Skip to main content

Table 4 Results of explanatory multivariate analysis: SI (log) models (I)

From: Dioxin-like compounds and bone quality in Cree women of Eastern James Bay (Canada): a cross-sectional study

SI (log) model

Main exposure variable

 

DLCs (pg TCDD-EQ/L)

DL-PCB 105 (μg/L)

DL-PCB 118 (μg/L)

DL-PCB 156 (μg/L)

N

245

 

246

 

244

 

203

 

R 2 (Adjusted R 2 )

0.3935 (0.3675)

 

0.3631 (0.3471)

 

0.3932 (0.3725)

 

0.3766 (0.3475)

 

Regression coefficient (SE) (a)

p-value

Regression coefficient (SE) (a)

p-value

Regression coefficient (SE) (a)

p-value

Regression coefficient (SE) (a)

p-value

Variables

DLCs

0.3806

DL-PCB 105

0.0414

DL-PCB 118

0.0483

DL-PCB 156

0.1379

 

0.07 (0.08)

 

−219.24 (106.90)

 

−37.84 (19.07)

 

−69.81 (46.87)

Age (years)

−13.87 (2.58)

<0.0001

−12.74 (1.49)

<0.0001

−12.55 (1.57)

<0.0001

−12.21 (1.76)

<0.0001

Weight (kg)

0.11 (0.75)

0.8815

NR

 

NR

 

NR

 

Height (cm)

−2.09 (2.36)

0.3762

−0.75 (2.18)

0.7321

−2.38 (2.23)

0.2860

−2.60 (2.50)

0.2994

Number of children

NR

 

6.22 (5.09)

0.2232

4.71 (5.00)

0.3477

25.2 (5.17)

0.6264

Omega-3 PUFAs (%)

18.36 (9.89)

0.0646

NR

 

23.90 (9.94)

0.0170

17.64 (11.36)

0.1221

Vitamin D (nmol/L)

0.57 (0.72)

0.4298

NR

 

NR

 

NR

 

Mercury (nmol/L)

NR

 

0.22 (0.25)

0.3769

NR

 

0.31 (0.31)

0.3215

Glucose (mmol/L)

NR

 

NR

 

2.88 (4.18)

0.4914

NR

 

Menopausal status

−26.05 (40.99)

0.5256

NR

 

NR

 

NR

 

Level of education

95.19 (31.70)

0.0030

NR

 

71.25 (31.37)

0.0240

72.72 (34.94)

0.0388

Smoking status

−36.58 (27.40)

0.1831

NR

 

NR

 

NR

 

Supplements use(b)

NR

 

NR

 

NR

 

−12.67 (43.91)

0.7733

Total lipid(c) (g/L)

−5.28 (9.51)

0.5796

7.31 (10.41)

0.4829

3.04 (10.38)

0.7696

7.19 (11.28)

0.5244

  1. a Regression coefficients and standard errors values are multiplied by 103;
  2. b Calcium and vitamin D supplements;
  3. c Final adjustment for the total plasma lipid concentration;
  4. NR – Variable not retained in the model (non-confounding factor; proportional change of the regression coefficient of the main exposure variable < 10%).