The Association Between Six Surrogate Insulin Resistance Indexes and Hypertension: A Population-Based Study

Abstract

Background:

The relationship between insulin resistance and hypertension is well established, but the association of different surrogate insulin resistance indexes with the presence of hypertension is still under debate. The aim of this study was to compare the strength of the association between the presence of hypertension and six indexes: triglyceride/HDL cholesterol ratio (TG/HDL-C), Triglyceride Glucose (TyG) Index, Visceral adiposity index (VAI), Lipid accumulation product (LAP), TyG-Body mass index (TyG-BMI), and TyG-Waist circumference (TyG-WC).

Methods:

Data from a cross-sectional epidemiological study enrolling a sample representative for the Romanian population aged 18–80 years, excluding those with diabetes or requiring treatment for hypertriglyceridemia, were used to calculate the six indexes. The association with the presence of hypertension was examined with binomial and multinomial logistic regression.

Results:

In multinomial logistic models, which included age, gender, smoking, drinking, sedentary lifestyle, estimated glomerular filtration rate, urinary sodium, urinary albumin creatinine ratio, and use of medications known to influence insulin resistance as covariates, all individual components and surrogate insulin resistance indexes were independently associated with the presence of hypertension. Values of pseudo R square ranged from 0.342 for the multivariate model including TG/HDL-C to 0.357 for the model including TyG-WC, but with no clear superiority of any of the tested indexes over all others. Models including BMI and WC had a similar ability to predict the presence of hypertension as most of the surrogate indexes and they were slightly superior to TG/HDL-C and TyG.

Conclusions:

Although TG/HDL-C, VAI, LAP, TyG, TyG-BMI, and TyG-WC were independently associated with the presence of hypertension, no superiority could be demonstrated over the use of BMI and WC as predictors of hypertension in this cross-sectional study.

Introduction

Hypertension is one of the most prevalent cardiovascular risk factors, with over 34% of males and 28% of women aged ≥25 years being affected globally by raised blood pressure.¹ Obesity plays an important role in the development of hypertension, with at least 75% of cases of incident hypertension directly related to excessive weight.² The main pathogenetic pathways linking obesity and hypertension are thought to be central obesity through insulin resistance/hyperinsulinemia and increased leptin, increased activity of the sympathetic nervous system and of the renin-angiotensin-aldosterone system, and increased renal sodium reabsorption.³ Insulin has complex actions on the vascular system as a result of the interplay between the activation of the phosphatidylinositol-3-kinase (PI3K) and mitogen-activated protein kinase (MAPK) pathways.⁴ It was shown that in insulin resistance states, nitric oxide (NO) production is decreased due to selective impairment of PI3K pathway, whereas MAPK pathway is activated normally and leads to vasoconstriction, and that these mechanisms may contribute to the development of hypertension.⁴ The association between insulin resistance and risk of incident hypertension was demonstrated in a recent meta-analysis of 11 studies, including over 55,000 participants, in which those in the highest category of homeostasis model assessment insulin resistance (HOMA-IR) had a pooled adjusted relative risk of hypertension of 1.43 (95% CI 1.27–1.62) compared to those in the lowest category,⁵ suggesting that insulin resistance could be used as an adjunctive tool to identify individuals at risk for hypertension.

In clinical practice, the use of HOMA-IR is restricted due to the relative high cost of measuring insulin. Therefore, some easy-to-determine surrogate indexes of insulin resistance have been proposed and validated against HOMA-IR: triglyceride/HDL cholesterol ratio (TG/HDL-C),⁶ Triglyceride Glucose (TyG) Index,⁷ Visceral adiposity index (VAI),⁸ and Lipid accumulation product (LAP).⁹ More recently, two indexes derived from TyG have also been proposed: TyG-Body mass index (TyG-BMI) and TyG-Waist circumference (TyG-WC).¹⁰ These indexes were studied as predictors in various clinical conditions and compared with adiposity indexes (e.g., BMI, WC) in studies reporting conflicting results. TG/HDL-C has been shown to predict mortality from coronary heart disease and cardiovascular disease (CVD) and risk of diabetes mellitus (DM) in men, whereas TyG only predicts risk of DM.¹¹ TG/HDL-C, but not VAI, is also a predictor of incident CVD.¹² VAI has been associated with an increased risk for type 2 DM, but not better than other fatness indices.^13,14 The ability of LAP to identify adverse levels of cardiovascular risk factors is higher than that of BMI for seven lipid parameters, including total, HDL and LDL cholesterol as well as for uric acid and heart rate, but not for systolic blood pressure (SBP) and diastolic blood pressure (DBP).⁹ LAP, but not BMI, has been shown to predict all-cause mortality in high-risk, nondiabetic individuals¹⁵ and performed better than BMI and WC in identifying individuals with metabolic syndrome.¹⁶ Other index proposed for early identification of metabolic syndrome was hypertensive waist, but this only applies in later stages when hypertension is already installed.¹⁷ Only few studies examined the association between surrogate insulin resistance indexes and the presence of hypertension,¹⁸ and none applied them concomitantly in the same population.

