The association of the visceral adiposity index with insulin resistance and beta-cell function in Korean adults with and without type 2 diabetes mellitus

Abstract

This study was conducted to assess the association of the visceral adiposity index (VAI) with insulin resistance and beta cell function in Korean adults with and without type 2 diabetes mellitus. The study was carried out using data from the 2015 Korean National Health and Nutrition Examination Survey (KNHANES VI-3) and included 4,922 adults, aged 20 or older. There were several key findings in the present study. First, in subjects without type 2 diabetes mellitus, homeostasis model assessment of insulin resistance (HOMA-IR) (p < 0.001) and beta cell function (HOMA-B) (p < 0.001), insulin (p < 0.001), fasting blood glucose (FBG) (p < 0.001), and metabolic syndrome (MetS) score (p < 0.001) were positively associated with quartiles of VAI. Second, in subjects with type 2 diabetes mellitus, HOMA-IR (p = 0.038), FBG (p = 0.007), and MetS score (p < 0.001) were positively associated with quartiles of VAI, but associations with HOMA-B (p = 0.879) and insulin (p = 0.104) were not significant. In conclusions, the visceral adiposity index is positively associated with insulin resistance and beta cell function in Korean adults without type 2 diabetes mellitus. The visceral adiposity index is positively associated with insulin resistance but not beta cell function in Korean adults with type 2 diabetes mellitus.

DIABETES is a common endocrine disease affecting global public health and is a principal factor contributing to cardiovascular events and mortality [1, 2]. An increase in insulin resistance is the most powerful predictor for the future development of type 2 diabetes (T2DM) [3]. A decrease in pancreatic beta-cell function is an early symptom of diabetes pathogenesis [4], and beta-cell dysfunction plays a role in the development of T2DM [5]. Therefore, normal maintenance of beta-cell function in the non-diabetic population is important for preventing the onset of T2DM [6]. For the measurement of insulin resistance and beta-cell function, homeostatic model assessment (HOMA) indices require only a single basal state measurement of insulin and glucose and are considered more appropriate for use in large epidemiological studies [7].

Adipose tissue plays a role in the regulation of glucose and lipid metabolism [8]. Visceral adiposity is associated with cardiovascular risk and insulin resistance [9, 10]. The visceral adiposity index (VAI) is an indicator of visceral fat distribution and dysfunction and is a gender-specific mathematical index based on simple anthropometric parameters (body mass index [BMI] and waist circumference [WC]) and metabolic parameters (triglycerides [TGs] and high-density lipoprotein cholesterol [HDL-C]) [11]. Some studies have reported that an increase in VAI is an important risk factor for cardiovascular disease and T2DM [12-14] and is positively associated with insulin resistance in populations with or without diabetes mellitus [15, 16].

Currently, despite extensive research on VAI and insulin resistance, the relationship between VAI and beta-cell function has received little attention, and results from studies of VAI and beta-cell function are not consistent across ethnic groups and countries. Thus, the present study investigated the association of VAI with HOMA-IR and HOMA-B in Korean adults with and without T2DM. We used data from the 2015 Korean National Health and Nutrition Examination Survey (KNHANES VI-3), which is representative of the Korean population.

Methods

Study Subjects

KNHANES is a cross-sectional survey conducted nationwide by the Division of Korean National Health and Welfare. The KNHANES VI-3 (2015) dataset is the most recent of the KNHANES datasets to include HOMA records and was performed from January 2015 to December 2015, sampling 7,380 individuals over one year of age. Among the 5,855 subjects who responded to the KNHANES VI-3 survey, we limited the analysis to adults aged ≥20 years. We excluded 895 subjects where data were missing for important analytic variables, including HOMA-IR, HOMA-B, and various blood chemistry tests. In addition, we excluded 38 subjects who received insulin treatment. Finally, 4,922 subjects were included in the statistical analysis. The KNHANES VI-3 (2015) study was conducted according to the principles expressed in the Declaration of Helsinki (Institutional Review Board No, 2015–01–02–6C). All survey participants agreed with the use of epidemiological research to identify risk factors and death causes of chronic diseases. Participants’ records and information in the KNHANES were anonymous and de-identified prior to analysis. Further information can be found in “The KNHANES VI-3 (2015) Sample,” which is available on the KNHANES website. The official website of KNHANES (http://knhanes.cdc.go.kr) is currently operating an English-​language information homepage. The data of the respective year are available to everyone free of charge. If the applicant completes a simple subscription process and provides his/her email address on the official website of KNHANES, the data of the respective year can be downloaded free of charge. If additional information is required, the readers may contact the department responsible for the storage of data directly (Su Yeon Park, [e-mail]).

General Characteristics and Blood Chemistry

Research subjects were classified by gender (men or women), anthropometric measurements, including body mass index (BMI) and waist circumference (WC), and systolic blood pressure (SBP) and diastolic blood pressure (DBP) measurements. Blood chemistry included measurement of total cholesterol (TC), triglycerides (TGs), high-density lipoprotein cholesterol (HDL-C), blood urea nitrogen (BUN), serum creatinine (Crea), high-sensitivity C-reactive protein (hs-CRP), insulin, and fasting blood glucose (FBG). Metabolic syndrome (MetS) was defined using the diagnostic criteria of the revised National Cholesterol Education Program Adult Treatment Panel III [17], including SBP, DBP, WC, TGs, HDL-C, and FBG. MetS comprises at least three of the five following medical conditions: abdominal obesity (WC ≥90 cm and 80 cm for men and women), abnormal blood pressure (BP ≥130/85 mmHg or treatment with antihypertensive medication), abnormal TGs (TGs ≥150 mg/dL or treatment for dyslipidemia), abnormal HDL-C (HDL-C <40 mg/dL and 50 mg/dL for men and women, respectively, or treatment for dyslipidemia), and abnormal FBG (FBG ≥100 mg/dL or treatment for hyperglycemia). Subjects without any of the five risk factors received a metabolic syndrome score (MetS score) of 0, while those with one, two, three, four, or five of the risk factors received a MetS score of 1, 2, 3, 4, or 5, respectively [18].

