Association Between Waist-to-Height Ratio and Endothelial Dysfunction in Patients With Morbidity　― A Report From the FMD-J Study ―

Abstract

Background: Waist circumference (WC), waist-to-height ratio (WHtR) and body mass index (BMI) are known as easy anthropometric markers of abnormal obesity and screening tools for predicting cardiovascular outcomes, but which indices are best is unclear. We therefore investigated the superiority and association between each index and low flow-mediated dilatation (FMD) as a surrogate marker for cardiovascular outcomes in patients with morbidity in a large Japanese prospective cohort.

Methods and Results: A total of 1,645 Japanese patients who had coronary artery disease and hypertension or diabetes mellitus were enrolled, and 1,087 of them were analyzed. The high-WHtR group (≥0.5) showed greater morbidity and increased inflammation in association with atherosclerosis compared with the low-WHtR group. High WHtR and advanced age were identified as predictors of low FMD (odds ratio (OR) 1.39, 95% confidence interval (CI) 1.02–1.88, P=0.037 and OR 1.55, 95% CI 1.19–2.01, P=0.001, respectively). However, WC was not associated with that risk in either sex (male: OR 1.37, 95% CI 0.97–1.93, P=0.076; female: OR 1.08, 95% CI 0.68–1.73, P=0.74), and no association was evident between high BMI and low FMD (OR 0.92, 95% CI 0.71–1.19, P=0.54).

Conclusions: WHtR offers a superior predictor of decreased FMD than other anthropometric indices, and progression of arteriosclerosis might be detected more sensitively. Further study is needed to investigate the relationship between cardiovascular mortality and WHtR.

Obesity, and abdominal adiposity in particular, is related to increased risks of mortality and morbidity.^1,2 Similarly, recent studies and a systematic review have reported that metabolic syndrome (MetS) causes endothelial dysfunction, oxidative stress, inflammation and is associated with an increased risk of cardiovascular events.^3,4 Defining optimal cutoffs for waist circumference (WC), which reflects abnormal adiposity and is one of the diagnostic criteria for MetS,⁵ appears difficult because of the influences of various factors, such as age, sex, ethnic group, height, and weight.^6,7 In fact, the current Japanese cutoffs of WC for MetS (85 cm for men; 90 cm for women) remain controversial.^8,9 Simple anthropometric indices used as markers of abnormal obesity and in screening tools for predicting cardiovascular outcomes include WC, waist-to-height ratio (WHtR) and body mass index (BMI), but which of these works best remains unclear. Atherosclerosis, which leads to cardiovascular disease (CVD) events, is caused by endothelial dysfunction, and because the measurement of flow-mediated dilatation (FMD) is able to clarify endothelial function noninvasively, early detection and treatment of atherosclerosis may be feasible.^10,11 However, little is known about the comparison of FMD among the anthropometric indices. We therefore investigated the superiority and associations between anthropometric indices and low FMD of the brachial artery as a surrogate marker for endothelial dysfunction (and hence, increased risk of cardiovascular morbidity) in patients with a diagnosis of coronary artery disease (CAD), hypertension or diabetes mellitus (DM) in the Flow-mediated Dilation Japan Registry (FMD-J) study, which recruited a large prospective cohort.

Methods

Study Design and Subjects

The FMD-J study was a prospective multicenter study conducted to examine the utility of FMD assessment using a semi-automatic device at individual institutions in the management of patients at risk of CVD.¹² Groups A and B in that study comprised patients being followed at one of the participating centers (22 university hospitals and affiliated clinics). The ethics committees of the participating institutions approved the study protocol. We obtained written informed consent for participation from all subjects in the FMD-J study.

The study design and patient enrollment of the FMD-J have been described in detail.^12,13 A total of 1,645 Japanese patients who had CAD (n=679 from Study A) or hypertension or DM (n=966 from Study B) who gave informed consent for their vascular function to be measured were enrolled in the FMD-J Registry between 1 April 2010 and 31 August 2012 at 3 hospitals in Japan. We excluded subjects for whom measurements of WC were unavailable. The final study population was 1,087 patients for this analysis (Figure).

Figure.

Study flow chart. We excluded patients for whom waist circumference was not measured to obtain final study population.

The study protocols of the multicenter study groups A and B conformed to the principles of the Declaration of Helsinki, and were approved by the ethics committees of each of the participating institutions.¹²

Definitions

The meta-analysis showed that the area under receiver operator characteristic analyses indicated that mean boundary values for WHtR, covering all cardiometabolic outcomes, from studies in 14 different countries and including Caucasian, Asian and Central American subjects, were 0.5 for men and 0.5 for women.¹⁴ For this reason, we adopted 0.5 as the cutoff point of WHtR.

