|Year : 2021 | Volume
| Issue : 3 | Page : 331-337
Association of high-density lipoprotein cholesterol subfractions with insulin resistance in Nigerians with type 2 diabetes mellitus
Chikezie Hart Onwukwe1, Nkiru Ifeoma Chikezie2, Kalu Okorie3, Paul Osemeke Nwani4, Augustine Efedaye Ohwovoriole5
1 Al Isawiya General Hospital, Directorate of Al Gurayat, Ministry of Health, Kingdom of Saudi Arabia
2 Department of Community Medicine, Nnamdi Azikiwe University Teaching Hospital, Nnewi, Anambra State, Nigeria
3 Cedarcrest Hospitals, Gudu, Abuja, Nigeria
4 Faculty of Medicine, College of Health Sciences, Nnamdi Azikiwe University, Nnewi Campus, Nigeria
5 College of Medicine, University of Lagos, Lagos, Nigeria
|Date of Submission||14-Feb-2021|
|Date of Decision||30-Mar-2021|
|Date of Acceptance||12-Apr-2021|
|Date of Web Publication||30-Sep-2021|
Dr. Chikezie Hart Onwukwe
Al Isawiya General Hospital, Directorate of Al Gurayat, Ministry of Health.
Kingdom of Saudi Arabia
Source of Support: None, Conflict of Interest: None
Background: There are conflicting reports on the relationship of each of the two subfractions of high-density lipoprotein cholesterol (HDL-C) with insulin resistance in patients with type 2 diabetes mellitus (T2DM). Information on the relationship between HDL-C subfractions and insulin resistance in Nigerian patients with T2DM is not available in the literature. Aim: The aim of this article is to determine the association between subfractions of HDL-C and insulin resistance in Nigerian patients with T2DM. Materials and Methods: Patients with T2DM who were being managed by the Endocrinology, Diabetes, and Metabolism unit of the Nnamdi Azikiwe University Teaching Hospital (NAUTH), Nnewi, South-East Nigeria were recruited for this study, whereas individuals with normal glucose tolerance (NGT) were recruited from Nnewi town. The study was carried out within a 5-month period. History was taken and physical examination was done on study participants. Fasting venous samples were collected for plasma glucose, plasma total HDL-C, plasma HDL-C phenotypes, serum C-peptide, serum creatinine, and serum alanine transferase estimation. Homeostasis Model Assessment for Insulin Resistance (HOMA2-IRC-peptide) score was determined using plasma glucose and serum C-peptide concentrations. Data were analyzed using appropriate statistical software. Results: A total of 616 participants consisting of 400 participants with T2DM and 216 participants with NGT were recruited for the study. Difference in age, sex, and blood pressure was not statistically significant between participants with T2DM and those with NGT. There were statistically significant differences in body mass index, fasting plasma glucose, C-peptide, HOMA-IR, total HDL-C, and HDL-C subfractions between subjects with T2DM and those with NGT. There was a significant negative correlation between HOMA-IR score and each of HDL2-C (rs= −0.513, P < 0.01), HDL3-C (rs= −0.471, P < 0.01), and HDL2-C/HDL3-C ratio (rs= −0.416, P < 0.01) in subjects with T2DM. Total HDL-C (odds ratio (OR)=3, P = 0.02), HDL2-C (OR=3.87, P = 0.01), and HDL3-C (OR=2.54, P = 0.02) were significant predictors of insulin resistance in individuals with T2DM after univariate and multivariate logistic regressions. Conclusion: This study showed a negative correlation between insulin resistance and each of the HDL-C subfractions in Nigerian patients with T2DM with HDL2-C having the strongest correlation with HOMA-IR score.
Keywords: High-density lipoprotein, insulin resistance, Nigeria, subfraction
|How to cite this article:|
Onwukwe CH, Chikezie NI, Okorie K, Nwani PO, Ohwovoriole AE. Association of high-density lipoprotein cholesterol subfractions with insulin resistance in Nigerians with type 2 diabetes mellitus. J Diabetol 2021;12:331-7
|How to cite this URL:|
Onwukwe CH, Chikezie NI, Okorie K, Nwani PO, Ohwovoriole AE. Association of high-density lipoprotein cholesterol subfractions with insulin resistance in Nigerians with type 2 diabetes mellitus. J Diabetol [serial online] 2021 [cited 2021 Dec 2];12:331-7. Available from: https://www.journalofdiabetology.org/text.asp?2021/12/3/331/327302
| Introduction|| |
Impaired insulin secretion, insulin resistance, reduced tissue glucose uptake in the splanchnic circulation, and hepatic gluconeogenesis play key roles in the pathogenesis of type 2 diabetes mellitus (T2DM).
