Free triiodothyronine/free thyroxine (FT3/FT4) ratio is strongly associated with insulin resistance in euthyroid and hypothyroid adults: a cross-sectional study

Abstract

Introduction:
Insulin resistance (IR) is a key and early pathogenetic mechanism of cardiometabolic diseases with huge potential if being early detected and mitigated, for lowering the burden of the diseases. Available data are conflicting to what extent adult thyroid dysfunction is associated with IR. Therefore, we aimed to investigate the association and to identify which thyroid parameters are predictors of IR.

Material and methods:
After undergoing basic anthropometric and biochemical studies including thyroid hormones, oral glucose tolerance test (OGTT) and insulin, 1425 middle-aged individuals were divided into three groups according to the thyroid parameters: overt hypothyroidism (OH), subclinical hypothyroidism (SH) and euthyroidism (EU).

Results:
The homeostasis model assessment of IR (HOMA-IR), fasting insulin and 2-hour glucose levels of OGTT showed a steady, yet insignificant, increase from EU through SH to OH. The strongest noted correlations were those of insulin levels with free triiodothyronine/free thyroxine (FT3/FT4) ratio (r = 0.206, p < 0.001) and FT3 (r = 0.205, p < 0.001). Also in the case of HOMA-IR, the only statistically significant correlations were observed with FT3 (r = 0.181, p < 0.001) and the FT3/FT4 ratio (r = 0.165, p < 0.001). Among other thyroid hormones, linear logistic regression proved the FT3/FT4 ratio as the only significant predictor of HOMA-IR (linear coefficient = 5.26, p = 0.027) and insulin levels (linear coefficient = 18.01, p = 0.023), respectively. Thyroid-stimulating hormone was not associated with IR in either correlation or regression analysis.

Conclusions:
The FT3/FT4 ratio should be more emphasised in the diagnosis and treatment of thyroid disorders. Patients could benefit from a pharmacological reduction of the FT3/FT4 ratio, potentially leading to a decrease in insulin resistance, and thus a risk of the cardiometabolic diseases.

Full text

8
Original paper
ORIGINAL PAPER
Free triiodothyronine/free thyroxine (FT3/FT4) ratio
isstrongly associated with insulin resistance in euthyroid
andhypothyroid adults: a cross-sectional study
Ladislav Štěpánek 1, 2, Dagmar Horáková 1, Lubomír Štěpánek3, Vladimír Janout 1,
Jana Janoutová 4, Kateřina Bouchalová 5, Karel Martiník6
1Department of Public Health, Faculty of Medicine and Dentistry, Palacký University Olomouc, Olomouc, Czech Republic
2Department of Occupational Medicine, University Hospital Olomouc and Faculty of Medicine and Dentistry,
Palacký University Olomouc, Olomouc, Czech Republic
3Institute of Biophysics and Informatics, First Faculty of Medicine, Charles University, Prague, Czech Republic
4Department of Healthcare Management, Faculty of Health Sciences, Palacký University Olomouc, Olomouc, Czech Republic
5Department of Pediatrics, University Hospital Olomouc and Faculty of Medicine and Dentistry, Palacký University Olomouc,
Olomouc, Czech Republic
6Ambulance Prof. MUDr. Karla Martiníka DrSc. s.r.o, Hradec Králové, Czech Republic
Abstract
Introduction: Insulin resistance (IR) is a key and early pathogenetic mechanism of cardiometabolic diseases with huge potential if de-
tected early and mitigated, for lowering the burden of the disease. Available data are conicting to what extent adult thyroid dysfunction
is associated with IR. Therefore, we aimed to investigate the association and to identify which thyroid parameters are predictors of IR.
Material and methods: After undergoing basic anthropometric and biochemical studies including thyroid hormones, oral glucose tol-
erance test (OGTT), and insulin, 1425 middle-aged individuals were divided into three groups according to thyroid parameters: overt
hypothyroidism (OH), subclinical hypothyroidism (SH), and euthyroidism (EU).
