Skip to main content

Table 5 Multivariable models of HOMA-IR in men without diabetes: parameter estimates and 95% confidence intervals for PCB exposure group, age, BMI group and lipids, n = 52 a

From: Polychlorinated biphenyl exposure, diabetes and endogenous hormones: a cross-sectional study in men previously employed at a capacitor manufacturing plant

Variables in model

PCBsbc

Lipid

Qrts

Job

Dioxin-

Non-

Estrogenic

Anti-

  

PCBsb

workedb

scoreb

like

dioxin

 

estrogenic

     

PCBsbc

PCBsbc

PCBsbc

PCBsbc

EUC exposured

-.02

-.01

-.03

-.03

-.02

-.02

-.03

-.03

 

(−.13, .10)

(−.13, .10)

(−.11, .04)

(−.09, .04)

(−.11, .07)

(−.14, .10)

(−.14, .09)

(−.11, .06)

Age

-.00

-.00

-.00

-.00

-.00

-.00

-.00

-.00

 

(−.01, .01)

(−.01, .01)

(−.01, .01)

(−.01, .01)

(−.01, .01)

(−.01, .01)

(−.01, .01)

(−.01, .01)

BMI Group

.37

.39

.36

.36

.36

.37

.37

.36

 

(.19, .55)

(.22, .56)

(.18, .54)

(.18, .54)

(.18, .55)

(.19, .55)

(.19, .55)

(.18, .54)

Lipids

.04

n/a

.04

.03

.04

.04

.04

.04

 

(−.05, .13)

 

(−.05, .13)

(−.05, .12)

(−.05, .13)

(−.05, .13)

(−.05, .13)

(−.05, .13)

  1. aAnalyses excluded men without a diagnosis of diabetes who did not have insulin or glucose measurements. Multivariable regression models included variables listed in 1st column.
  2. bvariable not normally distributed, therefore natural log-transformed variable used in regression model.
  3. cPCB is a sum of all PCB congeners measured, <LOD = zero; Dioxin-like PCBs include: 105, 118, 156, 157, 167, 189; Non dioxin-like PCBs: sum of all other PCBs beside dioxin-like; Estrogenic PCBs: sum of 52, 99, 101, 110, 153; Anti-estrogenic PCBs: sum of 105, 156.
  4. dEUC Exposure refers to the exposure variable listed in the column headings of the table.