causal inference using directed acyclic graphs: the R package “Dagitty.” International Journal of Epidemiology 2016 45 1887–1894. (https://doi.org/10.1093/ije/dyw341)
- 30↑
Wan H, Jiang Y, Yang J, Ma Q, Liu L, Peng L, Liu H, Xiong N, Guan Z, Yang A, et al.Sex-specific associations of the urinary fourteen-metal mixture with NAFLD and liver fibrosis among US adults: a nationally representative study. Ecotoxicology and Environmental Safety 2022 248 114306. (https://doi.org/10.1016/j.ecoenv.2022.114306)
- 31↑
Hohman EE, Corr TE, Kawasaki S, Savage JS, & Symons Downs D. Nutritional status differs by prescription opioid use among women of reproductive age: NHANES 1999–2018. Nutrients 2023 15. (https://doi.org/10.3390/nu15081891)
- 32↑
Wan H, Wang Y, Zhang H, Zhang K, Chen Y, Chen C, Zhang W, Xia F, Wang N, & Lu Y. Chronic lead exposure induces fatty liver disease associated with the variations of gut microbiota. Ecotoxicology and Environmental Safety 2022 232 113257. (https://doi.org/10.1016/j.ecoenv.2022.113257)
- 33↑
Centers for Disease Control and Prevention 2011. Nationa l Health and Nutrition Examination Survey (NHANES). U.S. Department of Health and Human Services. (https://www.cdc.gov/nchs/nhanes/participant.htm) Accessed April 25,2023.
- 34↑
Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, et al.A new equation to estimate glomerular filtration rate. Annals of Internal Medicine 2009 150 604–612. (https://doi.org/10.7326/0003-4819-150-9-200905050-00006)
- 35↑
Xia PF, Pan XF, Chen C, Wang Y, Ye Y, & Pan A. Dietary intakes of eggs and cholesterol in relation to all-cause and heart disease mortality: a prospective cohort study. Journal of the American Heart Association 2020 9 e015743. (https://doi.org/10.1161/JAHA.119.015743)
- 36↑
Groothof D, Flores-Guerrero JL, Nolte IM, Bouma HR, Gruppen EG, Bano A, Post A, Kootstra-Ros JE, Hak E, Bos JHJ, et al.Thyroid function and risk of all-cause and cardiovascular mortality: a prospective population-based cohort study. Endocrine 2021 71 385–396. (https://doi.org/10.1007/s12020-020-02397-z)
- 37↑
Yeap BB, Alfonso H, Hankey GJ, Flicker L, Golledge J, Norman PE, & Chubb SA. Higher free thyroxine levels are associated with all-cause mortality in euthyroid older men: the health in men study. European Journal of Endocrinology 2013 169 401–408. (https://doi.org/10.1530/EJE-13-0306)
- 38↑
Okoye C, Arosio B, Carino S, Putrino L, Franchi R, Rogani S, Cesari M, Mari D, Vitale G, Malara A, et al.The free triiodothyronine/free thyroxine ratio is associated with frailty in older adults: a longitudinal multisetting study. Thyroid 2023 33 169–176. (https://doi.org/10.1089/thy.2022.0422)
- 39↑
Lang X, Li Y, Zhang D, Zhang Y, Wu N, & Zhang Y. FT3/FT4 ratio is correlated with all-cause mortality, cardiovascular mortality, and cardiovascular disease risk: NHANES 2007–2012. Frontiers in Endocrinology 2022 13 964822. (https://doi.org/10.3389/fendo.2022.964822)
- 40↑
Lv F, Cai X, Li Y, Zhang X, Zhou X, Han X, & Ji L. Sensitivity to thyroid hormone and risk of components of metabolic syndrome in a Chinese euthyroid population. Journal of Diabetes 2023 15 900–910. (https://doi.org/10.1111/1753-0407.13441)
- 41↑
Pappa T, & Refetoff S. Resistance to thyroid hormone beta: a focused review. Frontiers in Endocrinology (Lausanne) 2021 12 656551. (https://doi.org/10.3389/fendo.2021.656551)
- 42↑
Dumitrescu AM, Korwutthikulrangsri M, & Refetof S. Impai red sensitivity to thyroid hormone: defects of transport, metabolism, and action. Endotext 2000. (https://www.ncbi.nlm.nih.gov/books/NBK279066/)
- 43↑
Sinha RA, Singh BK, Zhou J, Wu Y, Farah BL, Ohba K, Lesmana R, Gooding J, Bay BH, & Yen PM. Thyroid hormone induction of mitochondrial activity is coupled to mitophagy via ROS-AMPK-ULK1 signaling. Autophagy 2015 11 1341–1357. (https://doi.org/10.1080/15548627.2015.1061849)
- 44↑
Stuijver DJ, van Zaane B, Romualdi E, Brandjes DP, Gerdes VE, & Squizzato A. The effect of hyperthyroidism on procoagulant, anticoagulant and fibrinolytic factors: a systematic review and meta-analysis. Thrombosis and Haemostasis 2012 108 1077–1088. (https://doi.org/10.1160/TH12-07-0496)
- 45↑
Chavda V, & Lu B. Reverse electron transport at mitochondrial complex I in ischemic stroke, aging, and age-related diseases. Antioxidants 2023 12. (https://doi.org/10.3390/antiox12040895)