Results: The homeostasis model assessment of IR (HOMA-IR), fasting insulin, and two-hour glucose levels of OGTT showed a steady, yet
insignicant, increase from EU through SH to OH. The strongest noted correlations were those of insulin levels with free triiodothyronine/
/free thyroxine (FT3/FT4) ratio (r = 0.206, p < 0.001) and FT3 (r = 0.205, p < 0.001). Also in the case of HOMA-IR, the only statistically
signicant correlations were observed for FT3 (r = 0.181, p < 0.001) and the FT3/FT4 ratio (r = 0.165, p < 0.001). Among other thyroid
hormones, linear logistic regression proved the FT3/FT4 ratio as the only signicant predictor of HOMA-IR (linear coefcient = 5.26,
p = 0.027) and insulin levels (linear coefcient = 18.01, p = 0.023), respectively. Thyroid-stimulating hormone was not associated with IR
in either correlation or regression analysis.
Conclusions: The FT3/FT4 ratio should be more emphasised in the diagnosis and treatment of thyroid disorders. Patients could benet
from a pharmacological reduction of the FT3/FT4 ratio, potentially leading to a decrease in insulin resistance, and thus a corresponding
decrease in the risk of the cardiometabolic diseases. (Endokrynol Pol 2021; 72 (1): 8–13)
Key words: thyroid disorder; insulin resistance; FT3/FT4 ratio; hypothyroidism; euthyroidism
Endokrynologia Polska
DOI: 10.5603/EP.a2020.0066
Volume/Tom 72; Number/Numer 1/2021
ISSN 0423–104X, e-ISSN 2299–8306
Introduction
Thyroid disorders (TDs) are common conditions with
potentially devastating health consequences that affect
all populations worldwide. Both hypothyroidism and
hyperthyroidism have their overt, clinically manifested
forms caused by serum abnormities of the thyroid gland
hormones (free thyroxine [FT4] and free triiodothyro-
nine [FT3]). Whereas subclinical forms of TDs show
normal serum levels of free thyroid hormones, there
are inverse feedback changes in thyroid-stimulating
hormone (TSH) levels [1–3].
The prevalence of overt hypothyroidism (OH) in
the general adult population varies between 0–2%
and 5–3% in Europe and between 0–3% and 3–7% in
the USA, depending on the denition used [1]. The
diagnosis of subclinical hypothyroidism (SH) is bio-
chemical, solely based on thyroid function testing. In
iodine-sufcient populations, SH affects up to 10% of
the population [2]. The prevalence of hyperthyroidism
is 0.8% in Europe and 1.3% in the USA [3].
Proper thyroid function is important for the pre-
vention of cardiometabolic diseases. The importance
of thyroid hormones in maintaining cardiovascular
homeostasis can be deduced from clinical and ex-
perimental data showing that even subtle changes in
thyroid hormone concentrations adversely inuence
the cardiovascular system. Some potential mechanisms
Ladislav Štěpánek; Department of Public Health, Faculty of Medicine and Dentistry, Palacký University Olomouc,
Hněvotínská 3, 775 15 Olomouc, Czech Republic, tel: (+42) 608757316; e-mail: ladislav.stepanek@upol.cz

This article is available in open access under Creative Common Attribution-Non-Commercial-No Derivatives 4.0 International (CC BY-NC-ND 4.0) license, allowing to download articles
and share them with others as long as they credit the authors and the publisher, but without permission to change them in any way or use them commercially

9
Endokrynologia Polska 2021; 72 (1)
ORIGINAL PAPER
insulin. None of them had a history of thyroid disorders, and they
were not treated with any drugs known to affect thyroid function.
The study was conducted according to the principles stated in the
Declaration of Helsinki. The study was approved by the Ethics Com-
mittee of University Hospital Olomouc and Faculty of Medicine and
Dentistry, Palacký University Olomouc (Approval No. 20/11). To be
included in the study, all subjects signed informed consent forms.