HOMA-IR and HOMA-B and T2DM and VAI and insulin and hs-CRP

Concentrations of insulin were measured by an immunoradiometric assay (INS-IRMA; Biosource, Belgium) using a 1470 Wizard Gamma Counter (PerkinElmer, Turku, Finland). Concentrations of hs-CRP were measured by an immunoturbidimetric method (Roche cardiac high-sensitivity C-reactive protein; Roche, Germany) using a Cobas analyzer (Roche, Germany). The formulae were as follows: HOMA-IR = [fasting insulin (μU/mL) × fasting blood glucose (mg/dL)]/405; HOMA-B = 360 × fasting insulin (μU/mL)/[fasting blood glucose (mg/dL) – 63] [19]. T2DM was defined as a fasting blood glucose level of ≥126 mg/dL or through the subject’s self-reported history of diabetes or use of diabetes medications [20]. VAI was calculated using the following formulae differentiated according to sex, where the TGs and HDL-C concentrations are expressed in mmol/L as described by Amato et al. [11]: for men, VAI = [WC/39.68 + (1.88 × BMI)] × (TGs/1.03) × (1.31/HDL-C); for women, VAI = [WC/36.58 + (1.89 × BMI)] × (TGs/0.81) × (1.52/HDL-C). The quartiles of VAI in non-T2DM were classified as follows: 1^st quartile, 0.90 or less; 2^nd quartile, 0.91–1.46; 3^rd quartile, 1.47–2.42; 4^th quartile, more than 2.43. The quartiles of VAI in T2DM were classified as follows: 1^st quartile, 1.38 or less; 2^nd quartile, 1.39–2.25; 3^rd quartile, 2.26–3.65; 4^th quartile, more than 3.66.

Data analysis

The collected data were statistically analyzed using SPSS WIN version 18.0 (SPSS Inc., Chicago, IL, USA). In statistical analyses, continuous variables were reported as mean ± standard deviation (M ± SD). Categorical variables were reported as percentages (%). Clinical characteristics grouped by men or women were analyzed using chi-square and an independent t-test (see Table 1). Clinical characteristics, according to the quartile of VAI, were calculated using chi-square and analysis of variance (ANOVA) in non-T2DM (see Table 2) and T2DM (see Table 3). We conducted multiple linear regression analysis for the independent factors determining HOMA-IR (see Table 4) and HOMA-B (see Table 5) in non-T2DM and T2DM. We conducted the analysis of covariance test (ANCOVA) for HOMA-IR and HOMA-B according to the quartiles of VAI after adjusting for age, gender, SBP, DBP, TC, BUN, Crea, and hs-CRP (see Table 6). The significance level for all the statistical data was set as p < 0.05.

Table 1 Clinical characteristics of research subjects

Variables	Category	Total (n = 4,922)	Non-T2DM (n = 4,380)	T2DM (n = 542)	p-value
Gender	Men	2,161 (43.9)	1,873 (42.8)	288 (53.1)	<0.001
	Women	2,761 (56.1)	2,507 (57.2)	254 (46.9)
Age (years)		51.65 ± 16.18	50.32 ± 16.14	62.37 ± 11.92	<0.001
SBP (mmHg)		119.40 ± 17.19	119.91 ± 17.77	126.74 ± 16.96	<0.001
DBP (mmHg)		75.29 ± 10.15	75.61 ± 10.43	74.47 ± 10.72	0.046
BMI (kg/m2)		23.96 ± 3.48	23.79 ± 3.41	25.34 ± 3.74	<0.001
WC (cm)		82.84 ± 9.94	82.14 ± 9.74	88.50 ± 9.70	<0.001
VAI		2.12 ± 2.14	2.00 ± 1.95	3.08 ± 3.15	<0.001
TC (mg/dL)		190.49 ± 35.99	191.69 ± 34.99	180.82 ± 41.99	<0.001
TGs (mg/dL)		136.57 ± 109.32	131.34 ± 100.95	178.83 ± 155.56	<0.001
HDL-C (mg/dL)		50.86 ± 12.90	51.56 ± 12.91	45.24 ± 11.43	<0.001
BUN (mg/dL)		14.67 ± 4.85	14.48 ± 4.79	16.18 ± 5.04	<0.001
Crea (mg/dL)		0.84 ± 0.34	0.84 ± 0.35	0.86 ± 0.21	0.152
hs-CRP (mg/L)		1.28 ± 2.37	1.21 ± 2.27	1.85 ± 3.04	<0.001
FBG (mg/dL)		100.66 ± 23.42	94.99 ± 9.77	146.45 ± 43.08	<0.001
Insulin (μU/mL)		8.01 ± 6.69	7.64 ± 5.68	11.02 ± 11.67	<0.001
MetS		1,469 (29.8)	1,116 (25.5)	353 (65.1)	<0.001
MetS score		1.73 ± 1.39	1.58 ± 1.34	2.94 ± 1.17	<0.001
HOMA-IR		2.07 ± 2.19	1.83 ± 1.47	4.04 ± 4.64	<0.001
HOMA-B		85.20 ± 65.24	88.78 ± 64.69	56.31 ± 62.50	<0.001

n (%), M ± SD

SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; WC, waist circumference; VAI, visceral adiposity index; TC, total cholesterol; TGs, triglycerides; HDL-C, high density lipoprotein cholesterol; BUN, blood urea nitrogen; Crea, creatinine; hs-CRP, high sensitivity C reactive protein; FBG, fasting blood glucose; MetS, metabolic syndrome HOMA-IR, homeostasis model assessment of insulin resistance; HOMA-B, homeostasis model assessment of beta-cell function. The continuous variables were reported as mean ± standard deviation (M ± SD). Categorical variables were reported as percentages (%). Clinical characteristics according to gender were analyzed using chi-square and an independent t-test.