We hypothesized that different surrogate insulin resistance indexes may be superior to their individual components and may have different ability to discriminate between individuals with and without hypertension and thus being of value to identify individuals at risk for hypertension and to further stratify the overall cardiovascular risk. To test this hypothesis, we aimed to compare the strength of the association between presence of hypertension and six surrogate insulin resistance indexes (TG/HDL-C, TyG, TyG-BMI, TyG-WC, VAI, and LAP) and to examine whether they are superior to the individual components using data from an epidemiological study representative for the general population.

Methods

SEPHAR (Epidemiologic Study of the Prevalence of Arterial Hypertension and Cardiovascular Risk in Romania) III was a cross-sectional epidemiological study aiming to determine the prevalence of arterial hypertension and other cardiovascular risk factors in a representative sample for Romanian population 18–80 years of age. The detailed protocol of the study was published elsewhere.¹⁹ Study participants underwent two study visits 4 days apart in the period of November 16th, 2015–April 25th, 2016. Fasting blood and urine samples were collected at Visit 2 and all measurements were performed at a central laboratory.

The research was conducted in accordance with Good Clinical Practice Guidelines and the Declaration of Helsinki as revised in 2013, and the protocol was approved by the Ethics Committee from Carol Davila University of Medicine and Pharmacy, Bucharest, Romania. Participants were included if they provided written informed consent before any study-related procedure.

For the purpose of this analysis, only participants who completed both Visit 1 and 2 were selected. The following variables were retained: age, gender, urban/rural area of residence, smoking (current, ex- or nonsmokers), alcohol consumption (no/yes), sedentary lifestyle (no/yes), history of CVD (ischemic heart disease, cerebrovascular disease, peripheral artery disease), use of medications known to influence insulin resistance-angiotensin-converting enzyme inhibitors (ACEIs), angiotensin-receptor blockers (ARBs), thiazide and thiazide-like diuretics, beta-blockers, statins, ezetimibe or other lipid-lowering drugs, and antiobesity medication, SBP and DBP, BMI, WC, fasting blood glucose, glycated hemoglobin A1c (HbA1c), total, HDL and LDL cholesterol, serum triglycerides (TG), serum creatinine, and uric acid. Urinary albumin creatinine ratio and urinary sodium were determined from a morning urine sample. Estimated glomerular filtration rate (eGFR) was calculated with CKD-EPI creatinine equation.²⁰

Participants were considered as having diabetes if they had a previous diagnosis of diabetes and/or were using antihyperglycemic medications or if they had fasting blood glucose ≥ 126 mg/dl and HbA1c ≥ 6.5% at study visit, according to American Diabetes Association criteria.²¹ Diagnosis of hypertension was based on history of hypertension or use of antihypertensive drugs for at least 2 weeks before study entry, or a mean SBP ≥ 140 mmHg and/or mean DBP ≥ 90 mmHg at both study visits, according to European Society of Cardiology recommendations.²²

The following insulin resistance indexes were calculated according to published formula:

– TG/HDL-C as the ratio between serum triglycerides and HDL cholesterol⁶;

– TyG Index as Ln[fasting triglycerides (mg/dL) × fasting glucose (mg/dL)/2]⁷;

– VAI as (WC [cm]/39.68 + (1.88 × BMI) × (TG [mM/L]/1.03) × (1.31/HDL-C [mM/L]) for men and (WC [cm]/36.58 + (1.89 × BMI) × (TG [mM/L]/0.81) × (1.52/HDL-C [mM/L]) for women⁸;

– LAP as (WC [cm]–65) × TG [mM/L]) for men and (WC [cm]–58) × TG [mM/L]) for women⁹;

