Is free triiodothyronine important in the development of insulin resistance in healthy people?

Accepted Manuscript Title: Is free triiodothyronine important in the development of insulin resistance in healthy people? Authors: Vicente A. Benites-Zapata, Diego Urrunaga-Pastor, Cristina Torres-Mallma, Christian Prado-Bravo, Mirella Guarnizo-Poma, Herbert L´azaro-Alc´antara PII: DOI: Reference:

S1871-4021(17)30105-4 http://dx.doi.org/doi:10.1016/j.dsx.2017.04.022 DSX 776

To appear in:

Diabetes & Metabolic Syndrome: Clinical Research & Reviews

Please cite this article as: Benites-Zapata Vicente A, Urrunaga-Pastor Diego, TorresMallma Cristina, Prado-Bravo Christian, Guarnizo-Poma Mirella, L´azaro-Alc´antara Herbert.Is free triiodothyronine important in the development of insulin resistance in healthy people?.Diabetes and Metabolic Syndrome: Clinical Research and Reviews http://dx.doi.org/10.1016/j.dsx.2017.04.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Is free triiodothyronine important in the development of insulin resistance in healthy people?

Authors: Vicente A. Benites-Zapata1, Diego Urrunaga-Pastor2, Cristina Torres-Mallma2, Christian PradoBravo2, Mirella Guarnizo-Poma3, Herbert Lázaro-Alcántara3

Affiliation: 1. Centre for Public Health Research, Research Institute, Faculty of Medicine, University of San Martin de Porres, Lima, Peru 2. Scientific Society of Medical Students at the University of San Martin de Porres, University of San Martin de Porres, Lima, Peru 3. Metabolic Medical Institute, Lima, Peru

Type of article: Original article

Funding: None

Conflict of interest: Authors declare having no conflicts of interest regarding this paper

Correspondence: Vicente A Benites-Zapata. Address: Alameda del Corregidor avenue #1531 (USMP Campus), La Molina, Lima, Peru. Tel: +51995523081 E-mail: [email protected]

I. Abstract Objective: To determine the association between thyroid hormones and insulin resistance in a population of healthy individuals. Materials and methods: We conducted an analytical cross-sectional study in adults who attended a private clinic from 2012 to 2014. We excluded those participants with fasting glucose values compatible with diabetes mellitus, abnormal thyroid hormone values, chronic use of corticosteroids, and incomplete medical records. Participants were divided into tertiles (low, intermediate and high) according to their free triiodothyronine and free thyroxine values. We defined Insulin resistance as a Homeostasis Model Assessment (HOMA-IR) value greater than 3.8. We conducted a univariate and multivariate Poisson regression model to assess the association between thyroid hormones and insulin resistance. The association measure reported was the prevalence ratio (PR) with their confidence interval (CI) at 95%. Results: We evaluated 600 participants. The mean age was 36.8 ± 14.2 years and 33% were male. The frequency of insulin resistance was 29.5%. In the univariate regression, we found association between free triiodothyronine tertiles and insulin resistance. In the multivariate regression adjusted for age, sex, body mass index and thyroid stimulating hormone, the association between free triiodothyronine tertiles and insulin resistance remained; intermediate tertile (PR=1.54; CI95%: 1.10-2.15) and high tertile (PR=1.70; CI95%: 1.21-2.39). We found no association between T4 and insulin resistance. Conclusions: High levels of free triiodothyronine are associated with insulin resistance. The use of free trioodothyronine to assess insulin resistance in healthy patients should be considered. Key words: Thyroid Hormones; Triiodothyronine; Thyroxine; Insulin Resistance; Adult. (Source: MeSH NLM/DeCS BIREME).