Hypertension was defined as systolic blood pressure (SBP) ≥140 mmHg or diastolic blood pressure ≥90 mmHg in a seated position on at least 3 different occasions. We excluded patients with secondary hypertension. Diabetes was defined according to the criteria of the American Diabetes Association.¹⁵ CAD included angina pectoris, unstable angina and myocardial infarction. Unstable angina was defined as present if a history of prolonged ischemic chest pain (>15 min in duration) was observed with transient ischemic ST-segment and T-wave abnormality on ECGs without development of Q-wave abnormality or elevation of serum enzyme with myocardial necrosis. Framingham Risk score was computed using the following risk factors: age, total cholesterol level, high-density lipoprotein cholesterol (HDL-C) level, SBP and smoking status.¹⁶ Definitions of other diseases were as previously reported.¹²

Measurement of FMD

All measurements were taken in the morning, after overnight fasting, in a quiet, dark, air-conditioned room (constant temperature, 22–25℃). Subjects remained supine throughout all FMD measurements. Vascular response to reactive hyperemia in the brachial artery was used for assessment of endothelium-dependent FMD. Details of measurements are provided in our previous report.¹²

Measurement of Small-Dense Low-Density Lipoprotein (LDL) and Oxidized LDL

The small-dense LDL levels were measured with a rapid assay using a modified version of a previously described heparin-magnesium precipitation method by Hirano et al.¹⁷ Serum levels of malondialdehyde-modified LDL (MDA-LDL), a major oxidized LDL, was measured at SRL Co. Ltd., Tokyo. An enzyme-linked immunosorbent assay was used for the detection of MDA-LDL, based on the principles previously reported by Kotani et al.¹⁸

Statistical Analysis

Descriptive statistics are presented as frequency (percentage) for categorical variables and mean±standard deviation or median (interquartile range (IQR)) for continuous variables. These values were compared between patients who had low-WHtR (<0.5) and high-WHtR (≥0.5) using a chi-square or Fisher’s exact test for categorical variables and using a two-sample t-test or Wilcoxon rank-sum test for continuous variables. Multiple logistic regression modeling was used to identify the effect of WHtR on low FMD (<4.8%), with effects expressed as odds ratios (ORs) and 95% confidence intervals (CIs).

All analyses were conducted using JMP version 12.0 software (SAS Institute, Cary, NC, USA). All statistical analyses were two-tailed and values of P<0.05 were considered statistically significant.

Results

Baseline Characteristics and Medications: WHtR <0.5 vs. ≥0.5 (Table 1)

As compared with patients in the low-WHtR (<0.5) group, age, BMI, WC, triglycerides, HDL-C, glucose, uric acid, and Framingham score were significantly higher in the high-WHtR (≥0.5) group. Furthermore, hypertension, dyslipidemia, DM, chronic kidney disease and cardio- and cerebrovascular disease were more frequent in the high-WHtR group. Regarding baseline medications, use of calcium antagonists, angiotensin converting enzyme inhibitors (ACEI) or angiotensin II receptor blockers (ARB), β-blockers, diuretics, statins, oral antidiabetic agents and insulin was more frequent in the high-WHtR group than in the low-WHtR group.

Table 1. Baseline Characteristics of Study Patients Subdivided by WHtR <0.5 vs. ≥0.5

