Relationship between neck circumference and body composition: a cross-sectional study based on a Chinese population.

Neck circumference was significantly positively correlated with BMI, FMI, VFA, and PBF in Chinese young adults.

• All correlations were statistically significant at p < 0.01.
• The study analyzed body composition data from 8,319 Chinese young adults.
• Statistical analyses included multi-model fitting to verify linear correlations, followed by Pearson correlation analysis and gender-stratified linear regression analysis.
• The study was cross-sectional and retrospective in design.

In females, neck circumference demonstrated very high diagnostic performance for obesity, with an AUC of 0.982 and an optimal cut-off value of 36.25 cm.

• The female NC AUC of 0.982 was higher than VFA's AUC of 0.976 in females.
• The optimal cut-off was determined using the Youden index.
• The female model showed better calibration with a Brier score of 0.0259.
• All AUC comparisons were significant at p < 0.001.

In males, neck circumference had lower diagnostic performance for obesity compared to BMI, PBF, and VFA, with an AUC of 0.918 and an optimal cut-off of 38.95 cm.

• Male NC AUC (0.918) was lower than BMI (0.979), PBF (0.985), and VFA (0.954) in males.
• The male model showed poorer calibration with a Brier score of 0.0648, compared to 0.0259 in females.
• All AUC comparisons were significant at p < 0.001.
• The optimal male cut-off of 38.95 cm was determined using the Youden index.

Decision Curve Analysis confirmed neck circumference's clinical utility for obesity screening across a wide range of probability thresholds.

• DCA was applied to evaluate net clinical benefit across threshold probabilities.
• The analysis confirmed NC's obesity clinical utility in both males and females.
• DCA was used alongside ROC curve analysis and calibration curve analysis as complementary evaluative tools.

Kappa testing revealed good consistency between neck circumference-based and BMI-based obesity classification, with higher agreement in females than males.

• Female Kappa value was 0.790, indicating good consistency.
• Male Kappa value was 0.644, indicating moderate-to-good consistency.
• Both Kappa values were statistically significant at p < 0.001.
• The Kappa test was used to validate NC's consistency with BMI-defined classification of obesity.