– TyG-BMI as TyG index × BMI¹⁰;

– TyG-WC as = TyG index × WC.¹⁰

Participant with diabetes and/or treated with fibrates were excluded from the analysis as TyG was validated only in nondiabetic populations⁷ and LAP excluded individuals with very high triglycerides levels.⁹

Statistical analysis

Statistical analysis was carried out using SPSS-PC 19.0 (SPSS, Inc., Chicago, IL). Data are presented as mean (standard deviation) for continuous variables with normal distribution, as median (interquartile range) for continuous variables with skewed distribution and as percentage for categorical variables. Student t-test was used to compare variables with normal distribution, Mann–Whitney U test for variables with abnormal distribution, and Chi-square test was used to compare frequencies among participants with and without arterial hypertension. The association between the presence of hypertension and surrogate insulin resistance indexes was assessed with binomial logistic regressions, in which those in quartile 4 of each index were compared with those in quartiles 1 through 3. Goodness-of-Fit analysis was performed using multinomial logistic regression to compare models that included separately each index and its individual components as predictors of hypertension. The covariates included in all models were age, gender, lifestyle, and biological parameters and use of medications known to influence insulin resistance. Comparisons were made using the values of pseudo R square (McFaden), Akaike information criterion (AIC) and Bayesian information criterion (BIC). A two-tailed P-value ≤0.05 was considered statistically significant.

Results

From a total number of 2124 participants who signed the informed consent and were included in the study, 154 did not return for Visit 2 and were excluded from the analysis. After further excluding participants with diabetes and/or treatment with fibrates (N = 240), the study group consisted of 1730 participants (Fig. 1), of whom 733 (42.4%) had hypertension.

FIG. 1.

Summary of patients’ identification from SEPHAR III database included in the analysis.

Compared to nonhypertensive participants, the hypertensive ones were older, less frequently smokers, with a higher mean SBP and DBP, with a less favorable metabolic profile (blood glucose, HbA1c, total and LDL-cholesterol, serum triglycerides), lower levels of eGFR, and had a higher frequency of CVDs. Gender, living in urban area, sedentary lifestyle, drinking habits, HDL-cholesterol, urinary albumin creatinine ratio, and urinary sodium excretion were similar between hypertensive and normotensive subjects. All surrogate insulin resistance indexes were significantly higher in participants with hypertension compared with those free of hypertension (Table 1).

Table 1.

Characteristics of the Study Population According to Presence or Absence of Hypertension

	No HTN (n = 997)	HTN (n = 733)	P
Age, years	41.3 (16.2)	54.4 (16.0)	<0.001
Male gender (%)	44.7	49.1	0.071
Urban (%)	60.5	58.1	0.322
Current smoker (%)	30.3	20.5	<0.001
Drinking (%)	51.9	50.8	0.940
Sedentary (%)	24.4	23.3	0.615
Cardiovascular disease (%)
Ischemic heart disease	19.8	29.6	<0.001
Cerebrovascular disease	1.2	4.5	<0.001
Peripheral artery disease	1.5	3.0	0.033
Body mass index, kg/m²	26.4 (5.2)	29.6 (5.8)	<0.001
Waist circumference, cm	91.3 (14.2)	98.3 (15.1)	<0.001
Blood glucose, mg/dL	93.0 (88.0–100.0)	99.0 (91.0–108.0)	<0.001
HbA1c, %	5.2 (5.0–5.5)	5.4 (5.1–5.7)	<0.001
Total cholesterol, mg/dL	191.4 (42.2)	202.6 (43.6)	<0.001
HDL cholesterol, mg/dL	56.3 (16.3)	55.1 (16.7)	0.108
Triglycerides, mg/dL	85.0 (61.0–124.0)	112.0 (80.0–155.0)	<0.001
LDL cholesterol, mg/dL	125.5 (39.3)	133.9 (41.8)	<0.001
Serum creatinine, mg/dL	0.76 (0.66–0.89)	0.78 (0.68–0.91)	<0.001
eGFR, mL/min/1.73 m²	111.4 (19.0)	99.7 (20.0)	<0.001
Uric acid, mg/dL	4.7 (1.3)	5.2 (1.4)	<0.001
Urinary albumin creatinine ratio, mg/g	3.7 (2.4–6.5)	4.6 (2.8–8.5)	0.05
Urinary sodium (spot), mEq/L	129.0 (83.0–166.0)	129.0 (90.0–172.0)	0.151
TG/HDL-C	1.6 (1.0–2.6)	2.1 (1.3–3.4)	<0.001
TyG	8.4 (0.6)	8.6 (0.5)	<0.001
TyG-BMI	221.9 (51.5)	256.4 (58.1)	<0.001
TyG-WC	759.4 (147.7)	859.0 (156.3)	<0.001
VAI	2.4 (0.4)	2.7 (0.4)	<0.001
LAP	27.3 (14.6–51.7)	48.1 (26.6–76.0)	<0.001