Table 2 Clinical characteristics of subjects according to the quartiles of VAI in non-T2DM (n = 4,380)

Variables	Visceral adiposity index (VAI)				p-value	Post hoc analysis
	a Quartile 1 (≤0.90) (n = 1,095)	b Quartile 2 (0.91–1.46) (n = 1,095)	c Quartile 3 (1.47–2.42) (n = 1,095)	d Quartile 4 (>2.43) (n = 1,095)
Men	446 (40.7)	464 (42.4)	473 (43.3)	490 (44.7)	0.288
Age (years)	44.90 ± 16.61	48.75 ± 15.87	53.24 ± 15.67	54.39 ± 14.57	<0.001	a < b < c, d
SBP (mmHg)	114.72 ± 17.23	117.93 ± 14.19	122.39 ± 124.58	124.58 ± 17.19	<0.001	a < b < c < d
DBP (mmHg)	72.82 ± 9.99	74.79 ± 10.29	76.53 ± 10.13	79.30 ± 10.50	<0.001	a < b < c < d
BMI (kg/m2)	22.01 ± 2.89	23.31 ± 3.10	24.39 ± 3.14	25.46 ± 3.50	<0.001	a < b < c < d
WC (cm)	76.38 ± 8.52	80.50 ± 8.99	84.15 ± 8.84	87.51 ± 8.92	<0.001	a < b < c < d
VAI	0.64 ± 1.63	1.16 ± 0.16	1.87 ± 0.27	4.34 ± 2.64	<0.001	a < b < c < d
TC (mg/dL)	181.83 ± 31.08	188.73 ± 33.71	194.70 ± 35.31	201.50 ± 36.72	<0.001	a < b < c < d
TGs (mg/dL)	58.95 ± 16.39	91.98 ± 21.65	129.51 ± 32.43	244.92 ± 138.91	<0.001	a < b < c < d
HDL-C (mg/dL)	62.93 ± 12.11	54.54 ± 10.13	48.27 ± 9.29	40.81 ± 8.26	<0.001	d < c < b < a
BUN (mg/dL)	14.44 ± 5.79	14.33 ± 4.32	14.76 ± 4.60	14.40 ± 4.28	0.156
Crea (mg/dL)	0.83 ± 0.44	0.83 ± 0.25	0.85 ± 0.35	0.85 ± 0.35	0.138
hs-CRP (mg/L)	0.96 ± 2.16	1.09 ± 2.11	1.30 ± 2.45	1.48 ± 2.31	<0.001	a < d
FBG (mg/dL)	91.33 ± 8.56	94.17 ± 9.45	96.49 ± 9.79	97.96 ± 9.94	<0.001	a < b < c < d
Insulin (μU/mL)	5.35 ± 3.45	6.89 ± 4.81	8.21 ± 5.49	10.12 ± 7.19	<0.001	a < b < c < d
MetS	24 (2.2)	73 (6.7)	280 (25.6)	739 (67.5)	<0.001
MetS score	0.54 ± 0.76	1.02 ± 0.94	1.78 ± 1.07	3.00 ± 1.05	<0.001	a < b < c < d
HOMA-IR	1.23 ± 0.87	1.63 ± 1.25	1.99 ± 1.46	2.48 ± 1.84	<0.001	a < b < c < d
HOMA-B	70.61 ± 43.70	83.06 ± 55.57	91.85 ± 58.57	109.59 ± 86.69	<0.001	a < b < c < d

n (%), M ± SD Clinical characteristics according to the quartiles VAI in non-T2DM were analyzed using chi-square and an analysis of variance.

Table 3 Clinical characteristics of subjects according to the quartiles of VAI in T2DM (n = 542)

