Table 4.
Model 1b | Model 2c | Model 3d | ||
---|---|---|---|---|
Chlordecone in blood at 7 years (µg/L) | N | β (95% CI) | β (95% CI) | β (95% CI) |
Boys | 221/283a | |||
< 0.017 | 53 | 0 (ref) | 0 (ref) | 0 (ref) |
0.017—0.051 | 51 | 0.20 (-0.35; 0.74) | 0.12 (-0.36; 0.60) | 0.16 (-0.39; 0.70) |
0.051—0.112 | 62 | -0.22 (-0.73; 0.30) | -0.48 (-0.95; -0.01) | -0.16 (-0.68; 0.36) |
≥ 0.112 | 55 | -0.07 (-0.61; 0.46) | -0.10 (-0.58; 0.38) | -0.01 (-0.56; 0.53) |
log10 | 221 | -0.07 (-0,.42; 0.28) | -0,09 (-0.38; 0.20) | -0.08 (-0.43; 0.27) |
Girls | 225/292a | |||
< 0.017 | 60 | 0 (ref) | 0 (ref) | 0 (ref) |
0.017—0.051 | 59 | 0.20 (-0.23; 0.62) | 0.28 (-0.12; 0.69) | 0.32 (-0.13; 0.77) |
0.051—0.112 | 49 | 0.06 (-0.40; 0.51) | 0.01 (-0.41; 0.44) | 0.01 (-0.45; 0.47) |
≥ 0.112 | 57 | -0.10 (-0.54; 0.33) | -0.12 (-0.54; 0.30) | 0.01 (-0.44; 0.46) |
log10 | 225 | -0.001 (-0.29; 0.29) | -0,07 (-0.36; 0.22) | 0.01 (-0.28; 0.30) |
Adiposity is a latent variable defined from the BMI z-score, % Fat Mass, Sum of subscapular and tricipital folds, and Waist-to-Height Ratio, expressed in the same units as the BMI z-scores
Indices of fit (Model 1, exposure in quartiles):
Boys: χ2 p = 0.02; RMSEA = 0.04 95% CI, 0.02–0.05; CFI = 0.98; GFI = 1; SRMR = 0.01
Girls: χ2 p = 0.06; RMSEA = 0.03 95% CI, 0.00–0.05; CFI = 0.99; GFI = 1; SRMR = 0.01
aMissing blood concentrations at seven years of age accounted for by the FIML estimation
bModel 1 was adjusted for: exact age at measurement (months), maternal place of birth (Guadeloupe / Martinique, other Caribbean Islands, Europe), maternal BMI before pregnancy (< 18.5, 18.5 to < 25, 25 to < 30, ≥ 30 kg/m2), maternal level of education (< 5, 5 to < 12, ≥ 12 years), duration of breastfeeding (four categories), time spent exercising per week, time spent watching TV or playing videogames per week, obesogenic dietary habits at seven years of age (four categories), blood total lipids at 7 years of age (g/L, log10 scale)
cModel 2 is the same as Model 1 but additionally adjusted for blood concentrations of PCB153 and pp’-DDE
dModel 3 is the same as Model 1 but additionally adjusted for cord blood concentrations of chlordecone (N = 131 boys and N = 158 girls with blood samples available at birth (cord blood) and at 7 years)