Laboratory analysis
Venous blood samples were drawn in the morning after a 12-hour
fast. After centrifugation, the serum was used for analyses on the
day of blood collection. Glucose was analysed on a Cobas 8000
(Roche Diagnostics GmbH, Manheim, Germany) using the hexo-
kinase method. Also on the Cobas 8000, TSH, FT4, and FT3 were
measured by electrochemiluminescence immunoassay. Insulin
was determined by chemiluminescent microparticle immunoassay
on Architect i1000SR (Abbott Laboratories, Chicago, IL, USA). All
analyses were performed according to the manufacturer’s instruc-
tions and after verication of methods.
Statistical analysis
Statistical analysis was performed using the computing environ-
ment R (R Foundation for Statistical Computing, Vienna, Austria;
http://www.r-project.org/). All numerical variables were charac-
terised with descriptive statistics. The Wilcoxon signed-rank test
(p-value) was used to compare numerical characteristics between
the given subgroups. Extreme values of variables were found and
excluded using the inner and outer fences method. All variables fol-
lowed normal or near-normal distribution. Correlations of selected
variables were quantied with Pearson’s correlation coefcient (r),
and the level of signicance (p) was determined. Multivariate linear
regression was performed to examine the relationships between
thyroid hormones, age, and sex, as regressors of the dependent
variables insulin and HOMA-IR. Due to a hyperbolic dependence
of IR through the whole spectrum of TDs, evidenced by studies [5],
the correlation and linear regression analyses were performed only
in hypothyroid (both overt and subclinical) and euthyroid subjects.
In other words, IR grows in both hypo- and hyper-thyroidism (in
comparison with normal IR state in euthyroidism), whereas thyroid
parameters change linearly through the whole spectrum. Hyper-
thyroidism was chosen for exclusion because of the expected lower
prevalence in the study population. P-value of less than 0.05 indicates
statistical signicance. The following formula was used to calculates
HOMA-IR (glucose levels in mmol/L, insulin levels in mIU/L) [10]:
HOMA-IR = (glucose × insulin)/22.5
Results
Characteristics of the study population
The basic characteristics of all subjects with respect
to a particular TD are shown in Table 2. Only 23 in-
dividuals from the whole group met the criteria for
hyperthyroidism; therefore, due to the disproportion
to the size of the other subgroups, we did not include
them into the table. As expected, most individuals had
normal thyroid parameters, followed by the OH and
SH subgroups, respectively. Between the subgroups,
differences of the tested parameters, apart from the
thyroid parameters, did not reach statistical signi-
cance. There were no apparent trends of the parameters
between subgroups except for insulin, HOMA-IR, and
linking the two conditions are dyslipidaemia, endo-
thelial dysfunction, blood pressure or body weight
changes, and direct effects of thyroid hormones on the
myocardium [4]. Moreover, thyroid hormones have
an effect on glucose metabolism and probably on the
development of insulin resistance (IR). In hyperthy-
roidism, impaired glucose tolerance may be the result
of mainly hepatic IR, whereas in hypothyroidism the
available data suggest that IR of peripheral tissues
prevails [5]. However, the precise relationship between
TDs, especially subclinical forms, and IR is still under
discussion. IR represents a key pathogenetic mechanism
of cardiometabolic diseases, and its prevention is crucial
in the current dominance of cardiometabolic diseases
in overall mortality [6, 7].
In addition to serum thyroid hormones, some stud-
ies have discussed the relationship between T3/T4 ratio
and various, mostly metabolic, parameters or clinical
outcomes [8, 9]. The T3/T4 ratio is a calculated index used
to indicate thyroid function and the action of hormones
on the tissue. A method for detecting IR that is easy to
use in common clinical practice is the homeostasis model
assessment of IR (HOMA-IR). Its satisfactory correla-
tion with the most accurate glucose clamp techniques
has been conrmed by numerous studies. Even fasting
insulin levels alone reect IR to some extent [10].