Variables	Visceral adiposity index (VAI)				p-value	Post hoc analysis
	a Quartile 1 (≤1.38) (n = 135)	b Quartile 2 (1.39–2.25) (n = 136)	c Quartile 3 (2.26–3.65) (n = 136)	d Quartile 4 (>3.66) (n = 135)
Men	90 (66.7)	74 (54.4)	59 (43.4)	65 (48.1)	0.001
Age (years)	64.87 ± 10.37	62.65 ± 11.87	63.11 ± 11.41	58.86 ± 13.17	<0.001	d < a
SBP (mmHg)	126.67 ± 17.65	125.88 ± 16.52	128.21 ± 18.17	126.19 ± 15.44	0.681
DBP (mmHg)	72.11 ± 9.65	74.24 ± 11.02	74.90 ± 9.94	76.62 ± 11.77	0.006	a < d
BMI (kg/m2)	23.99 ± 3.07	24.99 ± 3.65	26.25 ± 4.09	26.14 ± 3.64	<0.001	a < c, d
WC (cm)	84.92 ± 8.64	87.79 ± 9.65	90.19 ± 10.19	91.08 ± 9.41	<0.001	a < c, d
VAI	0.93 ± 0.27	1.81 ± 0.25	2.89 ± 0.40	6.70 ± 4.50	<0.001	a < b < c < d
TC (mg/dL)	173.93 ± 36.54	177.55 ± 44.66	179.27 ± 41.62	192.55 ± 42.79	0.002	a, b < d
TGs (mg/dL)	78.73 ± 26.37	122.57 ± 29.56	169.63 ± 40.92	344.88 ± 230.79	<0.001	a < b < c < d
HDL-C (mg/dL)	55.63 ± 11.45	46.47 ± 9.34	41.93 ± 7.90	36.95 ± 7.40	<0.001	d < c < b < a
BUN (mg/dL)	16.90 ± 5.22	16.26 ± 4.36	15.85 ± 4.91	15.70 ± 5.58	0.210
Crea (mg/dL)	0.87 ± 0.19	0.88 ± 0.22	0.87 ± 0.23	0.84 ± 0.21	0.471
hs-CRP (mg/L)	1.75 ± 3.37	1.74 ± 3.15	2.08 ± 3.07	1.83 ± 2.53	0.778
FBG (mg/dL)	143.56 ± 47.85	142.81 ± 39.01	142.20 ± 37.88	157.31 ± 45.44	0.009	a < d
Insulin (μU/mL)	9.66 ± 16.35	10.52 ± 10.11	10.77 ± 9.73	13.14 ± 8.81	0.085
MetS	28 (20.7)	76 (55.9)	122 (89.7)	127 (94.1)	<0.001
MetS score	1.83 ± 0.82	2.57 ± 0.89	3.37 ± 0.80	3.98 ± 0.87	<0.001	a < b < c < d
HOMA-IR	3.48 ± 6.16	3.67 ± 3.79	3.92 ± 4.26	5.09 ± 3.82	0.020	a < d
HOMA-B	52.76 ± 93.53	56.02 ± 52.56	56.03 ± 46.93	60.45 ± 44.60	0.794

n (%), M ± SD Clinical characteristics according to the quartiles VAI in T2DM were analyzed using chi-square and an analysis of variance.

Table 4 Multiple linear regression analysis for the independent factors determining HOMA-IR in non-T2DM and T2DM (n = 4,922)

Variables	HOMA-IR
	Non-T2DM (n = 4,380)					T2DM (n = 542)
	β	SE	t	95% CI	p-value	β	SE	t	95% CI	p-value
Age (years)	–0.143	0.002	–7.916	–0.016 – –0.010	<0.001	–0.114	0.021	–2.109	–0.086 – –0.003	0.035
Gender (Men)	0.003	0.047	0.166	–0.084 – 0.099	0.868	0.117	0.493	2.202	0.177 – 2.055	0.028
SBP (mmHg)	0.120	0.002	5.467	0.007 – 0.014	<0.001	–0.059	0.015	–1.086	–0.045 – 0.013	0.278
DBP (mmHg)	0.016	0.003	0.790	–0.003 – 0.008	0.430	0.130	0.026	2.194	0.006 – 0.106	0.029
TC (mg/dL)	0.016	0.001	1.066	–0.001 – 0.002	0.287	–0.027	0.005	–0.605	–0.013 – 0.007	0.546
BUN (mg/dL)	0.085	0.005	4.760	0.015 – 0.037	<0.001	0.027	0.043	0.569	–0.060 – 0.110	0.569
Crea (mg/dL)	–0.008	0.074	–0.472	–0.181 – 0.111	0.637	0.045	1.234	0.794	–1.443 – 3.403	0.427
hs-CRP (mg/L)	0.097	0.009	6.727	0.045 – 0.081	<0.001	0.071	0.064	1.690	–0.018 – 0.234	0.092
VAI	0.269	0.011	18.184	0.182 – 0.225	<0.001	0.110	0.065	2.477	0.033 – 0.290	0.014

Table 5 Multiple linear regression analysis for the independent factors determining HOMA-B in non-T2DM and T2DM (n = 4,922)

Variables	HOMA-B
	Non-T2DM (n = 4,380)					T2DM (n = 542)
	β	SE	t	95% CI	p-value	β	SE	t	95% CI	p-value
Age (years)	–0.294	0.073	–16.222	–1.321 – –1.036	<0.001	–0.032	0.290	–0.585	–0.739 – 0.400	0.559
Gender (Men)	–0.066	2.060	–4.170	–12.627 – –4.550	<0.001	0.155	6.774	2.866	6.111 – 32.727	0.004
SBP (mmHg)	0.100	0.084	4.554	0.217 – 0.545	<0.001	–0.068	0.204	–1.222	–0.652 – 0.152	0.222
DBP (mmHg)	–0.011	0.129	–0.573	–0.326 – 0.179	0.567	0.072	0.351	1.188	–0.273 – 1.108	0.235
TC (mg/dL)	0.006	0.028	0.405	–0.043 – 0.065	0.686	–0.075	0.068	–1.633	–0.245 – 0.023	0.103
BUN (mg/dL)	0.052	0.242	2.892	0.226 – 1.176	0.004	0.007	0.594	0.154	–1.075 – 1.259	0.878
Crea (mg/dL)	0.014	3.273	0.788	–3.837 – 8.995	0.431	0.142	16.944	2.446	8.157 – 74.728	0.015
hs-CRP (mg/L)	0.082	0.411	5.690	1.534 – 3.147	<0.001	–0.025	0.881	–0.591	–2.251 – 1.210	0.555
VAI	0.211	0.493	14.184	6.028 – 7.961	<0.001	0.041	0.896	0.901	–0.953 – 2.569	0.368

Table 6

Comparisons of HOMA-IR and HOMA-B according to the quartiles of VAI in non-T2DM and T2DM (n = 4,922)