Data are presented as mean (standard deviation) for normally distributed variables, as median (interquartile range) for skewed variables, and as percentage for categorical variables.

HTN, hypertension; HbA1c, glycated hemoglobin A1c; HDL cholesterol, high-density lipoprotein cholesterol; LDL cholesterol, low-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; TG/HDL-C, triglyceride to HDL cholesterol ratio; TyG, triglyceride glucose; TyG-BMI, triglyceride glucose-body mass index; TyG-WC, triglyceride glucose-waist circumference; VAI, visceral adiposity index; LAP, lipid accumulation product.

The use of medications known to influence insulin resistance in hypertensive vs. nonhypertensive subgroups was as follows: ACEIs 28.9 versus 0%, ARBs 7.0 versus 0.1%, thiazide and thiazide-like diuretics 17.9 versus 1.2%, beta-blockers 12.1 versus 5.4%, (P < 0.001 for all), and statins 3.9 versus 4.5% (P = NS). None of the study subjects was treated with ezetimibe or other lipid-lowering drugs or with antiobesity medication.

ORs and 95% CI for TG/HDL-C, TyG, TyG-BMI, TyG-WC, VAI, and LAP are presented in Table 2. In unadjusted model, all six indexes were significantly associated with the presence of hypertension. After adjustments for age, gender, smoking, drinking, and sedentary lifestyle (Model 2), and for all variables in model 2 plus eGFR, urinary sodium (spot), urinary albumin creatinine ratio (Model 3), all indexes remained significantly associated with hypertension.

Table 2.

ORs and 95% CI for Highest Versus First Three Quartiles in Binary Logistic Regressions Predicting Presence of Hypertension

	Model 1	Model 2	Model 3
TG/HDL-C	1.88 (1.48–2.39)	1.57 (1.21–1.582)	1.54 (1.18–2.01)
TyG	2.58 (2.01–3.31)	1.85 (1.41–2.44)	1.83 (1.39–2.41)
TyG-BMI	2.90 (2.25–3.73)	2.14 (1.64–2.81)	2.12 (1.62–2.78)
TyG-WC	3.46 (2.68–4.47)	1.81 (1.25–2.60)	1.78 (1.23–2.57)
VAI	2.90 (2.27–3.70)	1.97 (1.51–2.57)	1.94 (1.48–2.53)
LAP	2.97 (2.33–3.80)	2.13 (1.63–2.78)	2.09 (1.60–2.73)

Model 1, unadjusted. Model 2 was adjusted for age, gender, smoking, drinking, and sedentary lifestyle. Model 3 was adjusted for all variables in model 2 and eGRF, urinary sodium (spot), and urinary albumin creatinine ratio.

Multinomial logistic models that considered separately each index and their individual components as predictors of hypertension were constructed. Each model included age, gender, smoking, drinking, sedentary lifestyle, eGFR, urinary sodium, urinary albumin creatinine ratio, and use of medications known to influence insulin resistance (ACEIs, ARBs, thiazide and thiazide-like diuretics, beta-blockers, and statins) as covariates. All individual components and surrogate insulin resistance indexes were independently associated with the presence of hypertension, with OR (95% CI) of 1.41 (1.06–1.87) for fasting blood glucose (P = 0.019), 2.06 (1.56–2.72) for TG (P < 0.001), 0.68 (0.50–0.92) for HDL-C (P = 0.013), 2.40 (1.81–3.19) for WC (P < 0.001), 2.54 (1.93–3.35) for BMI (P < 0.001), 1.71 (1.27–2.29) for TG/HDL-C (P < 0.001), 2.11 (1.57–2.84) for TyG (P < 0.001), 2.43 (1.82–3.25) for TyG-BMI (P < 0.001), 2.85 (2.10–3.85) for TyG-WC (P < 0.001), 2.60 (1.95–3.46) for VAI (P < 0.001), and 2.46 (1.84–3.29) for LAP (P < 0.001).