II. Introduction Insulin resistance (IR) is frequently associated with type 2 diabetes (1). According to WHO, in 2014, the global prevalence of type 2 diabetes was 9% among adults older than 18 years (2). In addition, more than 80% of deaths caused by diabetes are registered in low- and middle-income countries, showing a problem for health systems worldwide (3). Thus, both individually and in the public health field, IR is a determining factor for various problems. IR is an alteration in insulin signalling, in which mainly the liver, muscle and adipose tissue do not adequately capture glucose, causing hyperglycemia and hyperinsulinemia in the long term. People with IR show certain common clinical features that might suggest their presence; including acanthosis nigricans, lipodystrophy, linear growth retardation, autoimmune reactions, and muscular cramps (4). However, IR may be present even years before the appearance of clinical presentation (5). The end result is the development of type 2 diabetes mellitus (6). IR is also associated with other disorders such as altered metabolism of thyroid hormones, cardiovascular diseases (7), polycystic ovary syndrome and neoplasms such as breast cancer (8), endometrial cancer (9) and colon cancer (10). It is known that thyroid function is associated with IR (11) and that there is a significant increase in insulin and glucose levels in patients with hypothyroidism (12). Likewise, serum levels of thyroid stimulating hormone (TSH) are significantly increased in patients with IR (13) whereas hyperinsulinemia is related to low serum levels of triiodothyronine (T3) and thyroxine (T4) (14). Previous studies have found a positive association between the presence of subclinical hypothyroidism and IR in individuals with no history of metabolic abnormalities (15). However, much is not known about this association in populations with no history of thyroid disorders (16, 17). Because IR is a pathological condition that can occur silently, its early detection and prevention is important in otherwise healthy appearing patients. For this reason, the objective of this study was to determine the association between thyroid hormones and IR in a population of individuals with no history of endocrine or metabolic diseases at a health centre in Lima-Peru. III. Methods Study Design Observational, retrospective study with an analytical cross-sectional design. Population and sample The study population consisted of adults of both sexes with no previous history of diabetes mellitus or other metabolic or endocrine disorders, who attended the outpatient service of a private clinic in Lima-Peru. The sample consisted of all patients who attended the outpatient service of the private clinic during the period from 2012 to 2014 and who met the eligibility criteria of the study.

Eligibility Criteria Participants included in the study were patients ≥ 18 years with no history of diabetes mellitus or other metabolic diseases (polycystic ovary syndrome, hypothyroidism, subclinical hypothyroidism, hyperthyroidism or metabolic syndrome). We excluded participants with fasting glucose values ≥ 126 mg/dl, thyroid hormones values outside the following ranges: free T3 (FT3): 2.3-4.2 pg/ml, free T4 (FT4): 0.89-1.76 ng/dl, TSH: 0.35-5.5 μU/ml (13), history of chronic use of corticosteroids, pregnant women and those individuals who did not clinically relevant data in their medical records. Exposure Groups Participants who met the eligibility criteria were divided into tertiles according to their FT3 and FT4 values. Thus, we generated three groups: low tertile, intermediate and high tertile for both FT3 and FT4 values separately. Main outcome Our primary outcome of interest was the IR defined by those participants with a Homeostasis Model Assessment (HOMA-IR) value greater or equal to 3.8 (15). Mathews et al. proposed HOMAIR in 1985 in a mathematical model that was used to assess hyperinsulinemia. HOMA-IR was derived from the interaction between β-cell function and insulin sensitivity, using the mathematical model, where fasting glucose and insulin concentrations values were used. The gold standard for the diagnosis of IR is the hyperinsulinemic euglycemic clamp technique, which consists of intravenous infusion of insulin to maintain a permanently elevated level of insulin above the corresponding fasting period. Simultaneously, glycemia determinations are performed every 2-5 minutes to infuse glucose at such a rate to maintain glycemia around 5 mmol/L stably. The required glucose infusion rate must be proportional to the insulin sensitivity, and therefore inversely proportional to the IR (18). HOMA-IR is correlated with the hyperinsulinemic euglycemic clamp and is a good measure of IR (19). HOMA-IR was calculated using the formula: fasting glucose (mg/dl) x fasting insulin (µU/ml)/405. Data collection procedures We reviewed the medical records of the patients treated during the study period. The values of fasting glucose and insulin levels were only collected if the patient laboratory tests were performed with a maximum of 30 days after they were attended the outpatient service. All participants had a minimum fasting period of eight hours for laboratory tests, according to the protocols established by the individual medical centers. We also collected other variables such as age, sex, body mass index (BMI), free triiodothyronine (FT3), free thyroxine (FT4) and thyroid stimulating hormone (TSH).

