Neck Circumference as Predictor of Cardio-Metabolic Syndrome
The anthropometric and metabolic characteristics of the participants by gender are presented in Table 1 . The 4201 subjects had a mean age of 43.7±12.0 years (men: 44.1±12.4, women: 43.1±11.3 years), mean BMI of 23.73±3.15 kg/m (men: 24.44±2.98, women: 22.67±3.10 kg/m) and mean WC and NC of 82.37±10.08 cm (men: 86.44±8.71, women: 76.36±8.88 cm) and 35.41±3.39 (men: 37.40±2.46, women: 32.46±2.24 cm), respectively. Of the 4201 subjects, MS was diagnosed in 443 men (17.66%) and 215 women (12.70%). High BP was diagnosed in 965 men (38.48%) and 429 women (25.34%). Increased FBG was found in 349 (13.92%) men and 150 (8.86%) women. Increased TG was found in 972 men (38.76%) and 287 women (16.95%). Decreased HDL-C was found in 646 men (25.76%) and 471 women (27.82%). Increased LDL-C was detected in 1395 men (55.62%) and 743 women (43.89%). Increased TC was detected in 1633 men (65.11%) and 1044 women (61.79%).
The partial correlations between cardio-metabolic risk factors and anthropometry indices by gender are shown in Table 2 . In male subjects, after adjusting for age, the NC revealed a positive correlation with SBP and DBP (r=0.250 and 0.261, both P<0.01), FBG (r=0.177, P<0.01), TG (r=0.240, p<0.01), TC (r=0.143, p<0.01) and LDL-C (r=0.088, p<0.01). In contrast, NC was negatively correlated with HDL-C (r=-0.202, p<0.01). Similarly, in female subjects, NC showed the same trend in the correlation with cardio-metabolic risk factors, with the exception of TC. The correlations of other anthropometric indices to the cardio-metabolic risk factors were the same as those of NC to the cardio-metabolic risk factors in both genders.
The areas under the ROC curves (AUCs) were constructed to evaluate the predictive values of anthropometric indices for cardio-metabolic risks ( Table 3 ). The AUCs of NC were 0.635 and 0.659 for the risk factor of BP, 0.659 and 0.678 for TG, 0.618 and 0.641 for HDL-C, 0.624 and 0.646 for glucose, 0.558 and 0.596 for LDL-C and 0.683 and 0.703 for MS-rf in men and women, respectively. An NC of ≥37 cm for men and ≥33 cm for women were the best values of combined sensitivity and specificity in identifying MS. The AUCs of NC for all cardio-metabolic risk factors, except BMI for glucose in men and BMI for HDL-C in women, were lower than those of other anthropometric measurements, including WC, BMI and WHpR.
The crude and adjusted ORs for the cardio-metabolic risk factors for per SD increase in NC are presented in Figure 1 Generally, greater NC was linearly associated with higher risk of metabolic abnormalities in both genders. Women tended to have closer associations than men. Based on the univariate logistic regression analysis with MS-rf as the dependent variable, the crude OR was 2.03 (95% CI: 1.84–2.23) in men and 2.03 (95% CI: 1.79–2.29) in women (p<0.01 in both).
(Enlarge Image)
Figure 1.
Gender-specific odds ratios (ORs) and 95% confidence intervals (95% CIs) of cardio-metabolic risk factors for neck circumference (NC). OR: unadjusted, OR: adjusted for age, OR: adjusted for age, BMI, WC and WHpR.
After adjusting for age, NC was significantly associated with MS-rf at ORs 2.06 in men and 1.88 in women (P<0.01 in both). With further adjustments for BMI, WC and WHpR, the significant associations remained between NC and MS-rf with ORs of 1.29 in men and 1.44 in women (P<0.01 in both). In the multivariate model including age, BMI, WC and WHpR, NC also showed obvious correlations with the occurrence of high BP (OR, 1.15 in men and 1.22 in women), increased TG (OR, 1.16 in men and 1.42 in women), increased FBP (OR, 1.26 in men and 1.32 in women) and decreased HDL-C (OR, 1.29 in women). We also made the adjustments for anthropometric indices (BMI, WC and WHpR) one by one due to close colinearity among them, similar results were found to those observed in the previous multivariate model included all of them (data not shown).