Variables	Category	HOMA-IR
		Non-T2DM (n = 4,380)		T2DM (n = 542)
		Non-adjusted	Adjusted	Non-adjusted	Adjusted
VAI	Quartile 1	1.23 ± 0.04 (1.14–1.31)	1.20 ± 0.43 (1.12–1.28)	3.48 ± 0.40 (2.70–4.26)	3.56 ± 0.41 (2.76–4.36)
	Quartile 2	1.63 ± 0.04 (1.55–1.71)	1.64 ± 0.42 (1.56–1.72)	3.67 ± 0.40 (2.90–4.45)	3.70 ± 0.40 (0.93–4.48)
	Quartile 3	1.99 ± 0.04 (1.91–2.07)	2.00 ± 0.42 (1.92–2.09)	3.92 ± 0.40 (3.15–4.70)	3.86 ± 0.40 (3.08–4.65)
	Quartile 4	2.48 ± 0.04 (2.40–2.56)	2.49 ± 0.43 (2.41–2.57)	5.09 ± 0.40 (4.31–5.87)	5.04 ± 0.40 (4.26–5.83)
	p-value	<0.001	<0.001	0.020	0.038

Variables	Category	HOMA-B
		Non-T2DM (n = 4,380)		T2DM (n = 542)
		Non-adjusted	Adjusted	Non-adjusted	Adjusted
VAI	Quartile 1	70.61 ± 1.91 (66.87–74.35)	64.99 ± 1.89 (61.28–68.70)	52.76 ± 5.39 (42.17–63.35)	55.95 ± 5.57 (45.02–66.88)
	Quartile 2	83.06 ± 1.91 (79.32–86.80)	81.97 ± 1.84 (78.37–85.57)	56.02 ± 5.37 (45.47–66.57)	55.14 ± 5.35 (44.64–65.65)
	Quartile 3	91.85 ± 1.91 (88.11–95.59)	94.69 ± 1.84 (91.08–98.30)	56.03 ± 5.37 (45.48–66.58)	54.15 ± 5.41 (43.52–64.78)
	Quartile 4	109.59 ± 1.91 (105.85–113.33)	113.45 ± 1.88 (109.76–117.14)	60.45 ± 2.39 (49.87–71.04)	60.04 ± 5.48 (49.28–70.80)
	p-value	<0.001	<0.001	0.794	0.879

Adjusted for age, gender, SBP, DBP, TC, BUN, Crea, and hs-CRP

Results

Clinical characteristics of research subjects

The clinical characteristics of the research subjects are shown in Table 1. Mean VAI, HOMA-IR, and HOMA-B levels were 2.12 ± 2.14, 2.07 ± 2.19, and 85.20 ± 65.24, respectively. Age (p < 0.001), BMI (p < 0.001), WC (p < 0.001), TGs (p < 0.001), hs-CRP (p < 0.001), MetS score (p < 0.001), VAI (p < 0.001), and HOMA-IR (p < 0.001) in the T2DM group were significantly higher than in the non-T2DM group. TC (p < 0.001), HDL-C (p < 0.001), and HOMA-B (p < 0.001) in the T2DM group were significantly lower than in the non-T2DM group.

Clinical characteristics of subjects according to the quartiles of VAI in non-T2DM

Clinical characteristics by quartile of VAI in non-T2DM are shown in Table 2. Age (p < 0.001), SBP (p < 0.001), DBP (p < 0.001), BMI (p < 0.001), WC (p < 0.001), TC (p < 0.001), TGs (p < 0.001), hs-CRP (p < 0.001), FBG (p < 0.001), insulin (p < 0.001), MetS score (p < 0.001), HOMA-IR (p < 0.001), and HOMA-B (p < 0.001) were positively associated with increasing quartiles of VAI. HDL-C (p < 0.001) was inversely associated with increasing quartiles of VAI.

Clinical characteristics of subjects according to the quartiles of VAI in T2DM

Clinical characteristics by quartile of VAI in T2DM are shown in Table 3. Age (p = 0.001), DBP (p = 0.006), BMI (p < 0.001), WC (p < 0.001), TC (p = 0.002), TGs (p < 0.001), FBG (p = 0.009), MetS score (p < 0.001), and HOMA-IR (p = 0.020) were positively associated with increasing quartiles of VAI. HDL-C (p < 0.001) was inversely associated with increasing quartiles of VAI. Insulin (p = 0.085) and HOMA-B (p = 0.794) showed no significant association.

Multiple linear regression analysis for the independent factors determining HOMA-IR and HOMA-B in non-T2DM and T2DM

Multiple linear regression analysis for the independent factors determining HOMA-IR and HOMA-B in non-T2DM and T2DM are shown in Tables 4 and 5. VAI in non-T2DM (β = 0.269, 95% confidence interval [CI] 0.182 to 0.225; p < 0.001) and T2DM (β = 0.110, 95% CI 0.033 to 0.290; p = 0.014) were significant independent factors determining HOMA-IR (Table 4). VAI in non-T2DM (β = 0.211, 95% CI 6.028 to 7.961; p < 0.001) were significant independent factor determining HOMA-B but not in T2DM (β = 0.041, 95% CI –0.953 to 2.569; p = 0.368) (Table 5).

Comparisons of HOMA-IR and HOMA-B according to the quartiles of VAI in non-T2DM and T2DM

Comparisons of HOMA-IR and HOMA-B with the quartiles of VAI in non-T2DM and T2DM are shown in Table 6. In non-T2DM, after adjusting for related variables, HOMA-IR (p < 0.001) and HOMA-B (p < 0.001) were positively associated with quartiles of VAI. In T2DM, after adjusting for related variables, HOMA-IR (p = 0.038) was positively associated with quartiles of VAI, but HOMA-B (p = 0.879) showed no significant correlation.