Pseudo R square, AIC, and BIC values of the models are presented in Table 3. Values of pseudo R square ranged from 0.342 for the multivariate model including TG/HDL-C to 0.357 for the model including TyG-WC, but with no clear superiority of any of the tested indexes over all others. Models including BMI and WC had similar ability to predict presence of hypertension as most of the surrogate indexes and were slightly superior to TG/HDL-C and TyG.

Table 3.

Goodness-of-Fit Analysis (Multinominal Logistic Regression) for Individual Components and for Surrogate Indexes of Insulin Resistance Predicting Presence of Hypertension

	Pseudo R square (McFaden)	AIC	BIC	P for model
Fasting blood glucose	0.339	1531	1612	<0.0001
TG	0.348	1511	1592	<0.0001
HDL-C	0.340	1530	1611	<0.0001
WC	0.353	1500	1581	<0.0001
BMI	0.356	1493	1574	<0.0001
TG/HDL-C	0.342	1524	1605	<0.0001
TyG	0.348	1512	1593	<0.0001
TyG-BMI	0.353	1501	1582	<0.0001
TyG-WC	0.357	1490	1571	<0.0001
VAI	0.356	1494	1575	<0.0001
LAP	0.353	1499	1581	<0.0001

Each model included age, gender, smoking, drinking, sedentary lifestyle, eGFR, urinary sodium, urinary albumin creatinine ratio, and use of medications known to influence insulin resistance (ACEIs, ARBs, thiazide and thiazide-like diuretics, beta-blockers, and statins) as covariates.

AIC, Akaike information criterion; BIC, Bayesian information criterion.

Discussion

The relationship between total and abdominal adiposity and hypertension is well demonstrated. In a cross-sectional study, including 25,196 adults aged 18–74 years,²³ it was shown that both BMI and WC are positively associated with elevated blood pressure. Nevertheless, not all individuals with excessive weight will develop hypertension. Abdominal adiposity is considered a surrogate for the effect of excessive visceral fat on risk of hypertension.²⁴ Numerous studies suggest that excess visceral fat represents the cause of metabolic abnormalities leading to increased insulin resistance and cardiometabolic risk, including risk of hypertension.^25,26 Therefore, measurement of insulin resistance could add valuable information to stratification of the risk for developing hypertension.

Identification of insulin-resistant individuals represents a challenge in clinical practice. The gold standard for estimating insulin resistance is the hyperinsulinemic euglycemic clamp,²⁷ a time- and resource-consuming procedure, which makes it unsuitable for large population samples. HOMA-IR is recognized as an indirect marker of insulin resistance, which reflects fasting (hepatic) and not postprandial (peripheral) insulin resistance.²⁸ A meta-analysis, including five prospective epidemiological studies, confirmed that individuals in the highest category of HOMA-IR had a 43% higher subsequent risk of hypertension compared with those in the lowest category.⁵

Nevertheless, use of HOMA-IR in clinical practice, especially in lower resource settings, is limited due to relative high cost and low accessibility of measuring plasma insulin. Therefore, more affordable surrogate indexes of insulin resistance incorporating routine biochemical parameters and measurements of adiposity such as BMI and WC could represent an alternative to identify insulin-resistant subjects at high risk of developing hypertension.

Our study showed that six easily measurable surrogate indexes of insulin resistance are independently associated with the presence of hypertension after adjustment for age, gender, smoking, drinking, sedentary lifestyle, eGFR, urinary sodium (spot), urinary albumin creatinine ratio, and medications known to influence insulin resistance. The most important finding of our study was that none of the indexes was superior to others as predictor of hypertension. Moreover, models including BMI and WC had similar ability to predict presence of hypertension as most of the surrogate indexes and were slightly superior to TG/HDL-C and TyG. To date, only a few studies reported the relationship between surrogate insulin resistance indexes and blood pressure or hypertension. In a sample of nearly 10,000 participants from the third National Health and Nutrition Examination Survey,⁹ LAP performed better than BMI for identifying the adverse level of nine cardiovascular risk factors, but not for adverse levels of systolic and diastolic blood pressure. Our study, having the same cross-sectional design, found that LAP is significantly associated with the presence of hypertension, but we did not analyze the relationship of LAP with systolic and diastolic blood pressures as continuous variables.