The study aimed to investigate the association of
IR with dened TDs and thyroid function parameters,
and through these ndings, identify such thyroid dis-
orders and parameters that are the highest risk for the
development of IR.
Material and methods
Study population
The study comprised 1425 middle-aged individuals (317 males, 1108
females) examined in a co-author’s internal medicine outpatient
centre in Hradec Králové, Czech Republic. Data were collected in
2009–2016 from the patients’ rst visits to the centre. The labora-
tory analyses below, including standard 75 g oral glucose tolerance
test (OGTT), were performed, and, at the same time, their basic
anthropometric parameters (waist circumference [WC], weight, and
height) were measured. The obtained data were used to calculate
body mass index (BMI) and HOMA-IR for each participant. Based
on their TSH and FT4 levels and the manufacturer ’s instructions,
all subjects were divided into subgroups according to Table 1. None
of the included subjects was treated with oral antidiabetic drugs or
Table 1. Study subgroups according to thyroid hormone levels
Subgroups/thyroid status TSH [mIU/L] FT4 [pmol/L]
Euthyroidism 0.4–4.0 and 9–23
Subclinical hypothyroidism > 4.0 and 9–23
Overt hypothyroidism > 4.0 and < 9
Hyperthyroidism < 0.4 or > 23
TSH — thyroid-stimulating hormone; FT4 — free thyroxine

10
ORIGINAL PAPER
Thyroid disorders and insulin resistance Ladislav Štěpánek et al.
two-hour glucose levels, which showed steady growth
from euthyroidism through SH to OH.
Correlation analysis
The highest correlation coefcients of thyroid param-
eters were noted with insulin levels and HOMA-IR as
seen from Table 3, namely correlations of insulin with
the FT3/FT4 ratio (r = 0.206) and FT3 (r = 0.205), re-
spectively. Although the correlations were weak, they
reached high statistical signicance (p < 0.001). There
were also highly signicant correlations of both FT3
and the FT3/FT4 ratio with HOMA-IR (r = 0.181 and
r = 0.165, respectively). Figure 1 depicting the correla-
tions of the FT3/FT4 ratio suggests no difference be-
Table 2. Basic characteristics in study subgroups — mean value (95% condence interval of the mean value)
Characteristics Euthyroidism p value Subclinical
hypothyroidism p value Overt hypothyroidism
N (males, females) 1036
(243, 793) –76
(11, 65) –290
(57, 233)
Age [yrs] 42.21
(41.61; 42.81) 0.397 43.94
(43.21; 44.67) 0.452 43.23
(42.72; 43.75)
BMI [kg/m²] 31.17
(29.64; 32.79) 0.204 32.98
(25.66; 42.40) 0.172 32.04
(30.19; 34.00)
Waist circumference [cm] 109.69 (107.28;
112.10) 0.656 108.22
(102.01; 114.43) 0.899 110.78
(108.78; 112.78)
Glucose [mmol/L] 5.62
(5.58; 5.66) 0.223 5.73
(5.56; 5.90) 0.372 5.63
(5.56; 5.70)
Insulin [mIU/L] 6.70
(6.42; 6.98) 0.35 7.12
(5.93; 8.55) 0.483 7.85
(7.26; 8.48)
2-hour OGTT 6.26
(6.16; 6.35) 0.973 6.27
(5.87; 6.71) 0.097 6.60
(6.42; 6.80)
HOMA-IR 1.68
(1.60; 1.75) 0.224 1.82
(1.50; 2.21) 0.804 1.96
(1.80; 2.13)
TSH [mIU/L] 1.91
(1.86; 1.96) < 0.001 5.32
(5.04; 5.61) 0.091 6.44
(2.58; 4.31)
FT4 [pmol/L] 10.81
(10.73; 10.90) 0.373 10.68
(10.37; 10.99) < 0.001 8.28
(8.20; 8.36)
FT3 [pmol/L] 5.21
(5.17; 5.26) 0.711 5.18
(5.00; 5.37) 0.876 5.13
(5.05; 5.22)
FT3/FT4 ratio 0.48
(0.48; 0.49) 0.557 0.48
(0.46; 0.51) < 0.001 0.62 (0.61; 0.63)
BMI — body mass index; OGTT — oral glucose tolerance test; HOMA-IR — homeostasis model assessment of insulin resistance; TSH — thyroid-stimulating hormone;