Analysis of data Numerical variables are presented as means with standard deviation, median with interquartile range, depending on their distributions. Categorical variables are presented as numbers with percentages. We used the analysis of variance (ANOVA) and the nonparametric Krustal-Wallis test to make comparisons between numerical variables in normally distributed variables and nonsymmetrically distributed variables respectively. The Pearson correlation coefficient (r) was used to assess the relationship between numeric variables as HOMA-IR and FT3 or FT4. To assess the association between IR and FT3 and FT4 tertiles we conducted simple and multivariate Poisson regression models. The measure of association reported is the prevalence ratio (PR) with their respective 95% confidence intervals (CI). The multivariate model was adjusted for age, sex, body mass index (BMI) and TSH. All the analysis was conducted using the statistical package STATA version 14. Ethical aspects The data was initially collected by an independent researcher to study epidemiological surveillance. For this specific study, the data of the participants was then transferred in an Excel file with no patient identifiers, thus maintaining the confidentiality of the participants’ information in the study. IV. Results General characteristics of the population Out of a total of 705 patients during the study period, 105 were excluded due to hyperthyroidism, hypothyroidism, subclinical hypothyroidism or diabetes mellitus, leaving 600 participants for the final analysis. The average age of the participants was 36.8 ± 14.2 years, 199 (33%) were male and the mean BMI was 27.8 kg/m2. The median HOMA-IR in the study population was 2.7 (1.64-4.30) while the mean values of FT3 and FT4 were 3.2 pg/ml ± 0.4 (SD) and 1.2 ng/dl ± 0.2 (SD) respectively. The exposure groups for FT3 tertiles were defined as follows: low (2.39-3.02 pg/ml), intermediate (3.03-3.38 pg/ml) and high (3.39-4.19 pg/ml). The FT4 tertiles were distributed from 0.9-1.11 ng/dl in the low tertile, 1.12-1.24 ng/dl in the intermediate tertile and 1.25-1.76 ng/dl for high tertile. For FT3 tertiles, we found that the participants in the high tertile were younger than those in the other tertiles (p<0.01). Also, a lower number of men were found in the low and intermediate tertile compared to the high tertile and this difference was statistically significant (p<0.01). We found a higher median BMI in the high tertile group compared with those in the low tertile group (p=0.01). TSH levels were found evenly distributed in the tertiles (p=0.14). Table 1 describes the characteristics of the participants according to their FT3 tertiles. As for the FT4 tertiles, we found that the participants in the high tertile were younger than those in the other tertiles (p<0.01). We found a lower number of men in the low and intermediate tertile

compared with the high tertile (p<0.01). In addition, we found a lower median BMI in the high tertile compared to the low tertile (p<0.01). However, TSH levels in the low tertile were greater than those found in intermediate and high tertile (p<0.01). Table 2 describes the characteristics of individuals according to their FT4 tertiles. The correlation between FT3 values and HOMA-IR was positive and weak (r=0.19, p<0.01), whereas the correlation between FT4 values and HOMA-IR was negative and weak (r=-0.10, p=0.01). The correlations between HOMA-IR and thyroid hormones (FT3 and FT4) are presented in Figures 1 and 2 respectively. The frequency of IR in the study participants was 29.5% (n=177). We found an increase in the frequency of IR in FT3 tertiles, from low to the high tertile. The prevalence of IR was 17.3% (n=35), 30.7% (n=62) and 40.8% (n=80) for the low, intermediate and high tertile, respectively and they were significantly different from each other (p<0.01). The frequency of IR was higher in the low FT4 tertile with 34.2% (n=69), and the frequency of IR in the intermediate and high tertile was similar with 26.9% (n=54) and 27.4% (n=54), respectively with no significant differences between IR and FT4 tertiles (p=0.20). In the crude analysis between FT3 tertiles and IR, using the low tertile as reference, we found an association with the IR in the intermediate tertile (PR=1.77; CI95%: 1.23-2.55) and the high tertile (PR=2.36; CI95%: 1.67-3.33), respectively. In addition, crude analysis showed association between age, male gender and BMI with IR (Table 3). In the multivariate analysis between FT3 tertiles and IR, after adjusting for age, sex, BMI and TSH, the association between FT3 tertiles and IR remained statistically significant. Using low tertile as reference, we found that the prevalence of IR increased by 54% in those participants located in the intermediate tertile (PR = 1.54; CI95%: 1.10 - 2.15), whereas the prevalence of IR increased by 70% in the individuals located in the high tertile (PR = 1.70; CI95%: 1.21 - 2.39). Besides, only BMI remained associated with the IR after multivariate analysis performed (Table 3). In the crude analysis between FT4 terciles and IR, using the low tertile as reference, we did not find an association in the middle tertile, (PR=0.79; CI95%: 0.58-2.55) and the high tertile, (PR=0.80; CI95%: 0.60-1.08), respectively. Furthermore, crude analysis showed association between age, male gender and BMI with IR (Table 4). In multivariate analysis after adjusting for age, sex, BMI and TSH, the association between FT4 tertiles and IR remained, but was not statistically significant. In addition, male gender and BMI remained associated with IR after adjusting for variables (Table 4).