Dietary triglycerides (TGs) in diet are digested by intestinal lipases into monoglycerides and free fatty acids (FFAs). Cholesterol esters are de-esterified by pancreatic esterase into cholesterol. Triglycerides are transported by chylomicrons from intestinal cells through lymphatics into the blood. Lipoprotein lipase converts chylomicron TG into FFA and glycerol for energy use by fat and skeletal muscle cells. High-density lipoproteins (HDLs) are produced by enterocytes and liver cells. The role of HDL is to transport cholesterol from peripheral tissues and other lipoproteins to other tissues and lipoproteins using cholesteryl ester transfer protein (CETP).
There are two subfractions of HDL: HDL2 and HDL3, with densities of 1.063–1.125 and 1.125–1.210 kg/L, respectively. HDL2 contains apo A-I, whereas HDL3 contains both apo A-I and A-II. HDL2 particles are partially depleted of cholesteryl esters and enriched in triglycerides. Hepatic lipase hydrolyzes the triglyceride-enriched HDL2 molecules, thereby regenerating HDL3 and yielding particles that accept cholesterol from peripheral cells. The apo A-I HDL particles are associated with the cholesterol efflux-promoting effects of HDL; particles containing both apolipoproteins (e.g. HDL3) are less effective at mobilizing cholesterol from peripheral stores and appear to have various other functions. The clinical significance of the different HDL subfractions is not fully understood but previous studies have demonstrated that the HDL2 cholesterol is a better predictor of coronary atherosclerosis and insulin resistance than HDL3 or total HDL cholesterol (HDL-C) in persons with T2DM.,
Dyslipidemia accelerates atherosclerosis in patients with T2DM, thereby worsening the cardiovascular disease risk burden in these individuals. Reduction in HDL-C is a common lipid abnormality among patients with T2DM. The two subfractions of HDL-C are implicated in the development of cardiovascular disease in persons living with T2DM. There is conflicting information on the relationship between HDL-C subfractions and insulin resistance in patients with DM in various regions in the world, but no data of such relationship among Nigerians with T2DM exist. The main objective of this study was to determine the association of HDL-C subfractions and insulin resistance in Nigerian patients with T2DM.
| Materials and Methods|| |
This research was a cross-sectional observational study carried out in the side laboratory of the medical ward complex of the Nnamdi Azikiwe University Teaching Hospital (NAUTH), Nnewi, South-East Nigeria. Nnewi is a commercial town in Anambra state, Nigeria occupying an area of 2789 km2 with a population of 391,227 according to data from the 2006 Nigerian National Population Census.,
Inclusion criteria: T2DM patients aged 30–69 years were selected from the DM Unit Register of the NAUTH Endocrinology, Diabetes, and Metabolism (EDM) unit which contains personal information and management details of DM patients managed by the unit irrespective of DM type or outcome. Study participants aged 30–69 years were selected from the general population of Nnewi town. Normal glucose tolerance (NGT) was defined as fasting plasma glucose (FPG) less than 6.1 mmol/L in an individual without previous history of DM, whereas type 2 DM was defined as FPG of 7.0 mmol/L and above or use of medications for T2 DM.
Exclusion criteria: These include insulin therapy, use of lipid medications, hormonal contraception, steroid therapy, renal disease (defined as estimated glomerular filtration rate of less than 60 mL/1.73 m2/min), and necro-inflammatory liver disease (defined as serum alanine transferase greater than 45 IU/L which was the upper reference limit for alanine transferase in the NAUTH Chemical Pathology laboratory). Patients with T2DM on lipid-lowering agents were excluded from the study. Obese NGT subjects were also excluded from the study. The study was carried out within a 5-month period.