FT4 — free thyroxine; FT3 — free triiodothyronine
Table 3. Pearson’s correlation coefcients (r) of thyroid function parameters
Characteristics TSH FT4 FT3 FT3/FT4
BMI 0.008 0.035 –0.055 –0.068*
Waist circumference 0.002 0.033 0.001 0.038
Glucose 0.014 0.064* 0.018 –0.029
Insulin –0.015 –0.081* 0.205** 0.206**
2-hour OGTT 0.007 –0.026 –0.007 0.044
HOMA-IR –0.015 –0.050 0.181** 0.165**
*p < 0.05, **p < 0.001; TSH — thyroid-stimulating hormone; FT4 — free thyroxine; FT3 — free triiodothyronine; BMI — body mass index; OGTT — oral glucose
tolerance test; HOMA-IR — homeostasis model assessment of insulin resistance

11
Endokrynologia Polska 2021; 72 (1)
ORIGINAL PAPER
tween males and females. As for the parameters not as-
sociated with thyroid functions, there were statistically
highly signicant correlations of WC and body weight
with both insulin and HOMA-IR, the strongest being
correlated body weight with HOMA-IR (r = 0.494).
Regression analysis
The impact of individual IR predictors is expressed with
multivariate linear regression models that use linear
coefcients to show the effects of one-point increases
of these predictors (independent variables) on the de-
pendent variables, HOMA-IR and insulin, respectively
(Tab. 4). Except for the sex and age, the FT3/FT4 ratio
was the only statistically signicant predictor of both
HOMA-IR and insulin. If the ratio increases by one
point, HOMA-IR and insulin will increase by a mean of
5.26 (p = 0.027) and 18.01 (p = 0.023), respectively. With
each year of age, HOMA-IR and insulin will increase by
a mean of 0.02 (p < 0.001) and 0.04 (p = 0.25), respec-
tively. Female sex was a signicant predictor of a lower
insulin level and HOMA-IR in our study population.
Discussion
The obtained results of the correlation analysis revealed
only a few statistically signicant and very weak asso-
ciations. The regression analysis conrmed a signicant
relationship only between IR and the FT3/FT4 ratio. The
Table 4. Effects of every 1-point increase of independent variables on the dependent variables HOMA-IR and insulin
Independent
variables
Linear coefcient Standard error t-value p-value
HOMA-IR Insulin HOMA-IR Insulin HOMA-IR Insulin HOMA-IR Insulin
Intercept –2.45 –5.82 1.33 4.44 –1.85 –1.31 0.065 0.19
Age 0.02 0.04 < 0.01 0.02 3.97 2.25 < 0.001 0.025
Sex (female) –0.8 –2.31 0.14 0.47 –5.75 –4.96 < 0.001 < 0.001
TSH –0.01 –0.05 0.02 0.05 –0.75 –0.96 0.455 0.338
FT4 0.16 0.44 0.12 0.39 1.39 1.12 0.166 0.261
FT3 0.04 0.23 0.25 0.84 0.17 0.27 0.862 0.785
FT3/FT4 ratio 5.26 18.01 2.37 7.92 2.22 2.27 0.027 0.023
HOMA-IR — homeostasis model assessment of insulin resistance; TSH — thyroid-stimulating hormone; FT4 — free thyroxine; FT3 — free triiodothyronine
Figure 1. Scatter diagrams showing correlations of free triiodothyronine/free thyroxine (FT3/FT4) ratio with fasting insulin levels (A)
and HOMA-IR (B) with 95% condence bands for the medians of variables on the y-axis for a given variable on the x-axis. There are
no obvious sex differences in the relationship of the ratio and insulin resistance
50
40
30
20
10
0
15
10
5
0
Insulin [mIU/L]
HOMA-IR
FT3/FT4 ratio FT3/FT4 ratio
0.4 0.6 0.8 1.0 1.2 0.4 0.6 0.8 1.0 1.2
Female Male

12
ORIGINAL PAPER
Thyroid disorders and insulin resistance Ladislav Štěpánek et al.