V. Discussion In our study conducted on people with no history of metabolic or endocrine disorders, we report an association between FT3 tertiles and IR, measured with the HOMA-IR model. We did not find an association between FT4 tertiles and IR. In addition, we found that BMI was associated with the presence of IR in the studied population. The association between FT3 tertiles and IR has been reported in a few previous studies. In a study conducted in Turkey involving 211 patients (187 women and 24 men) with a mean age of 39.7 ± 11.7 years and no previous hormonal pathologies and a BMI ≥30; the authors reported that elevated levels of FT3 and FT4 correlated with IR (20). These findings are similar to ours regarding an association between FT3 and IR but differ with our findings related to FT4 and IR. This could be because we excluded obese participants and patients with undiagnosed type 2 diabetes or thyroid disorders from our study. It is therefore possible that in "healthy" patients FT3 has a better prognostic value than FT4 to assess IR. The relationship between FT3 and IR can be explained due to the effects that T3 has on energy expenditure and thermogenesis in individuals who have developed IR (20). The authors hypothesized that the increase in carbohydrates intake would generate an increase in metabolism that would manifest as elevated levels of FT3. This would accelerate lipid metabolism avoiding the excessive increase of fat and BMI in these patients. In many of these individuals, IR could be caused by a nutritional disorder and a chronic inflammatory process, thus, increased FT3 levels could mean a tissue response to the increase in the energy supply independent of the TSH level. Similarly, in a study by Ozcelik et al. on 315 obese men, it was reported that T3 values may be affected by IR (21). The authors suggested that this relationship could be due to insulin and T3`s similar roles in terms of regulation of glucose homeostasis. They therefore hypothesized that T3 levels increase as a compensatory mechanism to maintain homeostasis of glucose level and lipid metabolism up to a certain limit, exhausting its secretion, leading to subclinical hypothyroidism which is associated with IR (22). In a cross-sectional study involving 40 patients aged 18 to 45 years, diagnosed with subclinical hypothyroidism; the authors reported a significant increase in insulin, HOMA-IR and glucose levels (12). Similarly, another study reported that the levels of T3 had a strong correlation with insulin levels; a moderate correlation with the values of HOMA-IR and a weak correlation between T4 and insulin and HOMA-IR values (13). However, it is worth mentioning that these findings were in patients with subclinical hypothyroidism. We found no significant association between FT4 tertiles and IR measured with HOMA-IR model. Previous studies have reported similar findings. For example, in a study conducted in healthy euthyroid men in Iran, the main study finding was a negative association between the levels of FT4 and IR measured with the HOMA-IR model (23). In our study, we reported a negative correlation between the values of FT4 and HOMA-IR, however when the association was assessed by tertiles we did not find differences in relation to IR. The lack of association in our study between FT4 tertiles and IR may be explained because T4 is produced only by the thyroid gland while T3 is synthesized by deiodination of T4 in many other tissues aside from thyroid gland. In another study