Relevant history was taken and physical examination was done on study participants. Weight, height, and waist circumference (WC) were determined using the WHO STEPS protocol for anthropometric measurements. Weight was measured with an electronic weighing scale (Secca 770 Floor Digital Scale, Hamburg, Germany), with the subject wearing light clothing and standing upright and with bare feet on each side of the scale. At the same time, the height is read off from the back using a portable stadiometer (Secca 240 wall-mounted, Hamburg, Germany) with the feet together and the patient looking straight ahead. Body mass index (BMI) was calculated as weight in kilograms divided by square of height in meters. The WC was measured midway between the inferior margin (lowest point) of the last rib and the crest of the ilium (top of the hip hone) using a measuring tape. Global obesity was defined as a BMI of 30 kg/m2, whereas abdominal obesity was defined as a WC of 94 cm (for males) and 80 cm (for females)., Blood pressure measurement was done using a mercury sphygmomanometer (Accoson, UK) with the appropriate cuff snuggly applied on the arm and a stethoscope applied on the brachial artery with the cuff inflated to determine the systolic BP (Korotkoff I) and diastolic BP (Korotkoff 5). Hypertension was defined as a systolic blood pressure of 140 mmHg and above or diastolic blood pressure of 90 mmHg or use of anti-hypertensive medications.
An aliquot of 10 mL of venous blood was collected from each participant between 8 and 10 a.m. after an overnight fast of 12–14 h; 2 mL of the collected blood sample was put in a fluoride oxalate tube for plasma glucose estimation using the Trinder glucose oxidase method (QCA, Spain). The remaining volume of blood was placed in a plain tube for plasma HDL-C subfraction estimation (precipitation methods), serum C-peptide estimation (enzyme-linked immunosorbent assay), serum creatinine estimation (alkaline picrate method), and serum alanine transferase estimation (dinitrophenylhydrazine method) in NAUTH Chemical Pathology Laboratory.,,, An aliquot of 0.5 mL of plasma was used to determine total HDL-C and HDL3-C with a kit (Technoclone GmbH, Vienna, Austria) based on precipitation methods. Plasma HDL2-C was obtained by subtraction of HDL3 from total HDL-C. Serum C-peptide levels were estimated using a kit (DIAsource ImmunoAssays SA, Belgium) based on the ELISA technique with 100 µL of serum. Samples were preserved in a freezer (Haier Thermocool Ltd, UK) at a temperature of −40°C before laboratory analysis using a spectrophotometer (Spectronic ZOD, Milton Roy Company, UK). Unknown concentrations of biochemical analytes were determined by interpolation of respective absorbances on the standard curves. The intra- and inter-assay coefficients of variation for assay performance were all within the acceptable performance limit of less than 10%. The University of Oxford Homeostasis Model Assessment (HOMA2) model software was used to determine the Homeostasis Model Assessment for Insulin Resistance (HOMA2-IRC-peptide) score using plasma glucose and serum C-peptide concentrations. HOMA2-IRC-peptide score greater than the 97.5th percentile (1.89 for males and 1.72 for females) value in participants with NGT defined insulin resistance.
Data were entered in a study protocol and subsequently transferred to Microsoft Office Excel® 2010 software for data management which involved data verification, data cleaning without replacement, data handling with confidentiality, and use of codes where necessary. Managed data were then transferred to the statistical software SPSS version 26 (IBM Corporation) for statistical analyses. The Kolmogorov–Smirnov test was used to determine normality of data distribution. Normal reference ranges were derived from respondents with NGT using the 2.5th percentile as the lower reference limit and 97.5th percentile as the upper reference limit. Continuous variables were presented as median (IQR), whereas categorical variables were presented as proportions, n (%). Logarithmic transformation of skewed data was done before correlation. Comparison of continuous variables between two groups was done using the Mann–Whitney U-test, whereas the strength of association between these variables was determined using the Spearman rank correlation coefficient, rs. P-value less than 0.05 defined statistical significance. Results were presented as tables, figures, and text.
| Results|| |
Six hundred and sixteen study participants consisting of 400 respondents with T2DM and 216 persons with NGT were recruited for this study. Baseline characteristics of study participants were summarized in [Table 1]. Participants with NGT comprised 83 (40.3%) males and 133 (59.7%) females, whereas those with T2DM consisted of 174 (43.5%) males and 226 (56.5%) females. The median age of participants with T2DM and NGT was 60 and 59 years, respectively (P = 0.55). There was no statistically significant sex difference between participants with T2DM and those with NGT (P = 0.21). There was also no statistically significant difference in systolic (P = 0.32) and diastolic BP (P = 0.28) between respondents with T2DM and those with NGT. There were statistically significant differences in BMI (P = 0.01), FPG (P = 0.01), C-peptide (P = 0.01), HOMA-IR score (P = 0.01), total HDL-C (P = 0.01), and HDL-C subfractions (P = 0.01) between individuals with T2DM and those with NGT.