results also demonstrate both HOMA-IR and insulin
levels increasing in the way from euthyroidism through
SH to OH, although without statistical signicance.
Epidemiological evidence indicates that both OH and
SH are connected with IR. Some studies showed a sta-
tistically signicant increase of HOMA-IR and insulin
levels in subjects with SH compared to the euthyroid
[11, 12]. However, some ndings showed no or insigni-
cant difference between SH and the euthyroid state [13,
14]. A study by Wang et al. also recorded an increase in
HOMA-IR between SH and OH, as was the case in our
study, although in our case without reaching statistical
signicance [12]. Comparable IR between SH and OH
subjects was noted in a study by Maratou et al. [15]. We
also demonstrate an increase of two-hour OGTT values
in SH and OH compared to euthyroidism.
According to the population prevalence of TDs
indicated in the introduction, hypothyroidism pre-
dominates. SH was estimated to affect up to 10%, with
the highest prevalence among women and elderly
individuals. However, SH frequently reverts to euthy-
roidism, and TSH levels rise as people without TDs
age, making it likely that the prevalence of SH may be
overestimated [2]. The occurrence of TDs in our study
sample does not exactly correspond to the proposed
population prevalence. However, the study did not
comprise a random adult population, but rather pa-
tients of a stated outpatient centre.
The most common cause of hypothyroidism in
iodine-sufcient populations is Hashimoto’s thyroiditis
(HT). A chronic inammation resulting from an imbal-
ance between pro-inammatory and anti-inammatory
cytokines in HT might be responsible for IR in hypothy-
roidism. Subjects with highly positive thyroid peroxi-
dase antibodies (TPOabs) had elevated fasting insulin
levels compared to those with lower titers of TPOabs, in
a study by Mazarehi et al. [16]. OH was associated with
obesity, promoting the production of pro-inammatory
cytokines and contributing to the development of IR.
Moreover, BMI was positively correlated with TPOabs
in men, suggesting that the pathogenesis of IR in hy-
pothyroidism is a complex and interrelated process [17,
18]. In the present study, subjects with both SH and OH
showed higher BMI values compared to euthyroidism.
Elevated levels of the FT3/FT4 ratio were associated
with metabolic syndrome and some of its components
such as obesity, elevated triglycerides, fasting blood
glucose levels, and lower high-density lipoprotein
cholesterol levels in euthyroid subjects [19, 20]. Based
on regression analysis in a sample of 132,346 adults,
Park et al. conrmed the FT3/FT4 ratio to have a better
predictive power for metabolic syndrome than TSH
[21]. In hypothyroidism and euthyroidism, as TSH
levels rise the FT3/FT4 ratio also rises. Levothyroxine
therapy was proved to decrease the FT3/FT4 ratio [22,
23]. Given the association of both hypothyroidism and
the FT3/FT4 ratio with IR, thyroxine supplementation
in indicated cases may help not only in achieving eu-
thyroidism but also in improving IR.
The study has certain limitations. We considered all
hypothyroid subjects as having a peripheral form (not
distinguishing between central and peripheral forms)
of hypothyroidism. However, due to the low prevalence
of central hypothyroidism, there should be no major
bias for this reason.
