INTRODUCTION

Metabolic Syndrome (MS) has been widely studied in the world due to its negative repercussions on the health of individuals. This syndrome is a major public health problem because of its strong association with cardiovascular diseases and diabetes type 2. It is characterized by a set of metabolic changes and grouped risk factors, including central obesity, high levels of triglycerides (TG), high arterial pressure, low levels of cholesterol of lipoprotein of high density (HDL-c), and hyperglycemia¹.

The higher rates of mortality and morbidity in Brazil are a consequence of cardiometabolic diseases and diabetes ². The prevalence of MS in the Brazilian adult population, encountered through the National Health Research (NHR) of 2013, is 38.4%, being higher among women, related to low educational level and with advanced age ³. In a study with a rural population carried out from 2004 to 2005, 14.9% of the patients were diagnosed ⁴.

The conventional anthropometric measures such as the body mass index (BMI), waist circumference (WC), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR) fat percentage (%F) can predict multiple metabolic risk factors ⁵. However, some of the conventional parameters have been less precise to predict the MS because they provide limited information about corporal fat distribution ⁶.

The setting of substitute markers to easily and effectively diagnose is essential to the MS screening. Recently, new predictors were also validated in different populations, such as the visceral adiposity index (VAI), calculated based on the body mass index (BMI), waist circumference (WC), and lipid characteristics (TG and HDL-c), the lipid accumulation product (LAP) based on the combination of serum triglycerides (TG) and WC levels, and the product of plasmatic glucose levels and TG (index TyG) ^{7, 8, 9}.

Nevertheless, the relationship between components of the metabolic syndrome, anthropometric measures, and lipid indices, to the best of our knowledge, still was not evaluated in rural workers. Hence, the objective of this study was to verify the relation of MS components with lipidic indices and anthropometric parameters in rural workers.

METHODS

Study design

Transversal study.

Study population

We used the data of 167 rural workers selected in a sample by convenience from the Project “Triagem de fatores de risco relacionados ao excesso de peso em trabalhadores da agroindústria usando novas tecnologias analíticas e de informação em saúde” collected from 2013 to 2016. We included rural workers from Vale do Rio Pardo/RS, over the age of 18 years old, who had held all the variables for the lipid index calculation. We excluded pregnant women, individuals with neurological or motor impairments that could hamper the evaluation, and those who used insulin.

Procedures and data collection

We obtained the data in a single day through undergraduate and graduate students and trained professionals registered in a data bank and collected according to the description. We collected the demographic data (age and gender) through a standardized questionnaire of the project.

The biochemical parameters concerned fasting serum dosage glucose (GLI), TG, and HDL-c. We carried out the blood collection, through venipuncture technique, during the morning after twelve hours of night fasting. We carried out the biochemical analysis in serum and plasma samples (EDTA/Fluoride) in the Miura 200 (I.S.E., Rome, Italy) automatized equipment, utilizing commercial kits from Kovalent (Kovalent from Brazil). We carried out the collection and the biochemical analysis in the Exercise Biochemistry Laboratory from UNISC.

The anthropometrics parameters investigated were BMI, WC, WHR, and F%. We assessed the weight, height, WC, and skinfolds in the Physical Activity Laboratory of UNISC. We calculated the body mass index (BMI) with weight/height. We assessed the weight and the height utilizing the anthropometric scale platform type (Welmy SA, Santa Bárbara do Oeste, Brazil), capacity 150kg with the division of 100g and stadiometer coupled with a 1mm precision. We used the non-extendable measuring tape at the midpoint between the lower and upper coastal border of the iliac crest to measure the WC in a perpendicular plane. We determined the WHR through the relationship between the WC (cm) and the height (cm). We estimated the %F obtaining the corporal density calculated by the sum of seven skinfolds (chest, triceps, subscapular, suprailiac, abdominal, thigh, and midaxillary, with three repetitions) and later calculated the Siri’s equation 10 %F = [(4,95/DENS) – 4,50] x 100. We utilized the compass Lange model to measure the skinfolds.

We measured the systolic blood pressure (SBP), and the diastolic (DBP) with the individual rested in 5 minutes in a calm environment, with an empty bladder, sitting, with feet supported on the ground, two times with a mercury sphygmomanometer, according to the 7^th Brazilian Guideline for Arterial Hypertension ¹¹.