Results
The anthropometric and metabolic characteristics of the participants by gender are presented in Table 1 . The 4201 subjects had a mean age of 43.7±12.0 years (men: 44.1±12.4, women: 43.1±11.3 years), mean BMI of 23.73±3.15 kg/m (men: 24.44±2.98, women: 22.67±3.10 kg/m) and mean WC and NC of 82.37±10.08 cm (men: 86.44±8.71, women: 76.36±8.88 cm) and 35.41±3.39 (men: 37.40±2.46, women: 32.46±2.24 cm), respectively. Of the 4201 subjects, MS was diagnosed in 443 men (17.66%) and 215 women (12.70%). High BP was diagnosed in 965 men (38.48%) and 429 women (25.34%). Increased FBG was found in 349 (13.92%) men and 150 (8.86%) women. Increased TG was found in 972 men (38.76%) and 287 women (16.95%). Decreased HDL-C was found in 646 men (25.76%) and 471 women (27.82%). Increased LDL-C was detected in 1395 men (55.62%) and 743 women (43.89%). Increased TC was detected in 1633 men (65.11%) and 1044 women (61.79%).
The partial correlations between cardio-metabolic risk factors and anthropometry indices by gender are shown in Table 2 . In male subjects, after adjusting for age, the NC revealed a positive correlation with SBP and DBP (r=0.250 and 0.261, both P<0.01), FBG (r=0.177, P<0.01), TG (r=0.240, p<0.01), TC (r=0.143, p<0.01) and LDL-C (r=0.088, p<0.01). In contrast, NC was negatively correlated with HDL-C (r=-0.202, p<0.01). Similarly, in female subjects, NC showed the same trend in the correlation with cardio-metabolic risk factors, with the exception of TC. The correlations of other anthropometric indices to the cardio-metabolic risk factors were the same as those of NC to the cardio-metabolic risk factors in both genders.
The areas under the ROC curves (AUCs) were constructed to evaluate the predictive values of anthropometric indices for cardio-metabolic risks ( Table 3 ). The AUCs of NC were 0.635 and 0.659 for the risk factor of BP, 0.659 and 0.678 for TG, 0.618 and 0.641 for HDL-C, 0.624 and 0.646 for glucose, 0.558 and 0.596 for LDL-C and 0.683 and 0.703 for MS-rf in men and women, respectively. An NC of ≥37 cm for men and ≥33 cm for women were the best values of combined sensitivity and specificity in identifying MS. The AUCs of NC for all cardio-metabolic risk factors, except BMI for glucose in men and BMI for HDL-C in women, were lower than those of other anthropometric measurements, including WC, BMI and WHpR.
The crude and adjusted ORs for the cardio-metabolic risk factors for per SD increase in NC are presented in Figure 1 Generally, greater NC was linearly associated with higher risk of metabolic abnormalities in both genders. Women tended to have closer associations than men. Based on the univariate logistic regression analysis with MS-rf as the dependent variable, the crude OR was 2.03 (95% CI: 1.84–2.23) in men and 2.03 (95% CI: 1.79–2.29) in women (p<0.01 in both).
(Enlarge Image)
Figure 1.
Gender-specific odds ratios (ORs) and 95% confidence intervals (95% CIs) of cardio-metabolic risk factors for neck circumference (NC). OR: unadjusted, OR: adjusted for age, OR: adjusted for age, BMI, WC and WHpR.
After adjusting for age, NC was significantly associated with MS-rf at ORs 2.06 in men and 1.88 in women (P<0.01 in both). With further adjustments for BMI, WC and WHpR, the significant associations remained between NC and MS-rf with ORs of 1.29 in men and 1.44 in women (P<0.01 in both). In the multivariate model including age, BMI, WC and WHpR, NC also showed obvious correlations with the occurrence of high BP (OR, 1.15 in men and 1.22 in women), increased TG (OR, 1.16 in men and 1.42 in women), increased FBP (OR, 1.26 in men and 1.32 in women) and decreased HDL-C (OR, 1.29 in women). We also made the adjustments for anthropometric indices (BMI, WC and WHpR) one by one due to close colinearity among them, similar results were found to those observed in the previous multivariate model included all of them (data not shown).
SHARE