Conclusion
Both SH and OH are associated with increased
HOMA-IR, fasting insulin levels, and two-hour OGTT
glucose levels compared to euthyroidism. From the
common laboratory parameters of the thyroid gland,
the FT3/FT4 ratio shows the strongest relationship to IR.
The FT3/FT4 ratio should be more emphasised in the di-
agnosis and treatment of TDs. After further verication,
pharmacological lowering of the ratio could be a pos-
sible targeted way to reverse the development of IR.
The results are valid for middle-aged European adults.
Conict of onterest
Nothing to declare.
Funding
The study was supported by the grant IGA_LF_UPOL
2020_031.
References
1. Chaker L, Bianco A, Jonklaas J, et al. Hypothyroidism. Lancet. 2017;
390(10101): 1550–1562, doi:10.1016/s0140-6736(17)30703-1, indexed in
Pubmed:28336049.
2. Biondi B, Cappola AR, Cooper DS. Subclinical Hypothyroidism: A Re-
view. JAMA. 2019; 322(2): 153–160, doi:10.1001/jama.2019.9052, indexed
in Pubmed:31287527.
3. Leo SDe, Lee S, Braverman L. Hyperthyroidism. Lancet. 2016;
388(10047): 906–918, doi:10.1016/s0140-6736(16)00278-6, indexed in
Pubmed:27038492.
4. Jabbar A, Pingitore A, Pearce SHS, et al. Thyroid hormones and cardio-
vascular disease. Nat Rev Cardiol. 2017; 14(1): 39–55, doi:10.1038/nrcar-
dio.2016.174, indexed in Pubmed:27811932.
5. Gierach M, Gierach J, Junik R. Insulin resistance and thyroid disor-
ders. Endokrynol Pol. 2014; 65: 70–76, doi:10.5603/EP.2014.0010, indexed
in Pubmed:24549605.
6. Garduño-Garcia Jd, Camarillo Romero E, Loe Ochoa A, et al. Thyroid
function is associated with insulin resistance markers in healthy adoles-
cents with risk factors to develop diabetes. Diabetol Metab Syndr. 2015;
7: 16, doi:10.1186/s13098-015-0011-x, indexed in Pubmed:25780389.
7. Samuel VT, Shulman GI. The pathogenesis of insulin resistance: integrat-
ing signaling pathways and substrate ux. J Clin Invest. 2016; 126(1):
12–22, doi:10.1172/JCI77812, indexed in Pubmed:26727229.
8. Emokpae MA, Obazelu PA. The Association of Triiodothyronine-to-Thy-
roxine Ratio with Body Mass Index in Obese Nigerian Children and
Adolescents. Med Sci (Basel). 2017; 5(4), doi:10.3390/medsci5040036,
indexed in Pubmed:29244733.
9. Pasqualetti G, Schirripa M, Dochy E, et al. Thyroid hormones ratio is
a major prognostic marker in advanced metastatic colorectal cancer:
Results from the phase III randomised CORRECT trial. Eur J Cancer.
2020; 133: 66–73, doi: 10.1016/j.ejca.2020.04.023, indexed in Pubmed:
32446145.

13
Endokrynologia Polska 2021; 72 (1)
ORIGINAL PAPER
10. Wallace TM, Levy JC, Matthews DR. Use and abuse of HOMA model-
ing. Diabetes Care. 2004; 27(6): 1487–1495, doi:10.2337/diacare.27.6.1487,
indexed in Pubmed:15161807.
11. Ebrahimpour A, Vaghari-Tabari M, Qujeq D, et al. Direct correlation
between serum homocysteine level and insulin resistance index in
patients with subclinical hypothyroidism: Does subclinical hypothyroid-
ism increase the risk of diabetes and cardio vascular disease together?
Diabetes Metab Syndr. 2018; 12(6): 863–867, doi:10.1016/j.dsx.2018.05.002,
indexed in Pubmed:29748035.