In another cross-sectional study, which included 2,244 healthy college students (17–24 years old), women with TG/HDL-C > 2.5 and men with TG/HDL-C > 3.5 had significantly higher systolic and diastolic blood pressure values, but whether TG/HDL was independently associated with levels of blood pressure was not examined.¹⁵ We found that TG/HDL-C was significantly higher in participants with hypertension and that TG/HDL-C was significantly associated with the presence of hypertension in unadjusted and adjusted models.

In a prospective study that included 1,417 first-degree relatives of patients with type 2 diabetes aged 30–70 years,¹⁸ free of diabetes and hypertension at inclusion and followed for 7 years, participants with higher values of VAI, BMI, waist-to-hip ratio (WHR), and hypertriglyceridemic-waist had a higher risk of developing hypertension after adjustment for age and sex, but the strongest predictors were waist-to-height ratio and WC. We also found that VAI was independently associated with hypertension and that models, including VAI, had a similar ability to predict the presence of hypertension in multinomial logistic regression analysis compared with BMI and WC

More recently, two prospective studies reported that VAI was a powerful predictor of incident hypertension in individuals with prehypertension at baseline and that its predictive value was superior to that of BMI, WC, and WHR if VAI is combined with WC.^29,30 Nevertheless, in these two studies the comparison was made with the receiver operating curve analysis and did not take into account the effect of any confounder.

The strength of our study is that we performed a comparison of six surrogate insulin resistance indexes and of their individual components regarding their ability to predict the presence of hypertension in a large sample representative for our population excluding those with diabetes and hypertriglyceridemia requiring pharmacological treatment. We were also able to account for the influence of most of the conventional factors known to influence the risk of hypertension (smoking, drinking, sedentary lifestyle, urinary albumin creatinine ratio, estimated glomerular filtration rate, and urinary sodium excretion, and use of different drug classes with known effects on insulin resistance).

The main limitation is the cross-sectional design of the study, which does not allow drawing conclusions regarding the value of these surrogate insulin resistance indexes as predictors of incident hypertension. Another limitation is that we were not able to directly measure insulin resistance in our study population and to further compare the surrogate indexes with direct markers of insulin resistance.

In conclusion, our study demonstrates that TG/HDL-C, VAI, LAP, TyG, TyG-BMI, and TyG-WC were independently associated with the presence of hypertension, but no superiority could be demonstrated over the use of BMI and WC as predictors of hypertension after full adjustment for conventional risk factors and use of medication in this cross-sectional study.

Footnotes

Acknowledgments

The authors express their gratitude to those who contributed to the SEPHAR III study, especially to all doctors, nurses, and residents involved in the conduct of this study.

This study was funded by the Romanian Society of Hypertension. The funding organization did not influence the study design, data collection and analysis, decision to publish, or preparation of the article.

Author Disclosure Statement

No competing financial interests exist in relationship to this work.

References

World Health Organization. World Health Statistics 2014. World Health Organization, 2014:175

Mozaffarian

, Benjamin

, Go

, et al. Heart disease and stroke statistics—2015 update: A report from the American Heart Association. Circulation, 2015; 131:e29–e322.

Landsberg

, Aronne

, Beilin

, et al. Obesity-related hypertension: Pathogenesis, cardiovascular risk, and treatment-a position paper of the tobesity society and the American society of hypertension. Obesity, 2013; 21:8–24.

Schulman

, Zhou

M-S

. Vascular insulin resistance: A potential link between cardiovascular and metabolic diseases. Curr Hypertens Rep, 2009; 11:48–55.

Wang

, Han

, Hu

. Fasting insulin, insulin resistance and risk of hypertension in the general population: A meta-analysis. Clin Chim Acta, 2017; 464:57–63.

Hanak

, Munoz

, Teague

, et al. Accuracy of the triglyceride to high-density lipoprotein cholesterol ratio for prediction of the low-density lipoprotein phenotype B. Am J Cardiol, 2004; 94:219–222.

Simental-Mendía

, Rodríguez-Morán

, Guerrero-Romero

. The product of fasting glucose and triglycerides as surrogate for identifying insulin resistance in apparently healthy subjects. Metab Syndr Relat Disord, 2008; 6:299–304.

Riscimanna

ANC

. Visceral adiposity index. Diabetes Care, 2010; 33:920–922.