Two hundred and ninety-two (73.0%) participants with T2DM had insulin resistance, whereas 108 (27.0%) persons with T2DM did not have insulin resistance [Table 2]. There were lower levels of total HDL-C and its subfractions in insulin-resistant patients with T2DM when compared with those without insulin resistance. The difference in total HDL-C (P = 0.03), HDL2-C (P = 0.01), and HDL3-C (P = 0.02) was statistically significant between insulin-resistant and non-insulin-resistant patients with T2DM.
|Table 2: Comparison of HDL-C subfractions in insulin-resistant and non-insulin-resistant participants with T2DM|
Click here to view
There was a strong negative correlation between HDL2-C and HOMA2-IRC-peptide score (rs= −0.513, P < 0.01) in participants with T2DM [Figure 1]. A significant inverse relationship between HDL3-C and HOMA2-IRC-peptide score (rs= −0.471, P < 0.01) was found in participants with T2DM [Figure 2]. The ratio of HDL2-C to HDL3-C had the weakest strength of association with HOMA2-IRC-peptide score (rs= −0.416, P < 0.01) in participants with T2DM when compared with the strength of association between HOMA2-IRC-peptide score and each HDL-C subfraction [Figure 3].
|Figure 1: Association between HDL2-C and Homeostasis Model for Insulin Resistance score in participants with T2DM. The inverse relationship between HDL2-C and HOMA2-IRC-peptide was defined by the linear regression equation: Log HDL2-C= −0.27 – 0.5 HOMA2-IRC-peptide (rs= −0.513, P < 0.01)|
Click here to view
|Figure 2: Correlation of HDL3-C with Homeostasis Model for Insulin Resistance score in study participants with T2DM. The negative correlation between HDL3-C and HOMA2-IRC-peptide was defined by the linear regression equation: Log HDL3-C= −0.06 – 0.17 HOMA2-IRC-peptide (rs= −0.471, P < 0.01)|
Click here to view
|Figure 3: Association of HDL2-C to HDL3-C ratio with Homeostasis Model for Insulin Resistance Score in study participants with T2DM. The negative correlation between HDL2-C and HOMA2-IRC-peptide was defined by the linear regression equation: Log HDL2-C/HDL3-C= −0.21 – 0.32 HOMA2-IRC-peptide (rs= −0.416, P < 0.01)|
Click here to view
Predictors of insulin resistance were determined in participants with T2DM [Table 3]. Univariate logistic regression analysis showed that hypertension (odds ratio (OR)=2.52, P = 0.02), BMI (OR=2.91, P = 0.01), WC (OR=3.87, P = 0.01), FPG (OR=4.04, P = 0.001), HDL-C (OR=2.40, P = 0.02), HDL2-C (OR=3.17, P = 0.01), and HDL3-C (OR=2.31, P = 0.02) were significant predictors of insulin resistance. When these significant predictors were replaced in a multivariate logistic regression model after controlling for age and sex, hypertension was not a significant predictor of insulin resistance, whereas the other variables still remained statistically significant predictors of insulin resistance. Total HDL-C, HDL2-C, and HDL3-C had adjusted OR of 3 (P = 0.02), 3.87 (P = 0.01), and 2.54 (P = 0.02), respectively, in predicting insulin resistance in patients with T2DM.
| Discussion|| |
This study found significant relationships between HDL-C subfractions and insulin resistance in Nigerian patients with T2DM, with HDL2-C having the stronger association with HOMA-IR score than HDL3-C and the HDL2-C/HDL3-C ratio. These findings have been compared with previous reports from other studies.