Variables	All (n=1,087)	WHtR <0.5 (n=234)	WHtR ≥0.5 (n=853)	P value
Age (years)	61.9±9.3	60.8±9.95	62.2±9.10	0.054
≥65 years	462 (43)	87 (37)	375 (44)	0.063
Male (n (%))	715 (66)	153 (65)	562 (66)	0.89
BMI (kg/m2)	24.9±3.7	21.7±2.82	25.8±3.41	<0.0001
Waist (cm)	88.0±10.3	76.0±6.6	91.3±8.5	<0.0001
Height (cm)	161.8±8.7	163.0±8.9	161.5±8.6	0.018
WHtR	0.54±0.06	0.47±0.03	0.57±0.05	<0.0001
Systolic BP (mmHg)	133.7±15.3	134.1±15.3	133.6±15.3	0.64
Diastolic BP (mmHg)	78.8±10.9	78.8±10.9	78.8±10.9	0.75
Heart rate (beats/min)	71.3±12.4	71.5±13.9	71.2±12.0	0.81
Total cholesterol (mmol/L)	4.93±0.87	5.02±0.86	4.91±0.86	0.091
Triglycerides (mmol/L)	1.23 (0.87–1.75)	0.95 (0.73–1.36)	1.32 (0.94–1.84)	<0.0001
HDL-C (mmol/L)	1.44±0.39	1.67±0.46	1.38±0.35	<0.0001
LDL-C (mmol/L)	3.20±0.79	3.12±0.78	3.22±0.79	0.071
Glucose (mmol/L)	5.61 (5.22–6.23)	5.42 (5.17–5.83)	5.67 (5.28–6.39)	<0.0001
Uric acid (mg/dL)	5.81±1.35	5.48±1.35	5.90±1.33	<0.0001
eGFR (mL/min/1.73 m2)	75.0±18.3	77.7±18.8	74.2±18.1	0.009
Framingham risk score	6.04±3.00	5.42±0.20	6.21±2.92	0.0005
Hypertension (n (%))	1,037 (95)	209 (89)	828 (97)	<0.0001
Dyslipidemia (n (%))	535 (49)	85 (36)	450 (53)	<0.0001
Diabetes mellitus (n (%))	169 (16)	17 (7.3)	152 (18)	<0.0001
Cardio- and cerebrovascular disease (n (%))	366 (34)	63 (27)	303 (36)	0.014
Current smoking (n (%))	150 (14)	41 (18)	109 (13)	0.061
Chronic kidney disease (n (%))	190 (17)	28 (12)	162 (19)	0.012
Flow-mediated vasodilation (%)	4.48±2.59	4.86±2.74	4.38±2.54	0.011
Baseline brachial artery diameter (mm)	4.16±0.63	4.02±0.58	4.20±0.65	0.0002
Mean ankle-brachial index	1.14±0.08	1.14±0.07	1.15±0.09	0.41
hs-CRP (mg/mL)	0.52 (0.26–1.08)	0.36 (0.16–0.74)	0.57 (0.29–1.15)	<0.0001
Small dense LDL (mg/dL)	36.6±15.6	31.9±12.8	37.9±16.1	<0.0001
Oxidized LDL (U/L)	119.7±39.7	111.9±35.6	121.8±40.5	0.0003
Medications
Calcium antagonists (n (%))	638 (59)	112 (48)	526 (62)	0.0001
ACEI/ARB (n (%))	775 (71)	149 (64)	626 (73)	0.004
β-blockers (n (%))	274 (25)	45 (19)	229 (27)	0.018
Diuretics (n (%))	271 (25)	41 (18)	230 (27)	0.003
Aldosterone antagonists (n (%))	37 (3.4)	6 (2.6)	31 (3.6)	0.42
Statins	486 (45)	77 (33)	409 (48)	<0.0001
Fibrate preparations	29 (2.7)	4 (1.7)	25 (2.9)	0.30
EPA preparations	32 (2.9)	7 (3.0)	25 (2.9)	0.96
Oral antidiabetic agents	157 (14)	17 (7.3)	140 (16)	0.0004
Use of insulin	23 (2.1)	1 (0.4)	22 (2.6)	0.041
Aspirin	401 (37)	79 (34)	322 (38)	0.26
Ticlopidine	54 (5.0)	13 (5.6)	41 (4.8)	0.64
Clopidogrel	121 (11)	22 (9.4)	99 (12)	0.34
Nitrate agents	63 (5.8)	9 (3.9)	54 (6.3)	0.15
Nicorandil	83 (7.6)	15 (6.4)	68 (8.0)	0.43

Continuous variables are shown as mean±SD or median (interquartile range). Categorical variables were compared with the chi-square test or Fisher’s exact test. Continuous variables were compared using Student’s t-test or the Wilcoxon rank-sum test based on distributions. ACEI, angiotensin converting enzyme-inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; BP, blood pressure; eGFR, estimated glomerular filtration rate; EPA, eicosapentaenoic acid; HDL-C, high-density lipoprotein cholesterol; hs-CRP, high-sensitivity C-reactive protein; LDL, low-density lipoprotein; LDL-C, LDL cholesterol; WHtR, waist-to-height ratio.

Factors related to inflammation and oxidative stress in association with atherosclerosis such as high-sensitivity C-reactive protein (hs-CRP), small-dense LDL and oxidized LDL were significantly higher in the high-WHtR group.

High WHtR as a Risk Factor for Low FMD (Table 2)

The risk factors for low FMD (<4.8%) in univariate analysis by logistic regression model were high WHtR and advanced age. After adjusting for confounders such as male sex, uric acid (≥7.0 mg/dL), hypertension, dyslipidemia, DM, current smoking, chronic kidney disease and cardio- and cerebrovascular disease, high WHtR and advanced age were identified as predictors (OR 1.39, 95% CI 1.02–1.88, P=0.037; OR 1.55, 95% CI 1.19–2.01, P=0.001, respectively).

Table 2. Adjusted Effect of WHtR for FMD <4.8%

Variables	N of events/present N of patients	N of events/absent N of patients	Univariate		Multivariate
			OR (95% CI)	P value	OR (95% CI)	P value
WHtR ≥0.5	515/853 (60%)	122/234 (52%)	1.40 (1.05–1.87)	0.024	1.39 (1.02–1.88)	0.037
Age ≥65 years	299/462 (65%)	338/625 (54%)	1.56 (1.22–2.00)	0.0004	1.55 (1.19–2.01)	0.001
Male	428/715 (60%)	209/372 (56%)	1.16 (0.90–1.50)	0.24	1.30 (0.98–1.73)	0.071
Uric acid ≥7.0 mg/dL	119/205 (58%)	518/880 (59%)	0.97 (0.71–1.32)	0.83	0.85 (0.61–1.20)	0.36
Hypertension	611/1,037 (59%)	26/50 (52%)	1.32 (0.75–2.34)	0.33	1.22 (0.67–2.19)	0.51
Dyslipidemia	308/535 (58%)	329/552 (60%)	0.92 (0.72–1.17)	0.5	0.85 (0.63–1.16)	0.31
Diabetes mellitus	101/169 (60%)	536/918 (58%)	1.06 (0.76–1.48)	0.73	1.02 (0.71–1.46)	0.93
Current smoking	90/150 (60%)	539/920 (59%)	1.06 (0.75–1.51)	0.74	1.14 (0.79–1.64)	0.49
Chronic kidney disease	122/190 (64%)	515/897 (57%)	1.33 (0.96–1.84)	0.082	1.26 (0.90–1.80)	0.18
Cardio- and Cerebrovascular disease	213/366 (58%)	424/721 (59%)	0.98 (0.76–1.26)	0.85	0.90 (0.64–1.26)	0.53