For the MS identification, we utilized the clinical criteria defined according to the harmonized criteria of the American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI) and the International Diabetes Federation (IDF)¹. The presence of any 3 or 5 risk factors constitutes a diagnosis of MS: WC ≥ 90 cm for men and ≥ 80cm for women (the cutoff for the South-American population), TG ≥ 150mg/dL or drug treatment, fasting blood glucose ≥ 100mg/dL or drug treatment, arterial pressure ≥ 130/85mmHg or drug treatment.

Regarding the lipid index:

● We calculated the VAI through the equations: Men= [WC/39,68 + (1,88 x BMI)] × [TG (mmol/L)/1,03] × [1,31/HDL-c (mmol/L)]; Women= [WC/36,58 + (1,89 × BMI)] × [TG (mmol/L)/0,81] × [1,52/HDL-c (mmol/L)] ⁷.

● We calculated the LAP through the equations: Men= [WC (cm) - 65] × [TG (mmol/l)]; Women= [WC (cm) - 58] × [TG (mmol/l)] ⁸.

● We calculated TyG as: Nl [fasting triglycerides (mg/dl) x fasting glucose (mg/dl) / 2], where Nl (natural logarithm) ⁹.

Ethical aspects

The Research Ethics Committee of the Universidade de Santa Cruz do Sul (CEP-UNISC) approved the project CAAE: 43252721.1.0000.5343. All the participants signed the Consent Form accordingly to Resolution 466/2012 of The National Health Council - Health Ministry.

Statistical analysis

We carried out the analysis in the SPSS (Statistical Package for the Social Sciences) software for Windows 23.0 version (IBM, Armonk, NY, USA). We presented the descriptive characteristics of the participants as frequency (relative frequency) and the continuous variables expressed as average ± standard deviation and/or median (interquartile range 25-75). We tested the normality using the Kolmogorov-Smirnov test. We utilized the t-Student test to compare the averages between genders, and we calculated the medians using the U test of Mann-Whitney. We utilized Spearman’s and Pearson’s correlation tests to calculate the correlation coefficients between continuous variables. The level of significance adopted was 5% (p<0.05).

We carried out the exploratory factorial analysis in two models: I- to investigate the relation between the components of MS (WC, PA, GLI, TG, and HDL-c) the anthropometric parameters (BMI, WHR, and %F); II- to investigate the relation between the components of MS and the lipidic indices (TyG, VAI e LAP).

We utilized the method of main components analysis aiming to reduce the number of original variables in less latent factors. The factorial analysis consists of three steps: factorial extraction, which produces the minimum number of factors that retain the maximum possible total variance of the original data; varimax rotation, to make the factors more easily interpretable; and interpretation-based on rotated factor loadings. Higher factor loadings represent a higher correlation between the variable and the latent factor.

The Kaiser-Mayer-Olkin (KMO) method was estimated in >0.6, and Bartlett's test of sphericity was significant (p <0.001) as an indicator of the adequacy of the sample in the analysis. We extracted the number of factors based on identified factors with eigenvalues >1. We interpreted the factors based on loads that relate the variables to the factor because higher factor loadings represent more correlation between the variable and the latent factor. We considered the factor loadings > 0.4 significant to identify the variables that compose a factor, according to Tsai et al.¹².

RESULTS

Out of the 167 rural workers who took part in this study, 79.0% (N= 132) were adults and 21.0% (N= 35) were older adults (over the age 60 years), and 60.5% (N= 101) were female and 39.5% were male. The average age was 50.49 ± 10.76 years (men = 50.62±10.82 years; women = 50.41±10.78 years), varying between 18 and 73 years.

The frequency of the metabolic syndrome was 61.1% (N= 102) in the final sample, 67.6% were female and 32.4% (N= 33) were male. We presented the anthropometric and metabolic characteristics of the final sample and by gender in Table 1. The female gender reported average/median significantly superior in comparison to males for BMI, WHR, %F, VAI, LAP. Males, on the other hand, presented a significantly superior average for WC.

