Table 4.

Relationship between log-transformed total blood dioxins (in pg TEQ/g lipid) and health parameters/clinical biomarkers, 2011–2016 SEDOCCH (N = 490)

	Total TEQ with ‘zero’ substituted for dioxin congeners < LOD			Total TEQ with dioxins imputed using PMM			Total TEQ with dioxins imputed using MICE
Predictor	Coefficient (β)	p value	Adjusted R2	Coefficient (β)	p value	Adjusted R2	Coefficient (β)	p value	Adjusted R2
			0.410			0.483			0.440
Age (years)	0.298	< 0.001		0.420	< 0.001		0.360	< 0.001
BMI (kg/m2)	0.164	< 0.001		0.112	0.003		0.136	0.001
DGLA (μg/mL)	− 0.380	< 0.001		− 0.360	< 0.001		− 0.376	< 0.001
DHA (μg/mL)	0.362	< 0.001		0.315	< 0.001		0.321	< 0.001
Triglycerides (mg/dL)	0.158	0.002		0.134	0.002		0.131	0.003
Blood glucose (mg/dL)	0.057	0.142		0.098	0.008		0.087	0.024
ALT (IU/L)	0.067	0.090		0.048	0.215		0.080	0.050

Variables included in the model: age, BMI, triglycerides, blood glucose, blood urea nitrogen, ALT, DGLA, DHA, and sex

BMI, triglycerides, blood glucose, blood urea nitrogen, ALT, DGLA, and DHA were log-transformed

SEDOCCH survey on the exposure to dioxins and other chemical compounds in humans, PMM predictive mean matching, MICE multivariate imputation by chained equations, BMI body mass index, ALT alanine aminotransferase, DGLA dihomo-gamma-linolenic acid, DHA docosahexaenoic acid, LOD limit of detection