Fig. 1Receiver operating characteristic analysis of waist circumference (WC) to detect the presence of two or more metabolic components, visceral obesity, and insulin resistance (IR) in (A) men and (B) women with type 2 diabetes. The higher the area under the curve, the greater the predictive power. A WC of 87 cm for men (black circles) and 81 cm for women (black triangles) was chosen as the discriminating value to predict two or more metabolic components, and the presence of visceral obesity and IR. These cutoff points had the highest predictive powers for the presence of visceral obesity. Moreover, they had higher sensitivity and specificity compared to the previous criteria—90 cm for men (black squares) and 85 cm for women (black diamonds)—proposed by the Korean Society for the Study of Obesity. Visceral obesity was defined as a visceral fat thickness of greater than 47.6 mm for men and 35.5 mm for women, and IR was defined as the low tertile of insulin sensitivity.
Fig. 2Differences in the prevalence of metabolic components according to Korean Society for the Study of Obesity (KOSSO) criteria and our criteria in (A) men and (B) women with type 2 diabetes. Compared to metabolic syndrome (MetS) as defined by the KOSSO criteria, the prevalence of each metabolic component, including hypertriglyceridemia (hyper-TG), low high density lipoprotein (HDL) cholesterol, and hypertension (HTN), was significantly greater in both sexes when our criteria were used to define MetS.
Table 1Baseline characteristics
Characteristic |
Total (n=4,252) |
Men (n=2,220) |
Women (n=2,032) |
Age, yr |
57.24±10.27 |
55.47±10.80 |
59.17±9.27 |
DM duration, yr |
7.22±6.89 |
7.04±7.04 |
7.43±6.71 |
Body mass index, kg/m2
|
24.69±3.39 |
24.61±3.24 |
24.77±3.55 |
Waist circumference, cm |
84.60±8.14 |
87.26±7.38 |
81.69±7.95 |
Systolic blood pressure, mm Hg |
136.04±17.93 |
134.39±17.22 |
137.86±18.51 |
Diastolic blood pressure, mm Hg |
86.91±11.24 |
88.24±11.31 |
85.44±10.97 |
Kitt, %/min |
1.96±0.92 |
1.95±0.91 |
1.98±0.93 |
Fasting glucose, mg/dL |
157.81±56.79 |
159.44±56.72 |
156.02±56.83 |
C-peptide, ng/mL |
2.11±0.85 |
2.14±0.84 |
2.08±0.87 |
Glycosylated hemoglobin, % |
8.28±1.83 |
8.30±1.90 |
8.26±1.74 |
Total cholesterol, mg/dL |
197.07±40.94 |
192.34±40.35 |
202.25±40.95 |
Triglyceride, mg/dL |
161.84±115.30 |
166.17±126.98 |
157.11±100.82 |
HDL-C, mg/dL |
48.94±12.74 |
46.90±12.14 |
51.19±13.00 |
LDL-C, mg/dL |
115.55±34.15 |
112.60±32.77 |
118.79±35.34 |
Visceral fat thickness, mm |
46.14±17.54 |
49.75±17.35 |
42.19±16.88 |
CIMT, mm |
0.833±0.182 |
0.847±0.185 |
0.818±0.179 |
Medications for dyslipidemia, % |
698 (16.4) |