Discussion

Using data from the KNHANES VI-3, this study investigated the association between VAI and HOMA-IR and HOMA-B in Korean adults with and without T2DM. VAI was positively associated with HOMA-IR in both non-T2DM and T2DM. VAI was positively associated with HOMA-B in non-T2DM, but not in T2DM.

The accumulation of fatty tissue is associated with excessive fatty acid secretion, an increase of inflammatory cytokines, and abnormal adipocyte hormone signaling, resulting in insulin resistance [21, 22]. In the present study, HOMA-IR and MetS scores were found to increase with increasing VAI in both non-T2DM and T2DM subjects. In previous studies on the relationship between VAI and HOMA-IR, Randrianarisoa et al. reported that VAI is positively associated with HOMA-IR in non-diabetic German (β = 0.42, p < 0.0001) [23]. Ji et al. found that a very high VAI is the main risk factor for increased HOMA-IR in Chinese adults [24]. In the Framingham Heart Study, Preis et al. reported that visceral adipose tissue and abdominal subcutaneous adipose tissue are positive correlates with insulin resistance and that visceral adipose tissue is a stronger correlate with insulin resistance than abdominal subcutaneous adipose tissue [25]. Visceral adipocytes secrete adipose-specific cytokines, such as leptin and adiponectin, and inflammatory cytokines (tumor necrosis factor-α and interleukin-6), which increase insulin resistance [26, 27]. Macrophages accumulate in visceral adipose tissue and release these inflammatory cytokines, including tumor necrosis factor-α and interleukin-6, which can impair insulin sensitivity [28]. Excessive adipose tissue can promote inflammation by increasing levels of resistin or tumor necrosis factor-α, which increase insulin resistance [29, 30]. Diminished adiponectin levels linked to excessive adipose tissue can exacerbate metabolic dysregulation and insulin resistance [31].

In the present study, we evaluated the relationship between VAI and beta-cell function in non-T2DM and T2DM subjects. HOMA-B (p < 0.001), insulin (p < 0.001), and MetS scores (p < 0.001) increased with increasing VAI in non-T2DM adults (Table 6 and Supplementary Table 1). In non-T2DM subjects, insulin secretion may be increased to compensate for the increase in insulin resistance and VAI because the mass and function of beta cells are normal. In our previous study conducted for non-diabetic adults, HOMA-IR and HOMA-B for MetS were higher than for non-MetS [32]. In addition, HOMA-IR and HOMA-B increased with an increase in the MetS score. Polonsky et al. reported that insulin levels increase two to three-fold to compensate for insulin resistance associated with obesity [33]. Linnemann et al. suggested that an expansion of beta-cell mass in non-diabetic obese adults can provide enough insulin and prevent DM [34]. Also, Gastaldelli et al. reported that visceral fat deposition in non-diabetic populations increases insulin resistance but is also positively associated with an increase in beta-cell function [35]. They suggested that insulin secretion increases to compensate for the insulin resistance and that the dynamics of beta-cell function are largely preserved.

However, high demands on beta cells can affect their ability to compensate for insulin resistance. Ferrannini et al. reported that insulin levels in obese populations increase to maintain glucose tolerance, leading to stress on beta-cell function [36]. Persistent insulin hypersecretion causes beta-cell exhaustion and dysfunction [37, 38]. A hyper response by beta cells in non-T2DM adults can lead to the onset of diabetes mellitus. Macor et al. reported that excessive accumulation of visceral adipose tissue in subjects with normal glucose tolerance causes hyper insulin secretion and worsens insulin sensitivity [39]. Song et al. reported that compared to populations with low HOMA-IR and high HOMA-B, subjects with high HOMA-IR and high HOMA-B have a higher relative risk (RR) of diabetes mellitus (RR, 5.34; 95% CI, 2.96–9.62) [40].

Beta-cell mass and function in T2DM adults decrease due to apoptosis and necrosis of pancreatic beta cells [41]. In our study, HOMA-IR (p = 0.038) and MetS scores (p < 0.001) tended to increase with increasing VAI in T2DM subjects, but not with HOMA-B (p = 0.879) or insulin (p = 0.104) (Table 6 and Supplementary Table 1). We consider these results to indicate that even with insulin resistance and increased VAI in subjects with T2DM, insulin is secreted due to compensatory action. However, this secretion is insufficient because beta-cell mass and function are significantly reduced. There is also previous study that is the opposite of our findings. Yang et al. reported that VAI is negatively associated with HOMA-B in subjects with T2DM [42]. In their study DM was diagnosed using the standard oral glucose tolerance test (GTT) and oral GTT is a very accurate method in the diagnosis of diabetes. In their study, BMI was used as a correction variable. However, we think that the correction of BMI is a double correction because BMI is a variable included in the calculation of VAI. Some studies have emphasized that physicians should aim to maintain normal beta-cell mass and function for prevention and treatment of diabetes; beta-cell failure is central to the development and progression of DM [43, 44]. Thus, for the prevention and treatment of diabetes mellitus, physicians should consider taking steps both to decrease insulin resistance and maintain normal beta-cell mass and function.

In conclusions, we investigated the association of the visceral adiposity index with insulin resistance and beta-cell function in Korean adults with and without T2DM. The visceral adiposity index was positively associated with insulin resistance and beta-cell function in non-T2DM subjects. The visceral adiposity index was positively associated with insulin resistance in Korean adults with T2DM but was not associated with beta-cell function.