A multiple logistic regression model was used to identify the effect of WHtR on low FMD (<4.8%), with effects expressed as ORs and 95% CIs. CIs, confidence intervals; FMD, flow-mediated dilatation; ORs, odds ratios; WHtR, waist-to-height ratio.

Association Between WC, BMI and FMD

In relation to WC, analysis stratified by sex was performed because definitions of MetS differed between male and female patients.

In male patients, risk factors for low FMD (<4.8%) in the univariate analysis by logistic regression modeling did not include high WC (≥85 cm), only advanced age. After adjusting for confounders such as uric acid (≥7.0 mg/dL), hypertension, dyslipidemia, DM, current smoking, chronic kidney disease and cardio- and cerebrovascular disease, high WC was not confirmed as a predictor (OR 1.37, 95% CI 0.97–1.93, P=0.076), and only advanced age was found to be a predictor (OR 1.59, 95% CI 1.15–2.21, P=0.005; Table 3). Similarly, among female patients, risk factors for low FMD (<4.8%) in the univariate analysis by logistic regression modeling again did not include high WC (≥90 cm), only advanced age. After adjusting for confounders such as uric acid (≥7.0 mg/dL), hypertension, dyslipidemia, DM, current smoking, chronic kidney disease and cardio- and cerebrovascular disease, high WC did not represent a predictor (OR 1.08, 95% CI 0.68–1.73, P=0.74), with only advanced age showing a predictive relationship (OR 1.56, 95% CI 1.00–2.43, P=0.0049) (Table 4).

Table 3. Adjusted Effect of WC ≥85 cm for FMD <4.8% in Males

Variables	N of events/present N of patients	N of events/absent N of patients	Univariate		Multivariate
			OR (95% CI)	P value	OR (95% CI)	P value
WC ≥85 cm	316/512 (62%)	112/203 (55%)	1.31 (0.94–1.82)	0.11	1.37 (0.97–1.93)	0.076
Age ≥65 years	186/280 (66%)	242/435 (56%)	1.58 (0.46–0.86)	0.004	1.59 (1.15–2.21)	0.005
Uric acid ≥7.0 mg/dL	108/186 (58%)	320/528 (61%)	0.90 (0.64–1.27)	0.54	0.83 (0.58–1.20)	0.32
Hypertension	412/683 (60%)	16/32 (50%)	1.52 (0.25–3.11)	0.25	1.43 (0.68–3.01)	0.34
Dyslipidemia	216/376 (57%)	212/339 (63%)	0.81 (0.60–1.09)	0.17	0.78 (0.52–1.17)	0.24
Diabetes mellitus	69/117 (59%)	359/598 (60%)	0.96 (0.64–1.44)	0.83	0.99 (0.65–1.52)	0.96
Current smoking	74/122 (61%)	350/583 (60%)	1.03 (0.69–1.54)	0.9	1.08 (0.72–1.64)	0.71
Chronic kidney disease	88/136 (65%)	340/579 (59%)	1.29 (0.88–1.91)	0.2	1.29 (0.85–1.97)	0.23
Cardio- and Cerebrovascular disease	178/309 (58%)	250/406 (62%)	0.85 (0.63–1.15)	0.28	0.91 (0.61–1.38)	0.67

Multiple logistic regression modeling was used to examine the association of WC with low FMD (<4.8%) in male, expressed as ORs and 95% CIs. WC, waist circumference. Other abbreviations as in Table 2.