314 (14.1) |
384 (18.9) |
Medications for hypertension, % |
1,207 (28.4) |
562 (25.3) |
645 (31.7) |
Antidiabetic treatment, % |
|
|
|
Insulin |
450 (10.6) |
193 (8.7) |
257 (12.6) |
Oral medicationsa
|
2,497 (58.7) |
1,234 (55.6) |
1,263 (62.2) |
Only diet and exercise |
1,305 (30.7) |
793 (35.7) |
512 (25.2) |
MetS components, %b
|
|
|
|
Hypertriglyceridemia |
2,127 (50.0) |
1,059 (47.7) |
1,068 (52.6) |
Low HDL-C |
1,600 (37.6) |
589 (26.7) |
1,011 (50.4) |
Hypertension |
3,147 (75.0) |
1,643 (74.9) |
1,504 (75.0) |
More than 2 factors |
2,288 (54.6) |
1,068 (48.7) |
1,220 (61.1) |
Table 2Cutoff values for waist circumference to identify subjects with two or more metabolic components, visceral obesity, and insulin resistance
Variable |
Two or more metabolic risk factorsa
|
Presence of visceral obesityb
|
Presence of IRc
|
Sensitivity, % |
Specificity, % |
Sensitivity, % |
Specificity, % |
Sensitivity, % |
Specificity, % |
Men |
|
|
|
|
|
|
85 |
74.1 |
43.9 |
83.6 |
56.0 |
74.3 |
40.2 |
86 |
67.7 |
50.1 |
77.9 |
62.8 |
68.8 |
46.8 |
87 |
61.0 |
55.9 |
73.0 |
70.5 |
62.1 |
52.7 |
88 |
55.5 |
61.0 |
67.6 |
75.5 |
57.5 |
58.4 |
89 |
49.3 |
66.1 |
60.5 |
79.8 |
51.2 |
63.8 |
90 |
43.4 |
71.3 |
54.2 |
84.4 |
45.7 |
69.2 |
95 |
18.5 |
88.7 |
24.5 |
95.9 |
21.9 |
88.9 |
AUC (95% CI) |
0.616 (0.590-0.641) |
0.777 (0.756-0.798) |
0.610 (0.584-0.636) |
Women |
|
|
|
|
|
|
80 |
66.0 |
49.5 |
76.0 |
66.0 |
70.9 |
45.0 |
81 |
60.7 |
55.5 |
70.3 |
71.4 |
66.1 |
51.0 |
82 |
55.8 |
58.8 |
65.6 |
74.7 |
61.1 |
54.8 |
83 |
49.7 |
63.8 |
59.7 |
80.0 |
55.3 |
60.4 |
84 |
44.6 |
68.5 |
53.9 |
83.9 |
50.8 |
65.7 |
85 |
39.6 |
73.6 |
47.6 |
86.9 |
45.1 |
70.5 |
90 |
18.3 |
90.0 |
21.5 |
95.2 |
21.8 |
88.1 |
AUC (95% CI) |
0.611 (0.584-0.638) |
0.778 (0.756-0.800) |
0.603 (0.575-0.632) |
Table 3Differences in insulin sensitivity (Kitt), VFT, and CIMT according to the presence of metabolic syndrome defined by KOSSO criteria and our criteria in subjects with type 2 diabetes
Variable |
KOSSO criteria |
Our criteria |
MetS |
No MetS |
P value |
MetS |
No MetS |
P value |
Men, n
|
463 |
1,746 |
|
652 |
1,554 |
|
Kitt, %/min |
1.63±0.74 |
2.04±0.94 |
<0.001 |
1.68±0.78 |
2.07±0.94 |
<0.001 |
VFT, mm |
63.49±16.73 |
46.13±15.56 |
<0.001 |
60.23±16.86 |
45.36±15.54 |
<0.001 |
CIMT, mm |
0.851±0.189 |
0.846±0.184 |
0.666 |
0.849±0.188 |
0.846±0.184 |
0.729 |
Women, n
|
483 |
1,530 |
|
740 |
1,265 |
|
Kitt, %/min |
1.66±0.79 |
2.07±0.95 |
<0.001 |
1.70±0.80 |
2.13±0.96 |
<0.001 |
VFT, mm |
54.97±17.43 |
38.15±14.52 |
<0.001 |
50.91±16.96 |
37.08±14.62 |
<0.001 |
CIMT, mm |
0.828±0.182 |
0.817±0.177 |
0.244 |
0.834±0.184 |
0.811±0.174 |
0.007 |