Skip to main content

Table 3 Performance of different VAT and waist optimal thresholds to detect two components of the metabolic syndrome (excluding waist circumference)

From: The discriminatory power of visceral adipose tissue area vs anthropometric measures as a diagnostic marker for metabolic syndrome in South African women

Thresholdsa

Methods

App prev

True prev

Sensitivity

Specificity

Accuracy

DOR

NND

Youden index

PPV

NPV

LR+

LR−

VAT area

 174.0 (137.7–181.5)

Youden index

47.3 (40.3–54.4)

40.9 (34.0–48.0)

0.723 (0.614–0.815)

0.700 (0.610–0.780)

0.709 (0.642–0.770)

6.02 (3.28–11.31)

2.36 (1.68–4.48)

0.423 (0.223–0.596)

0.625 (0.520–0.722)

0.785 (0.695–0.859)

2.41 (1.78–3.27)

0.40 (0.27–0.57)

 175.50 (154.6–180.1)

CTL

47.8 (40.7–54.9)

40.9 (34.0–48.0)

0.735 (0.627–0.826)

0.700 (0.610–0.780)

0.714 (0.647–0.775)

6.47 (3.46–12.08)

2.30 (1.65–4.23)

0.435 (0.236–0.606)

0.629 (0.525–0.725)

0.792 (0.703–0.865)

2.45 (1.81–3.31)

0.38 (0.26–0.55)

WC

 94.40 (89.7–102.3)

Youden index

36.4 (29.8–43.5)

40.9 (34.0–48.0)

0.614 (0.501–0.719)

0.808 (0.726–0.874)

0.729 (0.662–0.789)

6.72 (3.56–12.67)

2.36 (1.68–4.39)

0.423 (0.228–0.594)

0.689 (0.571–0.792)

0.752 (0.668–0.824)

3.21 (2.14–4.81)

0.48 (0.36–0.63)

 94.40 (93.3–101.2)

CTL

36.4 (29.8–43.5)

40.9 (34.0–48.0)

0.614 (0.501–0.719)

0.808 (0.726–0.874)

0.729 (0.662–0.789)

6.72 (3.56–12.67)

2.36 (1.68–4.39)

0.423 (0.228–0.594)

0.689 (0.571–0.792)

0.752 (0.668–0.824)

3.21 (2.14–4.81)

0.48 (0.36–0.63)

 80b

 

10.8 (6.9–15.9)

40.9 (34.0–48.0)

0.205 (0.124–0.307)

0.958 (0.905––0.986)

0.650 (0.580–0.716)

5.92 (2.09–16.79)

6.13 (3.40–33.87)

0.163 (0.029–0.294)

0.773 (0.546–0.922)

0.635 (0.561–0.705)

4.92 (1.89–12.80)

0.83 (0.74–0.93)

 90b

 

26.1 (20.2–32.7)

40.9 (34.0–48.0)

0.482 (0.371–0.594)

0.892 (0.822–0.941)

0.724 (0.657–0.784)

7.66 (3.73–15.71)

2.68 (1.87–5.19)

0.374 (0.193–0.535)

0.755 (0.617–0.862)

0.713 (0.634–0.784)

4.45 (2.54–7.78)

0.58 (0.47–0.72)

  1. App prev apparent prevalence, DOR diagnostic odd ratio, LR− likelihood of a negative test, LR+ likelihood of a positive test, NND number needed to diagnose, NPV negative predictive value, PPV positive predictive value, thresh threshold, True prev true prevalence
  2. aCI P2.5 and P97.5 from bootstrap, WC 80 cmb (JIS), WC 90 cmb [31]