Table 4. Adjusted Effect of WC ≥90 cm for FMD <4.8% in Females

Variables	N of events/present N of patients	N of events/absent N of patients	Univariate		Multivariate
			OR (95% CI)	P value	OR (95% CI)	P value
WC ≥90 cm	69/119 (58%)	140/253 (55%)	1.11 (0.72–1.73)	0.63	1.08 (0.68–1.73)	0.74
Age ≥65 years	113/182 (62%)	96/190 (51%)	1.60 (1.06–2.43)	0.025	1.56 (1.00–2.43)	0.049
Uric acid ≥7.0 mg/dL	11/19 (59%)	198/352 (56%)	1.07 (0.42–2.82)	0.89	0.97 (0.37–2.63)	0.95
Hypertension	199/354 (56%)	10/18 (56%)	1.03 (0.38–2.66)	0.96	1.00 (0.37–2.63)	0.99
Dyslipidemia	92/159 (58%)	117/213 (55%)	1.13 (0.74–1.71)	0.57	1.00 (0.61–1.62)	0.99
Diabetes mellitus	32/52 (62%)	177/320 (55%)	1.29 (0.71–2.39)	0.4	1.04 (0.53–2.09)	0.91
Current smoking	16/28 (57%)	189/337 (56%)	1.04 (0.48–2.32)	0.91	1.19 (0.53–2.72)	0.67
Chronic kidney disease	34/54 (63%)	175/318 (55%)	1.39 (0.77–2.55)	0.27	1.20 (0.64–2.29)	0.56
Cardio- and Cerebrovascular disease	22/57 (61%)	174/315 (55%)	1.29 (0.73–2.32)	0.39	1.03 (0.52–2.03)	0.94

Multiple logistic regression modeling was used to identify the effect of WC on low FMD (<4.8%) in males, expressed as ORs and 95% CIs. Abbreviations as in Table 2,3.

Regarding BMI in both males and females, a risk factor for low FMD (<4.8%) in the univariate analysis by logistic regression modeling was not low BMI (<25.0), only advanced age. After adjusting for confounders such as uric acid (≥7.0 mg/dL), hypertension, dyslipidemia, DM, current smoking, chronic kidney disease and cardio-and cerebrovascular disease, high BMI was not found to be a predictor (OR 0.92, 95% CI 0.71–1.19, P=0.54), only advanced age was found as a predictor (OR 1.57, 95% CI 1.21–2.05, P=0.0007) (Tables 5,6).

Table 5. Baseline Characteristics of Study Patients Subdivided by BMI <25 kg/m² vs. ≥25 kg/m²

Variables	All (n=1,087)	BMI <25 (n=618)	BMI ≥25 (n=469)	P value
Age (years)	61.9±9.3	63.0±8.9	60.5±9.6	<0.0001
≥65 years	462 (43)	290 (47)	172 (37)	0.0007
Male (n (%))	715 (66)	382 (62)	333 (71)	0.002
BMI (kg/m2)	24.9±3.7	22.5±1.8	28.2±3.0	<0.0001
Waist (cm)	88.0±10.3	82.5±7.4	95.3±8.9	<0.0001
Height (cm)	161.8±8.7	161.2±8.6	162.5±8.8	0.014
WHtR	0.54±0.06	0.51±0.05	0.59±0.06	<0.0001
Systolic BP (mmHg)	133.7±15.3	133.2±15.3	134.4±15.3	0.18
Diastolic BP (mmHg)	78.8±10.9	78.4±11.2	79.3±10.4	0.16
Heart rate (beats/min)	71.3±12.4	71.3±12.5	71.3±12.4	0.97
Total cholesterol (mmol/L)	4.93±0.87	4.98±0.87	4.88±0.87	0.055
Triglycerides (mmol/L)	1.23 (0.87–1.75)	1.12 (0.81–1.57)	1.39 (0.95–1.94)	<0.0001
HDL-C (mmol/L)	1.44±0.39	1.53±0.42	1.33±0.33	<0.0001
LDL-C (mmol/L)	3.20±0.79	3.19±0.80	3.22±0.77	0.49
Glucose (mmol/L)	5.61 (5.22–6.23)	5.53 (5.17–6.00)	5.78 (5.39–6.56)	<0.0001
Uric acid (mg/dL)	5.80±1.35	5.57±1.34	6.11±1.29	<0.0001
eGFR (mL/min/1.73 m2)	75.0±18.3	75.2±17.9	74.7±18.8	0.59
Framingham risk score	6.04±3.00	6.01±3.02	6.08±2.97	0.71
Hypertension (n (%))	1,037 (95)	580 (94)	457 (97)	0.005
Dyslipidemia (n (%))	535 (49)	273 (44)	262 (56)	0.0001
Diabetes mellitus (n (%))	169 (16)	65 (11)	104 (22)	<0.0001
Cardio- and cerebrovascular disease (n (%))	366 (34)	201 (33)	165 (35)	0.36
Current smoking (n (%))	150 (14)	81 (13)	69 (15)	0.45
Chronic kidney disease (n (%))	190 (17)	106 (17)	84 (18)	0.74
Flow-mediated vasodilation (%)	4.48±2.59	4.52±2.64	4.39±2.53	0.63
Baseline brachial artery diameter (mm)	4.16±0.63	4.06±0.64	4.29±0.60	<0.0001
Mean ankle-brachial index	1.14±0.08	1.14±0.08	1.15±0.09	0.022
hs-CRP (mg/dL)	0.52 (0.26–1.08)	0.44 (0.20–0.93)	0.64 (0.34–1.23)	<0.0001
Small-dense LDL (mg/dL)	36.6±15.6	35.1±14.7	38.7±16.5	0.0002
Oxidized LDL (U/L)	119.7±39.7	117.3±38.8	122.8±40.7	0.025
Medications
Calcium antagonists (n (%))	638 (59)	324 (52)	314 (67)	0.0001
ACEI/ARB (n (%))	775 (71)	423 (68)	352 (75)	0.017
β-blockers (n (%))	274 (25)	136 (22)	138 (29)	0.005
Diuretics (n (%))	271 (25)	123 (20)	148 (32)	<0.0001
Aldosterone antagonists (n (%))	37 (3.4)	17 (2.8)	20 (4.3)	0.17
Statins	486 (45)	244 (39)	242 (52)	<0.0001
Fibrate preparations	29 (2.7)	15 (2.4)	14 (3.0)	0.57
EPA preparations	32 (2.9)	18 (2.9)	14 (3.0)	0.94
Oral antidiabetic agents	157 (14)	61 (9.9)	96 (20)	<0.0001
Use of insulin	23 (2.1)	8 (1.3)	15 (3.2)	0.031
Aspirin	401 (37)	217 (35)	184 (39)	0.16
Ticlopidine	54 (5.0)	32 (5.2)	22 (4.7)	0.71
Clopidogrel	121 (11)	73 (12)	48 (10)	0.41
Nitrate agents	63 (5.8)	34 (5.5)	29 (6.2)	0.63
Nicorandil	83 (7.6)	50 (8.1)	33 (7.0)	0.52