We presented the correlation between the 12 analyzed variables that configure MS components, anthropometric measures, and TyG, LAP, and VAI indices in Table 2. In the correlation test among the anthropometric parameters and the components of MS, we found a significant correlation, positive and strong of BMI with WC (r=0.776) and weak of SBP, DBP, and TG. The WHR correlated positively, strongly, and significantly with WC (r=0.884), and with SBP, DBP, GLI, and TG. On the other hand, the %F obtained a weak correlation, positive and significant with WC, SBP, and TG. By correlating the lipid indices and the components of the metabolic syndrome, we determined a significant, strong, and positive correlation for TyG, LAP, and VAI with TG (r= 0.951, r= 0.864, r= 0.914, respectively). The VAI presented moderate and an inverse correlation with HDL-c (r= -0.607) and weak and significant correlation with WC (r= 0.295). The TyG and LAP indices presented an inverse proportional correlation with HDL-c (r= -0.380; r= -0.348) and positive with the other MS variables.

The factorial analysis of the metabolic syndrome components with anthropometric and lipid indices (VAI, LAP, TyG) identified three dominant factors with eigenvalue > 1 in both analyses. The analysis that included the lipid indices responded to the highest total variation of the data (78.4%) if compared with the analysis that included the anthropometric parameters (70.4%). We show the loading patterns of the factorial analysis, after the varimax rotation, in Table 3.

In the analysis that included the anthropometric parameters (model I), the BMI, WHtR, WC, and %F positively contributed to factor 1. The BMI presented a higher factorial load of this factor. This factor explained 40.02% of the total variance. Factor 3 explained 14.38% of the variation, and the variables TG and glucose presented a positive contribution, and the HDL-c negatively contributed to this factor. We interpreted factor 1 as a factor of weight excess/obesity. We interpreted factor 2 as the factor of blood pressure. We interpreted factor 3 as the factor of lipid/glycemic.

In the analysis that included the lipid indices (model II), factor 1 was responsible for the higher proportion of the total variance (46.62%) and was positively composed by VAI, LAP, TyG, and TG, while the HDL-c negatively contributed to this factor. VAI presented the most significant load of this factor. SBP, DBP, and WC positively contributed to factor 2, which explained 20.59% of the total variance. The glucose (GLI) positively contributed to factor 3, which explained 11.22% of the total variance. Thus, we interpreted factor 1 as a metabolic factor, factor 2 as cardiometabolic, and factor 3 as a glycemic factor.

DISCUSSION

In this study, we utilized the factorial analysis to reduce interrelated variables that are key components to the metabolic syndrome and/or predictors to three factors not correlated in a sample with rural workers. As far as we know, no previous study has investigated the grouping of components of MS in two distinct models, model I (anthropometric parameters) and model II (lipid indices).

Model II responded with the highest explained variation (78.43%). Included the metabolic factors (VAI, LAP, TyG, TG, and HDL-c), cardiometabolic (SBP, DBP, and WC), and glycemic. The model I explained 73.4% of the variation and included factors of weight excess, blood pressure, and lipid/glycemic. Our results suggest that the indices were better related to the components of the metabolic syndrome.

Similar to our findings, Shin and Kim ⁸, in their study, identified that LAP, VAI, TyG, and WHtR were positively correlated with WC, SBP, DBP, TG, and fasting glucose level, and negatively correlated with HDL-c. We found these findings in the total sample of men and women.

In a study carried out with individuals in the rural area of India also utilizing cardiometabolic variables in the factorial analysis, three factors were extracted and responded to 71% of the variation. Factor 1 was positively loaded by WC, TG, and very-low-density lipoprotein (VLDL) and negatively loaded by HDL-c. Factor 2 was positively loaded by total cholesterol low-density lipoprotein (LDL-c). Factor 3 was positively loaded by SBP and DBP ¹³.

Similar to the findings in our study in model I, a study carried out with rural and industry Brazilian workers identified that factor 1 also was strongly loaded with related variables to overweight, obesity, and visceral fat (BMI, visceral fat area, WC, %F). The second factor was loaded by TG, VAI, TyG, and LAP. Factor 3 included L-8, IL-6, IL-1β and GLI ¹⁴.

As we observed, similarly to our findings in model II, Shin and Kim⁸ and Deshmukh et al.¹³also found positive relation in TG and negative relation with HDL-c in factor 1. Our findings in the model I are similar to the results found in our previous research ¹⁴in factor 1, composed of variables related to the excess of weight/obesity.