The present investigation has several limitations. First, the background of the diabetic patients (duration of diabetes, diabetic complications, family history and the treatment drugs) can affect the relationship between VAI and HOMA. However, KNHANES data do not have accurate information about the background (duration of diabetes, diabetic complications, family history and the treatment drugs) of the diabetic patients. Second, the study assessed insulin resistance and beta-cell function using HOMA indices. In the measurements of insulin resistance and beta-cell function, the “gold-standard” methods (e.g., hyperinsulinemic-euglycemic clamp and the hyperglycemic clamp tests) are more accurate than HOMA indices. Although HOMA indices are not the “gold-standard” method, they may be more appropriate for use in large epidemiological studies [45]. Third, as a cross-sectional study, the ability to establish a causal relationship between VAI and beta-cell function is limited. Nevertheless, this study is the first to show a relationship between VAI and HOMA-B in Korean adults with and without T2DM. More accurate and generalizable results might be obtained by performing a cohort study.

Acknowledgement

This study was financially supported by Dongnam Health University.

Disclosure

The authors declare that there is no conflict of interest associated with this manuscript.

References

• 1   Olokoba  AB,  Obateru  OA,  Olokoba  LB (2012) Type 2 diabetes mellitus: a review of current trends. Oman Med J 27: 269–273.
• 2   Leon  BM,  Maddox  TM (2015) Diabetes and cardiovascular disease: epidemiology, biological mechanisms, treatment recommendations and future research. World J Diabetes 6: 1246–1258.
• 3  Taylor R (2012) Insulin resistance and type 2 diabetes. Diabetes 61: 778–779.
• 4   Chen  C,  Cohrs  CM,  Stertmann  J,  Bozsak  R,  Speier  S (2017) Human beta cell mass and function in diabetes: recent advances in knowledge and technologies to understand disease pathogenesis. Mol Metab 6: 943–957.
• 5   Cernea  S,  Dobreanu  M (2013) Diabetes and beta cell function: from mechanisms to evaluation and clinical implications. Biochem Med (Zagreb) 23: 266–280.
• 6   Page  KA,  Reisman  T (2013) Interventions to preserve beta-cell function in the management and prevention of type 2 diabetes. Curr Diab Rep 13: 252–260.
• 7   Wallace  TM,  Levy  JC, Matthews DR (2014) Use and abuse of HOMA modeling. Diabetes Care 27: 1487–1495.
• 8   Yin  J,  Gao  Z,  He  Q,  Zhou  D,  Guo  Z, et al. (2009) Role of hypoxia in obesity-induced disorders of glucose and lipid metabolism in adipose tissue. Am J Physiol Endocrinol Metab 296: E333–E342.
• 9  Després JP (1998) The insulin resistance-dyslipidemic syndrome of visceral obesity: effect on patients’ risk. Obes Res 6: S8–S17.
• 10   Amato  MC,  Giordano  C,  Galia  M,  Criscimanna  A,  Vitabile  S, et al. (2010) Visceral adiposity index: a reliable indicator of visceral fat function associated with cardiometabolic risk. Diabetes Care 33: 920–922.
• 11   Amato  MC,  Giordano  C (2014) Visceral adiposity index: an indicator of adipose tissue dysfunction. Int J Endocrinol 2014: 730827.
• 12   Dereziński  T,  Zozulińska-Ziółkiewicz  D,  Uruska  A,  Dąbrowski  M (2018) Visceral adiposity index as a useful tool for the assessment of cardiometabolic disease risk in women aged 65 to 74. Diabetes Metab Res Rev 34: e3052.
• 13   Wander  PL,  Boyko  EJ,  Leonetti  DL,  McNeely  MJ,  Kahn  SE, et al. (2013) Change in visceral adiposity independently predicts a greater risk of developing type 2 diabetes over 10 years in Japanese Americans. Diabetes Care 36: 289–293.
• 14   Bozorgmanesh  M,  Hadaegh  F,  Azizi  F (2011) Predictive performance of the visceral adiposity index for a visceral adiposity-related risk: type 2 diabetes. Lipids Health Dis 10: 88.
• 15   Uruska  A,  Zozulinska-Ziolkiewicz  D,  Niedzwiecki  P,  Pietrzak  M,  Wierusz-Wysocka  B (2018) TG/HDL-C ratio and visceral adiposity index may be useful in assessment of insulin resistance in adults with type 1 diabetes in clinical practice. J Clin Lipidol 12: 734–740.
• 16   Cardona-Alvarado  MI,  Lopez-Moreno  G,  Aguilar-Zavala  H,  Figueroa-Vega  N,  Perez-Luque  E (2018) Relationship of visceral adiposity index with the metabolic phenotype and cardiovascular markers in non-diabetic subjects. Health Sci J 12: 588.
• 17   Expert Panel on  Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (2001) Executive summary of the third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel III). JAMA 285: 2486–2497.
• 18   Yoon  H,  Kim  GS,  Kim  SG,  Moon  AE (2015) The relationship between metabolic syndrome and increase of metabolic syndrome score and serum vitamin D levels in Korean adults: 2012 Korean national health and nutrition examination survey. J Clin Biochem Nutr 57: 82–87.
• 19   Matthews  DR,  Hosker  JP,  Rudenski  AS,  Naylor  BA,  Treacher  DF, et al. (1985) Homeostasis model assessment: insulin resistance and beta cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28: 412–419.