Continuous variables are shown as mean±standard deviation or median (interquartile range). Categorical variables were compared with the chi-square test or Fisher’s exact test. Continuous variables were compared using Student’s t-test or the Wilcoxon rank-sum test based on distributions. Abbreviations as in Table 1.

Table 6. Adjusted Effect of BMI <25.0 kg/m² for FMD <4.8%

Variables	Present N of events/ N of patients	Absent N of events/ N of patients	Univariate		Multivariate
			OR (95% CI)	P value	OR (95% CI)	P value
BMI <25.0 kg/m2	360/637 (57%)	258/450 (57%)	0.97 (0.76–1.23)	0.79	0.92 (0.71–1.19)	0.54
Age ≥65 years	299/462 (65%)	338/625 (54%)	1.56 (1.22–2.00)	0.0004	1.57 (1.21–2.05)	0.0007
Male	428/715 (60%)	209/372 (56%)	1.16 (0.90–1.50)	0.24	1.28 (0.96–1.71)	0.088
Uric acid ≥7.0 mg/dL	119/205 (58%)	518/880 (59%)	0.97 (0.71–1.32)	0.83	0.87 (0.62–1.22)	0.42
Hypertension	611/1,037 (59%)	26/50 (52%)	1.32 (0.75–2.34)	0.33	1.32 (0.74–2.37)	0.34
Dyslipidemia	308/535 (58%)	329/552 (60%)	0.92 (0.72–1.17)	0.5	0.87 (0.64–1.18)	0.38
Diabetes mellitus	101/169 (60%)	536/918 (58%)	1.06 (0.76–1.48)	0.73	1.03 (0.72–1.49)	0.86
Current smoking	90/150 (60%)	539/920 (59%)	1.06 (0.75–1.51)	0.74	1.12 (0.78–1.62)	0.54
Chronic kidney disease	122/190 (64%)	515/897 (57%)	1.33 (0.96–1.84)	0.082	1.28 (0.91–1.82)	0.16
Cardio- and Cerebrovascular disease	213/366 (58%)	424/721 (59%)	0.98 (0.76–1.26)	0.85	0.90 (0.64–1.27)	0.56

A multiple logistic regression model was used to identify the effect of waist circumference on low FMD (<4.8%) in males, with effects expressed as ORs and 95% CIs. Abbreviations as in Table 2.

Discussion

This study showed that: (1) the high-WHtR group displayed a greater risk of morbidity and increased inflammation in association with atherosclerosis relative to the low-WHtR group; (2) high WHtR was predictive of low FMD (<4.8%), but neither WC nor BMI was associated with a risk of morbidity in either sex.

Overall or central obesity is related to cardiovascular risk factors such as DM, hypertension and dyslipidemia.^1,19 BMI is reflective of overall obesity, but has difficulty in distinguishing between fat and muscle. In contrast, previous reports have shown that WC is significantly associated with central obesity and visceral fat.^20,21 However, WC appears to be influenced by factors such as age, sex, ethnicity, height, and weight.^6,7 In particular, height seems to cause deviations in WC and visceral fat. In fact, the current Japanese cutoffs for WC in defining MetS (85 cm for men, 90 cm for women) remain controversial.^8,9,22 In the present study, WC and BMI were not associated with a risk of low FMD (<4.8% in both sexes). This was attributed to height causing deviations in WC and visceral fat, as stated above. Similarly, BMI is calculated from height and overall body weight, which can lead individuals with low levels of adiposity being identified as obese by mistake. In addition, endothelial dysfunction appears to be more significantly associated with abnormal fat distributions than with obesity per se. In fact, Hashimoto et al reported that subjects with visceral-type obesity, rather than subcutaneous-type, showed a stronger association with impaired flow-mediated endothelium-dependent vasodilatation of the brachial artery.²³ In addition, a systematic review based on meta-analysis reported WHtR as a better screening tool than WC and BMI for adult cardiometabolic risk factors.²⁴ On the other hand, Joseph et al reported no clear correlation among older sedentary men between adiposity or body fat distribution and impairment of flow-mediated endothelium-dependent vasodilation.²⁵ However, this result may have been because the subjects in that study were limited to older men. Because of the absence of easy and appropriate anthropometric indices for predicting endothelial dysfunction, we consider that better indices are needed.