The MS is a group of metabolic abnormalities that includes central obesity, insulin resistance, atherogenic dyslipidemia, and hypertension. The prevalence of MS corresponds, generally, to the prevalence of obesity ¹⁵, and parameters of general obesity, such as BMI, WC, and WHtR were associated with that syndrome. Among the mechanisms involved in obesity is the resistance to insulin in the adipose tissue, which damages the inhibition of lipolysis mediated through insulin, and there is an increase of circulating free fatty acids that inhibit the antilipolytic effect of insulin. However, the deposits of visceral fat contribute more to the resistance to insulin than the subcutaneous fat being the central obesity the most proposed because it is the main trigger for most of the endocrine and immune pathways of adipocytes involved in MS ^{15, 16..} There still is not a full comprehension of MS, but insulin resistance is the most accepted hypothesis for the underlying pathophysiology due to the excess of fatty acids because of inadequate lipolysis ¹⁷.

In the last decades, the TyG, VAI, and LAP indices were associated with strong predictive capability for the resistance against insulin ^{9, 18, 19}. These same indices are now usually tested to predict the MS ^{20, 21, 22}. In an evaluation of the predictive capacity and cut value of 11 parameters related to obesity (BMI, WC, WHR, WHtR, conicity index, VAI, TyG index, among others) in the identification of the MS in adults, we observed that the index TyG and VAI had the highest predictive output in different group ages (30-50 years old and 51-70 years old) in both genders ²³.

In our study, we applied factorial analysis in the general population, aiming to explore the higher variation between both models. The factorial analysis considers the underlying correlational structure between the individual markers, minimizes various test problems, and does not require biological suppositions, offering advantages regarding other approaches to create summary variables ²⁴. However, it is necessary larger explorations regarding the associations between the conventional anthropometric variables with the lipid indices in the prediction of the metabolic syndrome in the studied population, besides the transverse nature of this analysis and the sample size, not allowing us to extract the sample by gender, which constitutes a fragility in our study. In addition, and for future studies, we may test the results of the exploratory factorial analysis study using a set of independent data but including the same variables.

In conclusion, our study limited itself to exploring the grouping of factors in two independent models and demonstrated that model II, which included the lipid indices VAI, LAP, and TyG, explained the higher variation observed. Hence, the lipid indices investigated presented better relation to the MS components in rural workers, especially the lipid index VAI. However, we suggest that other studies must be carried out aiming to explore the relationship of the lipid indices with other variables such as age (adults and elderly people).

REFERENCES

1. Alberti KGMM, Eckel RH, Grundy SM, et al. Harmonizing the Metabolic Syndrome: A Joint Interim Statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation. 2009;120:1640–1645. https://doi.org/10.1161/circulationaha.109.192644

2. Marques, MV, Santos SSAN, Lima MV, et al. Distribuição espacial da mortalidade por diabetes no Brasil. Revista Saúde e Desenvolvimento Humano. 2020; 8(3): 113-122. http://dx.doi.org/10.18316/sdh.v8i3.6135

3. Oliveira LVA, Santos BNS, Machado IE, et al. Prevalência da Síndrome Metabólica e seus componentes na população adulta brasileira. Ciência & Saúde Coletiva 2020;25(11): 4269-4280. https://doi.org/10.1590/1413-812320202511.31202020

4. Pimenta AM, Gazzinelli A, Velasquez-Melendez G. Prevalência da síndrome metabólica e seus fatores associados em área rural de Minas Gerais (MG, Brasil). Ciênc saúde coletiva. 2011;16(7): 3297-3306. https://doi.org/10.1590/S1413-81232011000800029

5. Gu Z, Li D, He H, et al. Body mass index, waist circumference, and waist-to-height ratio for prediction of multiple metabolic risk factors in Chinese elderly population. Sci Rep. 2018;8(1):385. https://doi.org/10.1038/s41598-017-18854-1

6. Nascimento-Ferreira MV, Rendo-Urteaga T, Vilanova-Campelo RC, et al. The lipid accumulation product is a powerful tool to predict metabolic syndrome in undiagnosed Brazilian adults. Clin Nutr. 2017;36(6):1693-1700. https://doi.org/10.1016/j.clnu.2016.12.020

7. Amato MC, et al. Visceral Adiposity Index: a reliable indicator of visceral fat function associated with cardiometabolic risk. Diabetes Care. 2010;33(4):920-922. https://doi.org/10.2337/dc09-1825

8. Kahn HS. The "lipid accumulation product" performs better than the body mass index for recognizing cardiovascular risk: a population-based comparison. British Medical association Cardiovasc Disord. 2005; 5:1-26, 2005. https://doi.org/10.1186/1471-2261-5-26

9. Simental-Mendia LE, Rodriguez-Moran M, Guerrero-Romero F. 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. https://doi.org/10.1089/met.2008.0034

10. Heyward VH. Avaliação física e prescrição de exercício: técnicas avançadas. 6. ed. Porto Alegre: Artmed, 2013. xvi, 485 p.