• 20   Harris  MI,  Flegal  KM,  Cowie  CC,  Eberhardt  MS,  Goldstein  DE, et al. (1998) Prevalence of diabetes, impaired fasting glucose, and impaired glucose tolerance in U.S. adults. The Third National Health and Nutrition Examination Survey, 1988–1994. Diabetes Care 21: 518–524.
• 21   Guilherme  A,  Virbasius  JV,  Puri  V,  Czech  MP (2008) Adipocyte dysfunctions linking obesity to insulin resistance and type 2 diabetes. Nat Rev Mol Cell Biol 9: 367–377.
• 22  Schuster DP (2010) Obesity and the development of type 2 diabetes: the effects of fatty tissue inflammation. Diabetes Metab Syndr Obes 3: 253–262.
• 23   Randrianarisoa  E,  Lehn-Stefan  A,  Hieronimus  A,  Rietig  R,  Fritsche  A, et al. (2019) Visceral adiposity index as an independent marker of subclinical atherosclerosis in individuals prone to diabetes mellitus. J Atheroscler Thromb 26: 821–834.
• 24   Ji  B,  Qu  H,  Wang  H,  Wei  H,  Deng  H (2017) Association between the visceral adiposity index and homeostatic model assessment of insulin resistance in participants with normal waist circumference. Angiology 68: 716–721.
• 25   Preis  SR,  Massaro  JM,  Robins  SJ,  Hoffmann  U,  Vasan  RS, et al. (2010) Abdominal subcutaneous and visceral adipose tissue and insulin resistance in the Framingham heart study. Obesity (Silver Spring) 18: 2191–2198.
• 26   Rytka  JM,  Wueest  S,  Schoenle  EJ,  Konrad  D (2011) The portal theory supported by venous drainage-selective fat transplantation. Diabetes 60: 56–63.
• 27   Makki  K,  Froguel  P,  Wolowczuk  I (2013) Adipose tissue in obesity-related inflammation and insulin resistance: cells, cytokines, and chemokines. ISRN Inflamm 2013: 139239.
• 28   Hardy  OT,  Czech  MP,  Corvera  S (2012) What causes the insulin resistance underlying obesity? Curr Opin Endocrinol Diabetes Obes 19: 81–87.
• 29   Steppan  CM,  Bailey  ST,  Bhat  S,  Brown  EJ,  Banerjee  RR, et al. (2001) The hormone resistin links obesity to diabetes. Nature 409: 307–312.
• 30   Hivert  MF,  Sullivan  LM,  Fox  CS,  Nathan  DM,  D’Agostino RB  Sr, et al. (2008) Associations of adiponectin, resistin, and tumor necrosis factor-alpha with insulin resistance. J Clin Endocrinol Metab 93: 3165–3172.
• 31   Coker  RH,  Williams  RH,  Yeo  SE,  Kortebein  PM,  Bodenner  DL, et al. (2009) Visceral fat and adiponectin: associations with insulin resistance are tissue-specific in women. Metab Syndr Relat Disord 7: 61–67.
• 32   Yoon  H,  Yoon  YS,  Kim  SG,  Oh  HJ,  Choi  CW, et al. (2019) Relationship between metabolic syndrome and metabolic syndrome score with β-cell function by gender in non-diabetic Korean populations. Endocr Res 44: 71–80.
• 33   Polonsky  KS,  Given  BD,  Van Cauter  E (1988) Twenty-four-hour profiles and pulsatile patterns of insulin secretion in normal and obese subjects. J Clin Invest 81: 442–448.
• 34   Linnemann  AK,  Baan  M,  Davis  DB (2014) Pancreatic β-cell proliferation in obesity. Adv Nutr 5: 278–288.
• 35   Gastaldelli  A,  Sironi  AM,  Ciociaro  D,  Positano  V,  Buzzigoli  E, et al. (2005) Visceral fat and beta cell function in non-diabetic humans. Diabetologia 48: 2090–2096.
• 36   Ferrannini  E,  Camastra  S (1998) Relationship between impaired glucose tolerance, non-insulin-dependent diabetes mellitus and obesity. Eur J Clin Invest 28 Suppl 2: 3–6; discussion: 6–7.
• 37  Kasuga M (2006) Insulin resistance and pancreatic beta cell failure. J Clin Invest 116: 1756–1760.
• 38   Cernea  S,  Dobreanu  M (2013) Diabetes and beta cell function: from mechanisms to evaluation and clinical implications. Biochem Med (Zagreb) 23: 266–280.
• 39   Macor  C,  Ruggeri  A,  Mazzonetto  P,  Federspil  G,  Cobelli  C, et al. (1997) Visceral adipose tissue impairs insulin secretion and insulin sensitivity but not energy expenditure in obesity. Metabolism 46: 123–129.
• 40   Song  Y,  Manson  JE,  Tinker  L,  Howard  BV,  Kuller  LH, et al. (2007) Insulin sensitivity and insulin secretion determined by homeostasis model assessment and risk of diabetes in a multiethnic cohort of women: the women’s health initiative observational study. Diabetes Care 30: 1747–1752.
• 41  Tomita T (2016) Apoptosis in pancreatic β-islet cells in type 2 diabetes. Bosn J Basic Med Sci 16: 162–179.
• 42   Yang  Y,  Feng  Y,  Ma  X,  Chen  K,  Wu  N, et al. (2015) Visceral adiposity index and insulin secretion and action in first-degree relatives of subjects with type 2 diabetes. Diabetes Metab Res Rev 31: 315–321.
• 43  Saisho Y (2014) Importance of beta cell function for the treatment of type 2 diabetes. J Clin Med 3: 923–943.
• 44   Halban  PA,  Polonsky  KS,  Bowden  DW,  Hawkins  MA,  Ling  C, et al. (2014) β-cell failure in type 2 diabetes: postulated mechanisms and prospects for prevention and treatment. Diabetes Care 37: 1751–1758.
• 45   Wallace  TM,  Levy  JC,  Matthews  DR (2004) Use and abuse of HOMA modeling. Diabetes Care 27: 1487–1495.