In this study, the high-WHtR group showed a higher risk of morbidity relative to the low-WHtR group, related to increased inflammatory markers and oxidative stress in association with measures of atherosclerosis such as hs-CRP, small-dense LDL and oxidized LDL. Furthermore, high WHtR was predictive of low FMD (<4.8%), which is strongly associated with endothelial dysfunction. These results suggested that visceral adiposity causes inflammation and oxidative stress, leading to endothelial dysfunction. The mechanisms by which endothelial dysfunction is caused by obesity appear complicated,^26,27 but previous reports have implied that increased oxidative stress from accumulated fat plays a dominant role in eliciting endothelial dysfunction in obesity.^28,29

The current study differs from previous studies in several ways. The current study analyzed only baseline data as a cross-sectional study, whereas the FMD-J study was a prospective multicenter study in Japan, and no large cohort studies had been conducted prior to that investigation. FMD has been assessed by many individual institutions, but the reliability of FMD assessment has been increased using various methods.¹³ We therefore believe that the accuracy of the present results is significantly higher than in previous reports. WHtR is able to be measured easily, as are BMI and WC, during periodic health examinations, and does not require separation by sex. Furthermore, our findings indicated superior predictive ability of this factor for endothelial dysfunction compared with WC and BMI.

Study Limitations

Some limitations to this study need to be considered when interpreting the results. First, subjects in this study only comprised patients with morbidities such as DM, hypertension and CAD, and healthy individuals were not included. Second, because we did not obtain any data on other factors that are known to affect FMD (such as vitamin and phosphate levels) and other as-yet-unidentified confounding factors, the results of this study might have been inadvertently affected. Finally, this study used a cross-sectional design, which does not allow for definitive determinations of the direction of any relationship between high WHtR and low FMD.

Conclusions

WHtR was a superior predictor of decreased FMD compared with other anthropometric indices (i.e., WC and BMI), and progression of arteriosclerosis might be detected more sensitively. Accordingly, we propose that WHtR should be used more positively in clinical practice. Further study is needed to clarify the relationship between cardiovascular mortality and WHtR.

Acknowledgments

The authors acknowledge the valuable contributions of the FMD-J collaborative investigators.

Disclosures

The authors declare that there are no conflicts of interest.

Grant Support

The present study was supported by the Japanese Atherosclerosis Prevention Research Fund (A study of Multi-center Assessment on the Clinical Usefulness of Semi-automatic Measurement of Flow-mediated Vasodilatation of the Brachial Artery: FMD-J study).