11. Malachias MVB, Plavnik FL, Machado CA, et al. 7ª Diretriz Brasileira de Hipertensão Arterial: Capítulo 1 - Conceituação, Epidemiologia e Prevenção Primária. (2016). Arquivos Brasileiros de Cardiologia. 107(3, Suppl. 3), 1-6. https://doi.org/10.5935/abc.20160151

12. Tsai TY, Hsu PF, Lin CC, et al. Factor analysis for the clustering of cardiometabolic risk factors and sedentary behavior, a cross-sectional study. PloS one. 2020; 15(11): e0242365. https://doi.org/10.1371/journal.pone.0242365

13. Deshmukh PR, Kamble P, Goswami K, et al. Metabolic syndrome in the rural population of wardha, central India: an exploratory factor analysis. Indian J Community Med. 2013;38(1):33-38. https://doi.org/10.4103/0970-0218.106625

14. Couto AN, Kleinpaul WV, Renner JDP, et al. Nutritional status, insulin resistance and cytokine biomarkers: principal component analysis approach. IJDR. 2019; 09(10): 30659-30664. https://www.journalijdr.com/nutritional-status-insulin-resistance-and-cytokine-biomarkers-principal-component-analysis-approach

15. Rochlani Y, Pothineni NV, Kovelamudi S, et al. Metabolic syndrome: pathophysiology, management, and modulation by natural compounds. Ther Adv Cardiovasc Dis. 2017;11(8):215-225. https://doi.org/10.1177/1753944717711379

16. Nolan CJ, Prentki M. Insulin resistance and insulin hypersecretion in the metabolic syndrome and type 2 diabetes: Time for a conceptual framework shift. Diab Vasc Dis Res. 2019;16(2):118-127. https://doi.org/10.1177/1479164119827611

17. McCracken E, Monaghan M, Sreenivasan S. Pathophysiology of the metabolic syndrome. Clin Dermatol. 2018;36(1):14-20. https://doi.org/10.1016/j.clindermatol.2017.09.004

18. Guerrero-Romero F, Simental-Mendía LE, González-Ortiz M, et al. The product of triglycerides and glucose, a simple measure of insulin sensitivity. Comparison with the euglycemic-hyperinsulinemic clamp. J Clin Endocrinol Metab. 2010;95(7):3347-51. https://doi.org/10.1210/jc.2010-0288

19. Mirmiran P, Bahadoran Z, Azizi F. Lipid Accumulation Product Is Associated with Insulin Resistance, Lipid Peroxidation, and Systemic Inflammation in Type 2 Diabetic Patients. Endocrinology and Metabolism. 2014;29(4):443-449.

20. Li R, Li Q, Cui M, et al. Clinical surrogate markers for predicting metabolic syndrome in middle-aged and elderly Chinese. J Diabetes Investig. 2018;9(2):411-418. https://doi.org/10.1111/jdi.12708

21. Shin KA, Kim YJ. Usefulness Of Surrogate Markers Of Body Fat Distribution For Predicting Metabolic Syndrome In Middle-Aged And Older Korean Populations. Diabetes Metab Syndr Obes. 2019;12:2251-2259. https://dx.doi.org/10.2147%2FDMSO.S217628

22. Raimi TH, Dele-Ojo BF, Dada SA, et al. Triglyceride-Glucose Index and Related Parameters Predicted Metabolic Syndrome in Nigerians. Metab Syndr Relat Disord. 2021;19(2):76-82. https://doi.org/10.1089/met.2020.0092

23. Chiu TH, Huang YC, Chiu H, et al. Comparison of Various Obesity-Related Indices for Identification of Metabolic Syndrome: A Population-Based Study from Taiwan Biobank. Diagnostics (Basel). 2020;10(12):1081. https://doi.org/10.3390/diagnostics10121081

24. Brinkley TE, Hsu FC, Beavers KM, et al. Total and abdominal adiposity are associated with inflammation in older adults using a factor analysis approach. J Gerontol A Biol Sci Med Sci. 2012;67(10):1099-1106. https://doi.org/10.1093/gerona/gls077