from South West Asia, it was shown that drug therapy with thyroxine does not cause a meaningful change in insulin sensitivity (24). We found a direct association between BMI and IR measured by the HOMA-IR model. Previous studies have also reported a direct association between BMI and IR (25). BMI has been suggested to be a very good predictor of insulin resistance (25). The Minneapolis Children Blood Pressure Study has reported the same association in children (26). This occurs because the increase of adipose tissue is associated with an increased synthesis of pro inflammatory cytokines, responsible for the development of IR. Lastly, in patients with obesity, the capacity of adipose tissue to store lipids is altered leading to an overflow of lipids to other tissues where they could interfere with insulin signaling (27). According to our findings and to the evidence described in previous studies about the association between FT3 levels and IR, it would be prudent to evaluate FT3 values in laboratory tests of patients with potential risk factors for developing IR and diabetes. This may be beneficial at the preventive level of clinical practice allowing the clinicians to consider the possibility of the development of IR in people with normal FT3 values. Lastly, eating healthy and performing physical activity can indirectly prevent the appearance of diabetes mellitus by affecting this pathway. Our study has limitations. We used HOMA-IR to measure IR as the gold standard is the euglycemichyperinsulinemic clamp. However previous studies have shown a very high correlation between the two values. Another limitation was the study design; a cross-sectional study cannot establish causality and only allows to establishing associations. In addition, obtaining data for the study based on the review of medical records may have registry errors; however, we carried out a strict quality control of the data to overcome these difficulties. Finally, this study was conducted in a single medical center and the results may not be representative of a general population. Nevertheless, the considerable number of participants included, permitted us to perform an adjusted regression model by potential confounders. Our study demonstrated an association between FT3 and IR in a healthy adult population with no evidence of endocrine or metabolic diseases. We did not find an association between FT4 levels and IR in the same population. Prospective longitudinal studies in a cohort of healthy adults are needed to assess the role of thyroid hormones in the pathophysiology of IR, diabetes mellitus.

VII. References 1. Ríos MS, Diabetes GdTRalidlSEd. Resistencia a la insulina y su implicación en múltiples factores de riesgo asociados a diabetes tipo 2. Medicina Clínica. 2002;119(12):458-63. 2. Organization WH. Global status report on noncommunicable diseases 2014: World Health Organization; 2014. 3. Alwan A, Armstrong T, Bettcher D, Branca F, Chisholm D, Ezzati M. Informe sobre la situación mundial de las enfermedades no transmisibles. Organización Mundial de la Salud. 2010. 4. Mantzoros C. Insulin resistance: Definition and clinical spectrum. Up to Date Online. 2005;14:1-5. 5. DeFronzo RA. Pathogenesis of type 2 diabetes mellitus. Medical Clinics of North America. 2004;88(4):787-835. 6. Muoio DM, Newgard CB. Molecular and metabolic mechanisms of insulin resistance and βcell failure in type 2 diabetes. Nature reviews Molecular cell biology. 2008;9(3):193-205. 7. Eckel RH, Grundy SM, Zimmet PZ. The metabolic syndrome. The lancet. 2005;365(9468):1415-28. 8. Hernandez AV, Guarnizo M, Miranda Y, Pasupuleti V, Deshpande A, Paico S, et al. Association between insulin resistance and breast carcinoma: a systematic review and metaanalysis. PloS one. 2014;9(6):e99317. 9. Hernandez AV, Pasupuleti V, Benites-Zapata VA, Thota P, Deshpande A, Perez-Lopez FR. Insulin resistance and endometrial cancer risk: A systematic review and meta-analysis. European journal of cancer. 2015;51(18):2747-58. 10. Bowers K, Albanes D, Limburg P, Pietinen P, Taylor PR, Virtamo J, et al. A prospective study of anthropometric and clinical measurements associated with insulin resistance syndrome and colorectal cancer in male smokers. American journal of epidemiology. 2006;164(7):652-64. 11. de Jesús Garduño-Garcia J, Romero EC, Ochoa AL, Romero-Figueroa S, Bravo GH, García RT, et al. Thyroid function is associated with insulin resistance markers in healthy adolescents with risk factors to develop diabetes. Diabetology & metabolic syndrome. 2015;7(1):16. 12. Upadya U, Suma M, Srinath K, Prashant A, Parveen Doddamani SS. Effect of insulin resistance in assessing the clinical outcome of clinical and subclinical hypothyroid patients. Journal of clinical and diagnostic research: JCDR. 2015;9(2):OC01. 13. Vyakaranam S, Vanaparthy S, Nori S, Palarapu S, Bhongir AV. Study of Insulin Resistance in Subclinical Hypothyroidism. International journal of health sciences and research. 2014;4(9):147. 14. Farasat T, Cheema AM, Khan MN. Hyperinsulinemia and insulin resistance is associated with low T 3/T 4 ratio in pre diabetic euthyroid pakistani subjects. Journal of Diabetes and its Complications. 2012;26(6):522-5. 15. Ascaso JF, Real JT, Priego A, Carmena R, Romero P, Valdecabres C. Cuantificación de insulinorresistencia con los valores de insulina basal e índice HOMA en una población no diabética. Medicina clínica. 2001;117(14):530-3. 16. Farasat T, Cheema AM, Khan MN. Relationship of thyroid hormones with serum fasting insulin and insulin resistance in euthyroid glycemic anomalies. Pakistan J Zool. 2011;43:379-86. 17. Ortega E, Koska J, Pannacciulli N, Bunt JC, Krakoff J. Free triiodothyronine plasma concentrations are positively associated with insulin secretion in euthyroid individuals. European Journal of Endocrinology. 2008;158(2):217-21. 18. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. American Journal of Physiology-Gastrointestinal and Liver Physiology. 1979;237(3):G214-G23.