Kahn

. The “lipid accumulation product” performs better than the body mass index for recognizing cardiovascular risk: A population-based comparison. BMC Cardiovasc Disord, 2005; 5:26.

10.

, Wu

, Chou

, et al. Triglyceride glucose-body mass index is a simple and clinically useful surrogate marker for insulin resistance in nondiabetic individuals. PLoS One, 2016; 11:e0149731.

11.

Vega

, Barlow

, Grundy

, et al. Triglyceride-to-high-density-lipoprotein-cholesterol ratio is an index of heart disease mortality and of incidence of type 2 diabetes mellitus in men. J Invest Med, 2014; 62:345–349.

12.

Salazar

, Carbajal

, Espeche

, et al. Identification of cardiometabolic risk: Visceral adiposity index versus triglyceride/HDL cholesterol ratio. Am J Med, 2014; 127:152–157.

13.

Janghorbani

, Amini

. The Visceral Adiposity Index in Comparison with Easily Measurable Anthropometric Markers Did Not Improve Prediction of Diabetes. Can J Diabetes, 2016; 40:393–398.

14.

Wang

, He

, et al. Predictive value of visceral adiposity index for type 2 diabetes mellitus: A 15-year prospective cohort study. Herz, 2015; 40:277–281.

15.

Ioachimescu

, Brennan

, Hoar

, et al. The Lipid Accumulation Product and All-cause Mortality in Patients at High Cardiovascular Risk: A PreCIS Database Study. Obesity, 2010; 18:1836–1844.

16.

Ray

, Ravichandran

, Nanda

. Comparison of lipid accumulation product index with body mass index and waist circumference as a predictor of metabolic syndrome in Indian population. Metab Syndr Relat Disord, 2018; 16:240–245.

17.

Nita

, Hancu

, Rusu

, et al. Hypertensive waist: First step of the screening for metabolic syndrome. Metab Syndr Relat Disord, 2009; 7:105–110.

18.

Janghorbani

, Aminorroaya

, Amini

. Comparison of different obesity indices for predicting incident hypertension. High Blood Press Cardiovasc Prev, 2017; 24:157–166.

19.

Dorobantu

, Darabont

, Dimulescu

, et al. New national epidemiological survey for the assessment of trend in hypertension's prevalence, treatment and control among the adult population of Romania: SEPHAR III - design and methodology. J Hypertens Res, 2016; 2:143–152.

20.

Levey

, Stevens

, Schmid

, et al. CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med, 2009; 150:604–612.

21.

American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care, 2019; 42:S13–S28.

22.

Task Force for the management of arterial hypertension of the European Society of Hypertension; Task Force for the management of arterial hypertension of the European Society of Cardiology. 2013 ESH/ESC Guidelines for the management of arterial hypertension. Blood Press, 2013; 22:193–278.

23.

Deng

W-W

, Wang

, Liu

M-M

, et al. Body Mass Index Compared with Abdominal Obesity Indicators in Relation to Prehypertension and Hypertension in Adults: The CHPSNE Study. Am J Hypertens, 2013; 26:58–67.

24.

Ostchega

, Hughes

, Terry

, et al. Abdominal Obesity, Body Mass Index, and Hypertension in US Adults: NHANES 2007–2010. Am J Hypertens, 2012; 25:1271–1278.

25.

Kotchen

, Grim

, Kotchen

, et al. Altered Relationship of Blood Pressure to Adiposity in Hypertension. Am J Hypertens, 2008; 21:284–289.

26.

Manrique

, Lastra

, Gardner

, et al. The Renin Angiotensin Aldosterone System in Hypertension: Roles of Insulin Resistance and Oxidative Stress. Med Clin North Am, 2009; 93:569–582.

27.

Tam

, Xie

, Johnson

, et al. Defining insulin resistance from hyperinsulinemic-euglycemic clamps. Diabetes Care, 2012; 35:1605–1610.

28.

Matthews

, Hosker

, Rudenski

, et al. Homeostasis model assessment: Insulin resistance and B-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia, 1985; 28:412–419.

29.

Zhang

, Shi

, Zhang

, et al. Visceral adiposity index (VAI), a powerful predictor of incident hypertension in prehypertensives. Intern Emerg Med, 2018; 13:509–516.

30.

Fiorentino

. Visceral adiposity index (VAI), a powerful predictor of incident hypertension in prehypertensives. Intern Emerg Med, 2018; 13:471–473.