References

• 1.   Hubert H, Feinlieb M, McNamara P, Castelli W. Obesity as an independent risk factor for cardiovascular disease: A 26-year follow-up of participants in the Framingham Heart Study. Circulation 1983; 67: 968–977.
• 2.   Tchernof A, Després JP. Pathophysiology of human visceral obesity: An update. Physiol Rev 2013; 93: 359–404.
• 3.   Tziomalos K, Athyros VG, Karagiannis A, Mikhailidis DP. Endothelial dysfunction in metabolic syndrome: Prevalence, pathogenesis and management. Nutr Metab Cardiovasc Dis 2010; 20: 140–146.
• 4.   Mottillo S, Filion KB, Genest J, Joseph L, Pilote L, Poirier P, et al. The metabolic syndrome and cardiovascular risk: A systematic review and meta-analysis. J Am Coll Cardiol 2010; 56: 1113–1132.
• 5.   Matsuzawa Y. Metabolic syndrome--definition and diagnostic criteria in Japan. J Atheroscler Thromb 2005; 12: 301.
• 6.   Klein S, Allison DB, Heymsfield SB, Kelley DE, Leibel RL, Nonas C, et al. Waist circumference and cardiometabolic risk: A consensus statement from Shaping America’s Health: Association for Weight Management and Obesity Prevention; NAASO, the Obesity Society; the American Society for Nutrition; and the American Diabetes Associat. Diabetes Care 2007; 30: 1647–1652.
• 7.   Nakamura K, Nanri H, Hara M, Higaki Y, Imaizumi T, Taguchi N, et al. Optimal cutoff values of waist circumference and the discriminatory performance of other anthropometric indices to detect the clustering of cardiovascular risk factors for metabolic syndrome in Japanese men and women. Environ Health Prev Med 2011; 16: 52–60.
• 8.   Oka R, Kobayashi J, Yagi K, Tanii H, Miyamoto S, Asano A, et al. Reassessment of the cutoff values of waist circumference and visceral fat area for identifying Japanese subjects at risk for the metabolic syndrome. Diabetes Res Clin Pract 2008; 79: 474–481.
• 9.   Shimajiri T, Imagawa M, Kokawa M, Konami T, Hara H, Kyoku I, et al. Revised optimal cut-off point of waist circumference for the diagnosis of metabolic syndrome in Japanese women and the influence of height. J Atheroscler Thromb 2008; 15: 94–99.
• 10.   Ross R. Atherosclerosis: An inflammatory disease. N Eng J Med 1999; 340: 115–126.
• 11.   Kullo IJ, Malik AR. Arterial ultrasonography and tonometry as adjuncts to cardiovascular risk stratification. J Am Coll Cardiol 2007; 49: 1413–1426.
• 12.   Tomiyama H, Kohro T, Higashi Y, Takase B, Suzuki T, Ishizu T, et al. A multicenter study design to assess the clinical usefulness of semi-automatic measurement of flow-mediated vasodilatation of the brachial artery. Int Heart J 2012; 53: 170–175.
• 13.   Tomiyama H, Kohro T, Higashi Y, Takase B, Suzuki T, Ishizu T, et al. Reliability of measurement of endothelial function across multiple institutions and establishment of reference values in Japanese. Atherosclerosis 2015; 242: 433–442.
• 14.   Browning LM, Hsieh SD, Ashwell M. A systematic review of waist-to-height ratio as a screening tool for the prediction of cardiovascular disease and diabetes: 0·5 could be a suitable global boundary value. Nutr Res Rev 2010; 23: 247–269.
• 15.   American Diabetes Association: Clinical practice recommendations 1999. Diabetes Care 1999; 22(Suppl 1): S1–S114.
• 16.   Wilson PW, Castelli WP, Kannel WB. Coronary risk prediction in adults (the Framingham Heart Study). Am J Cardiol 1987; 59: 91G–94G.
• 17.   Hirano T, Ito Y, Yoshino G. Measurement of small dense low-density lipoprotein particles. J Atheroscler Thromb 2005; 12: 67–72.
• 18.   Kotani K, Maekawa M, Kanno T, Kondo A, Toda N, Manabe M. Distribution of immunoreactive malondialdehyde-modified low-density lipoprotein in human serum. Biochim Biophys Acta (BBA)/Lipids Lipid Metab 1994; 1215: 121–125.
• 19.   Bermudez OI, Tucker KL. Total and central obesity among elderly Hispanics and the association with Type 2 diabetes. Obes Res 2001; 9: 443–451.
• 20.   Lemieux S, Prud’homme D, Bouchard C, Tremblay A, Després JP. A single threshold value of waist girth identifies normal-weight and overweight subjects with excess visceral adipose tissue. Am J Clin Nutr 1996; 64: 685–693.
• 21.   Han TS, McNeill G, Seidell JC, Lean ME. Predicting intra-abdominal fatness from anthropometric measures: The influence of stature. Int J Obes Relat Metab Disord 1997; 21: 587–593.
• 22.   Hara K, Matsushita Y, Horikoshi M, Yoshiike N, Yokoyama T, Tanaka H, et al. A proposal for the cutoff point of waist circumference for the diagnosis of metabolic syndrome in the Japanese population. Diabetes Care 2006; 29: 1123–1124.
• 23.   Hashimoto M, Akishita M, Eto M, Kozaki K, Ako J, Sugimoto N, et al. The impairment of flow-mediated vasodilatation in obese men with visceral fat accumulation. Int J Obes Relat Metab Disord 1998; 22: 477–484.
• 24.   Ashwell M, Gunn P, Gibson S. Waist-to-height ratio is a better screening tool than waist circumference and BMI for adult cardiometabolic risk factors: Systematic review and meta-analysis. Obes Rev 2012; 13: 275–286.
• 25.   Joseph LJO, Ryan AS, Sorkin J, Mangano C, Brendle DC, Corretti MC, et al. Body fat distribution and flow-mediated endothelium-dependent vasodilation in older men. Int J Obes Relat Metab Disord 2002; 26: 663–669.
• 26.   Gustafson B. Adipose tissue, inflammation and atherosclerosis. J Atheroscler Thromb 2010; 17: 332–341.
• 27.   Boden G. Obesity and free fatty acids. Endocrinol Metab Clin North Am 2008; 37: 635–646.
• 28.   Böger RH. Asymmetric dimethylarginine (ADMA) modulates endothelial function: Therapeutic implications. Vasc Med 2003; 8: 149–151.
• 29.   Eid HMA, Arnesen H, Hjerkinn EM, Lyberg T, Seljeflot I. Relationship between obesity, smoking, and the endogenous nitric oxide synthase inhibitor, asymmetric dimethylarginine. Metabolism 2004; 53: 1574–1579.