12. Wang G, Liu J, Yang N, et al. Levothyroxine treatment restored the
decreased circulating broblast growth factor 21 levels in patients
with hypothyroidism. Eur J Intern Med. 2016; 31: 94–98, doi:10.1016/j.
ejim.2016.03.028, indexed in Pubmed:27085392.
13. Maleki N, Kazerouni F, Hedayati M, et al. Assessment of cardiovas-
cular risk factors in patients with subclinical hypothyroidism. Acta
Cardiol. 2016; 71(6): 691–697, doi:10.2143/AC.71.6.3178188, indexed in
Pubmed:27920457.
14. Aksoy N, Yeler MT, Ayan NN, et al. Association between thyroid hor-
mone levels and insulin resistance and body mass index. Pak J Med
Sci. 2015; 31(6): 1417–1420, doi:10.12669/pjms.316.7560, indexed in
Pubmed:26870107.
15. Maratou E, Hadjidakis DJ, Kollias A, et al. Studies of insulin resistance
in patients with clinical and subclinical hypothyroidism. Eur J En-
docrinol. 2009; 160(5): 785–790, doi:10.1530/EJE-08-0797, indexed in
Pubmed:19141606.
16. Mazaheri T, Shari F, Kamali K. Insulin resistance in hypothyroid patients
under Levothyroxine therapy: a comparison between those with and
without thyroid autoimmunity. J Diabetes Metab Disord. 2014; 13(1): 103,
doi:10.1186/s40200-014-0103-4, indexed in Pubmed:25364704.
17. Sanyal D, Raychaudhuri M. Hypothyroidism and obesity: An
intriguing link. Indian J Endocrinol Metab. 2016; 20(4): 554–557,
doi:10.4103/2230-8210.183454, indexed in Pubmed:27366725.
18. Wang B, Song R, He W, et al. Sex Differences in the Associations of
Obesity With Hypothyroidism and Thyroid Autoimmunity Among Chi-
nese Adults. Front Physiol. 2018; 9: 1397, doi:10.3389/fphys.2018.01397,
indexed in Pubmed:30337885.
19. Xu R, Huang F, Zhang S, et al. Thyroid function, body mass index,
and metabolic risk markers in euthyroid adults: a cohort study. BMC
Endocr Disord. 2019; 19(1): 58, doi:10.1186/s12902-019-0383-2, indexed
in Pubmed:31174521.
20. Urrunaga-Pastor D, Guarnizo-Poma M, Moncada-Mapelli E, et al. Insulin
Resistance and Metabolic Syndrome Research Group. High free triiodo-
thyronine and free-triiodothyronine-to-free-thyroxine ratio levels are
associated with metabolic syndrome in a euthyroid population. Diabetes
Metab Syndr. 2018; 12(2): 155–161, doi:10.1016/j.dsx.2017.12.003, indexed
in Pubmed:29254889.
21. Park SoY, Park SeE, Jung SW, et al. Free triiodothyronine/free thyroxine
ratio rather than thyrotropin is more associated with metabolic param-
eters in healthy euthyroid adult subjects. Clin Endocrinol (Oxf). 2017;
87(1): 87–96, doi:10.1111/cen.13345, indexed in Pubmed:28374508.
22. Ettleson MD, Bianco AC. Individualized Therapy for Hypothyroidism:
Is T4 Enough for Everyone? J Clin Endocrinol Metab. 2020; 105(9),
doi:10.1210/clinem/dgaa430, indexed in Pubmed:32614450.
23. Strich D, Chay C, Karavani G, et al. Levothyroxine Therapy Achieves
Physiological FT3/FT4 Ratios at Higher than Normal TSH Lev-
els: A Novel Justification for T3 Supplementation? Horm Metab
Res. 2018; 50(11): 827–831, doi:10.1055/a-0751-0498, indexed in Pubmed:
30396211.