Maeda et al. found similar significant negative correlation between each of the HDL-C subfractions and insulin resistance, with HDL2-C having the stronger association with HOMA-IR score than HDL3-C. This was similar to the present study, although the study participants who partook in the study by Maeda et al. were Japanese Americans with and without T2DM. Ko et al. observed lower concentrations of HDL2-C and HDL3-C in male Chinese subjects with insulin resistance compared with those without insulin resistance. This is also in keeping with the present study, although the study participants were only males. Moriyama et al. found a negative association between HDL2-C/HDL3-C ratio and insulin resistance in non-diabetic Japanese subjects. This was similar to findings from the present study. The significant negative correlation between HDL2-C/HDL3-C and HOMA-IR was not as strong as that between each of the HDL-C subfractions and HOMA-IR score. The stronger association of the HDL2 molecule with insulin resistance has been explained by the possibility of the HDL2 playing a more functional role in the cardiovascular protectiveness of the HDL particle than the HDL3 subfraction.
Few studies have reported a stronger association between HDL3-C and HOMA-IR score compared with that between HDL2-C and HOMA-IR score. Stampfer et al. reported a strong association between HDL3-C and HOMA-IR score. This was a prospective study in which individuals were followed up for 5 years unlike the present study that is a cross-sectional study. Sweetnam et al. in the Caerphilly and Speedwell Collaborative Heart Disease Study also reported that the HDL3 particle correlated better with insulin resistance than the HDL2 subfraction. The use of different methodologies for HDL subfraction estimation in different laboratories may have contributed to disparity in study findings.
The anti-atherogenic properties of the HDL2-C subfraction have been studied, and it has been shown that metabolic pathways based on reverse cholesterol transport have linked it strongly to insulin resistance compared with the HDL3-C subfraction. The major strength of the present study is the large sample size of 616 study participants, but the cross-sectional design of the study limited monitoring of study participants in order to evaluate causality in study participants.
| Conclusion|| |
The present study showed a significant inverse relationship between each HDL-C subfraction and insulin resistance, with HDL2-C having the stronger association with insulin resistance. BMI, WC, FPG, total HDL-C, and HDL-C subfractions were significant predictors of insulin resistance in patients with T2DM. The odds of HDL2-C predicting insulin resistance were significantly higher than those of total HDL-C and HDL3-cholesterol.
The clinical implication of this finding is that the HDL2-C subfraction is a better correlate of insulin resistance than total HDL-C and HDL3-cholesterol. The HDL2-C subfraction should be measured in patients with T2DM in clinical practice because of its strong relationship with insulin resistance in these patients.
Ethical approval and patient consent
This research work was approved by the NAUTH Research Ethics Committee, Nnewi, Nigeria before commencement of the study. Written consent was obtained from volunteers after details of the research were explained to them. The research was carried out at no cost to study participants, and they were free to withdraw from the study at any stage if they so wished. Procedures carried out in this research work were in accordance with the ethical standards of the responsible committee on human experimentation (institutional or regional) and with the Helsinki Declaration of 1975, as revised in 2000.
Financial support and sponsorship
The study was funded by the research team. No grant or donation was provided by any external source.
Conflicts of interest
| References|| |
DeFronzo RA. Pathogenesis of type 2 diabetes mellitus. Med Clin North Am 2004;88:787-835, ix.
Link JJ, Rohatgi A, de Lemos JA. HDL cholesterol: Physiology, pathophysiology, and management. Curr Probl Cardiol 2007;32:268-314.
Syvänne M, Ahola M, Lahdenperä S, Kahri J, Kuusi T, Virtanen KS, et al
. High density lipoprotein subfractions in non-insulin-dependent diabetes mellitus and coronary artery disease. J Lipid Res 1995;36:573-82.
Gowri MS, Van der Westhuyzen DR, Bridges SR, Anderson JW. Decreased protection by HDL from poorly controlled type 2 diabetic subjects against LDL oxidation may be due to the abnormal composition of HDL. Arterioscler Thromb Vasc Biol 1999;19:2226-33.
Semenkovich CF, Goldberg AC, Goldberg IJ. Disorders of lipid metabolism. In: Melmed S, Polonsky KS, Larsen PR, Kronenberg HM, editors. Williams Textbook of Endocrinology. 13th ed. Philadelphia: Saunders Elsevier; 2016. p. 1660-87.