19. Bonora E, Targher G, Alberiche M, Bonadonna RC, Saggiani F, Zenere MB, et al. Homeostasis model assessment closely mirrors the glucose clamp technique in the assessment of insulin sensitivity: studies in subjects with various degrees of glucose tolerance and insulin sensitivity. Diabetes care. 2000;23(1):57-63. 20. Tarcin O, Abanonu GB, Yazici D, Tarcin O. Association of metabolic syndrome parameters with TT3 and FT3/FT4 ratio in obese Turkish population. Metabolic syndrome and related disorders. 2012;10(2):137-42. 21. Ozcelik F, Yuksel C, Arslan E, Genc S, Omer B, Serdar MA. Relationship between visceral adipose tissue and adiponectin, inflammatory markers and thyroid hormones in obese males with hepatosteatosis and insulin resistance. Archives of medical research. 2013;44(4):273-80. 22. Maratou E, Hadjidakis DJ, Kollias A, Tsegka K, Peppa M, Alevizaki M, et al. Studies of insulin resistance in patients with clinical and subclinical hypothyroidism. European journal of Endocrinology. 2009;160(5):785-90. 23. Amouzegar A, Kazemian E, Gharibzadeh S, Mehran L, Tohidi M, Azizi F. Association between thyroid hormones, thyroid antibodies and insulin resistance in euthyroid individuals: A population-based cohort. Diabetes & metabolism. 2015;41(6):480-8. 24. Nada AM. Effect of treatment of overt hypothyroidism on insulin resistance. World J Diabetes. 2013;4(4):157-61. 25. Molist-Brunet N, Jimeno-Mollet J, Franch-Nadal J. Correlación entre las diferentes medidas de obesidad y el grado de resistencia a la insulina. Atención primaria. 2006;37(1):30-6. 26. Prineas RJ, Gillum RF, Horibe H, Hannan PJ, Stat M. The Minneapolis children's blood pressure study. Hypertension. 1980;2(4):I-18–I-23. 27. Pérez MR, Medina-Gómez G. Obesidad, adipogénesis y resistencia a la insulina. Endocrinología y nutrición. 2011;58(7):360-9.

