Table 5.

Association between the WQS index and obesity in NHANES 2005–2010 (N = 2372)

Outcomes	OR/ β	95% CI of OR	P value
Obesity
Model 1	1.50	(1.19, 1.90)	< 0.001
Model 2	1.51	(1.19, 1.91)	< 0.001
Model 3	1.48	(1.16, 1.89)	0.002
BMI z-score
Model 1	0.028	(−0.09, 0.15)	0.643
Model 2	0.033	(−0.09, 0.15)	0.584
Model 3	0.001	(−0.12, 0.12)	0.983

NHANES National Health and Nutrition Examination Survey, CI Confidence interval. The weighted quantile sum (WQS) regression was fitted for the obesity and BMI z-score, which scored all the chemical exposures into quantiles and estimated the weight index. OR estimates represent the odds ratios of obesity as 1 quartile increased in the WQS index. β estimates represent the mean differences in the BMI z-score as 1 quartile increased in the WQS index. Model 1: Adjusted for age, gender, ethnicity, and log-transformed creatinine. Model 2: Adjusted for age, gender, ethnicity, caloric intake, serum cotinine, and log-transformed creatinine. Model 3: Adjusted for age, gender, ethnicity, educational levels, family income-to-poverty ratio, caloric intake, serum cotinine, and log-transformed creatinine