Structure Plan for Nnewi and Satellite Towns. United Nations Human Settlements Programme (UN-HABITAT), 2009.
National Population Commission. 2006 Provisional Census Results, Abuja: Federal Republic of Nigeria Official Gazette; 2007.
World Health Organization. Definition and Diagnosis of Diabetes Mellitus and Intermediate Hyperglycaemia. Report of a WHO/IDF Consultation. Geneva: WHO; 2006.
WHO STEPS Instruments. Available from: http//www.who.int/chp/steps. Accessed February 5, 2018.
International Diabetes Federation. The IDF Consensus worldwide definition of the Metabolic Syndrome. Consensus workshop on behalf of the IDF Task Force on Epidemiology and Prevention 2006.
Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr, et al
; National Heart, Lung, and Blood Institute Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure; National High Blood Pressure Education Program Coordinating Committee. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: The JNC 7 Report. J Am Med Assoc 2003;289:2560-72.
Marks V. An improved glucose-oxidase method for determining blood, C.S.F. And urine glucose levels. 1959. Clin Chim Acta 1996;251:19-24.
Hirano T, Nohtomi K, Koba S, Muroi A, Ito Y. A simple and precise method for measuring HDL-cholesterol subfractions by a single precipitation followed by homogenous HDL-cholesterol assay. J Lipid Res 2008;49:1130-6.
Little RR, Rohlfing CL, Tennill AL, Madsen RW, Polonsky KS, Myers GL, et al
. Standardization of C-peptide measurements. Clin Chem 2008;54:1023-6.
Hursden H, Rapaport A. Estimation of creatinine by the Jaffe reaction. A comparison of three methods. Clin Chem 1968;14:222-38.
Bergmeyer HU, Hørder M. International Federation of Clinical Chemistry. Scientific Committee. Expert panel on enzymes. IFCC document stage 2, draft 1; 1979-11-19 with a view to an IFCC recommendation. IFCC methods for the measurement of catalytic concentration of enzymes. Part 3. IFCC method for alanine aminotransferase. J Clin Chem Clin Biochem 1980;18:521-34.
Logan JE, Bayse DD, Koedam JC, Mather A, Wilding P. International Federation of Clinical Chemistry (IFCC). Scientific Committee, Analytical Section. IFCC/WHO principles and recommendations on evaluation of diagnostic reagent sets used in health laboratories with limited resources. Part 3. Selection and evaluation using reference materials. General considerations. J Clin Chem Clin Biochem 1984;22:573-82.
HOMA Calculator Version 2.2.2. Diabetes Trials Unit, University of Oxford. Available from: http://www.dtu.ox.ac.u/homa. Accessed February 12, 2018.
Maeda S, Nakanishi S, Yoneda M, Awaya T, Yamane K, Hirano T, et al
. Associations between small dense LDL, HDL subfractions (HDL2, HDL3) and risk of atherosclerosis in Japanese-Americans. J Atheroscler Thromb 2012;19:444-52.
Ko GT, Chan JC, Cockram CS. The association between dyslipidaemia and obesity in Chinese men after adjustment for insulin resistance. Atherosclerosis 1998;138:153-61.
Moriyama K, Negami M, Takahashi E. HDL2-cholesterol/HDL3-cholesterol ratio was associated with insulin resistance, high-molecular-weight adiponectin, and components for metabolic syndrome in Japanese. Diabetes Res Clin Pract 2014;106:360-5.
Stampfer MJ, Sacks FM, Salvini S, Willett WC, Hennekens CH. A prospective study of cholesterol, apolipoproteins, and the risk of myocardial infarction. N Engl J Med 1991;325:373-81.
Sweetnam PM, Bolton CH, Yarnell JW, Bainton D, Baker IA, Elwood PC, et al
. Associations of the HDL2 and HDL3 cholesterol subfractions with the development of ischemic heart disease in British men. The Caerphilly and Speedwell Collaborative Heart Disease studies. Circulation 1994;90:769-74.
Generoso G, Bensenor IM, Santos RD, Santos IS, Goulart AC, Jones SR, et al
. Association between high-density lipoprotein subfractions and low-grade inflammation, insulin resistance, and metabolic syndrome components: The ELSA-Brasil study. J Clin Lipidol 2018;12:1290-1297.e1.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3]