1

2

3

Figure 1. Scatter plot for the correlation between FT3 and the log of the HOMA-IR value

-1

0

r=0.19; p<0.01

2.5

3 patients

3.5 Free T3

4 Fitted values

4.5

Supplementary Figure

3

Figure 2. Scatter plot for the correlation between FT4 and the log of the HOMA-IR value

-1

0

1

2

r= -0.10; p=0.01

.8

1

1.2

1.4 Free T4

patients

Fitted values

1.6

1.8

Table 1. Characteristics of the study population by FT3 tertiles FT3 tertiles (pg/ml) Low

Intermediate

High

p

(n=202)

(n=202)

(n= 196)

Value

Age (Years)

40.08 ± 13.6

37.62 ± 14.1

32.83 ± 14

<0.01

Male

39 (19.3)

62 (30.7)

98 (50.0)

<0.01

BMI (Kg/m2)

26.94 (23.3 a 29.8)

27.72 (23.4 a 31.5)

28.72 (24.1 a 32)

0.01

TSH (μU/ml)

2.53 ± 1.2

2.64 ± 1

2.42 ± 1.2

0.15

Variables

Abbreviation: FT3, free triiodothyronine; TSH, thyroid-stimulating hormone; BMI, body mass index Data expressed as mean ± standard deviation, median (interquartile range) or number (percentage)

Table 2. Characteristics of the study population by FT4 tertiles

FT4 tertiles (ng/dl) Low

Intermediate

High

P

(n=202)

(n=201)

(n= 197)

Value

Age (Years)

39.10 ± 14.3

37.47 ± 14.2

34.00 ± 13.8

<0.01

Male

36 (17.8)

63 (31.3)

100 (50.8)

<0.01

BMI (Kg/m2)

28.55 (24 a 32.4)

27.66 (23.5 a 31.1)

27.11 (23.3 a 30.1)

0.03

TSH (μU/ml)

2.76 ± 1.2

2.44 ± 1

2.39 ± 1.14

<0.01

Variables

Abbreviation: FT4, free thyroxine; TSH, thyroid-stimulating hormone; BMI, body mass index Data expressed as mean ± standard deviation, median (interquartile range) or number (percentage)

Table 3. Crude and adjusted Poisson Regression model to assess the association between IR and FT3 tertiles Variables

PR Crude (CI 95%)

P value

PR adjusted (CI 95%)

P value

Low

Reference

-

-

-

Intermediate

1.77 (1.23 – 2.55)

<0.01

1.54 (1.10 – 2.15)

0.01

High

2.36 (1.67 – 3.33)

<0.01

1.70 (1.21 – 2.39)

<0.01

Age (Years)

1.01 (1.00 - 1.02)

0.01

1.00 (0.99 – 1.01)

0.66

Male

2.04 (1.60 - 2.59)

<0.01

1.21 (0.95 – 1.55)

0.12

BMI (Kg/m2)

1.12 (1.09 - 1.14)

<0.01

1.11 (1.08 – 1.13)

<0.01

TSH (μU/ml)

1.10 (0.99 - 1.22)

0.08

1.02 (0.92 – 1.14)

0.67

FT3 Tertiles (pg/ml)

Abbreviation: IR, insulin resistance; FT3, free triiodothyronine TSH, thyroid-stimulating hormone; BMI, body mass index

Table 4. Crude and adjusted Poisson Regression model to assess the association between IR and FT4 tertiles Variables

PR Crude (CI 95%)

P value

PR adjusted (CI 95%)

P value

Low

Reference

--

-

-

Intermediate

0.79 (0.58 – 1.06)

0.11

0.86 (0.65 – 1.14)

0.30

High

0.80 (0.60 – 1.08)

0.15

0.91 (0.68 – 1.22)

0.54

Age (Years)

1.01 (1.00 - 1.02)

0.01

1.01 (0.99 – 1.01)

0.82

Male

2.04 (1.60 - 2.59)

<0.01

1.35 (1.05 – 1.73)

0.02

BMI (Kg/m2)

1.12 (1.09 - 1.14)

<0.01

1.11 (1.08 – 1.14)

<0.01

TSH (μU/ml)

1.10 (0.99 - 1.22)

0.08

1.01 (0.91 – 1.13)

0.79

FT4 Tertiles (ng/dl)

Abbreviation: IR, insulin resistance; FT4, free thyroxine; TSH, thyroid-stimulating hormone; BMI, body mass index