These include such factors as: (i) Adjustments for smoking may produce an apparent "negative confounding" effect that biases E-R trends because current smoking was not associated with DE exposure, and therefore was not a confounder (Silverman et al., 2012 ; Attfield et al., 2012 ).
We used three levels of adjustment: (i) adjusted for age, population and attending examination in a clinic (Model 1); (ii) adjusted for age, population, attending examination in a clinic and socioeconomic status indicators (Model 2) and (iii) adjusted for age, population, attending examination in a clinic, socioeconomic status indicators and health behaviours (Model 3).
Model I included adjustment for field center or the variables used in estimating the 90th percentiles for birth weight and percent body fat for gestational age (sex, ethnicity, center, and parity).
Indicator variables for specific hospitals were added to the model to estimate odds of surfactant use for each hospital relative to the hospital with the median rate of surfactant use (hospital I), after adjustment for patient characteristics.
The findings from this small prospective study are consistent with a moderate increase in breast cancer risk associated with high levels of IGF-I, after adjustment for IGFBP-3.
We express the full model at marker locus j as y i = ∑ k x i k a k + ∑ s = 1 36 p i j (s ) β s + γ i + ε i, (1 where y i is the phenotype for animal i, x ik is an indicator for covariate k and animal i, a k is the effect of covariate k, p ij s) is the diplotype probability at a marker locus j for animal i generated by the HMM, γ i is an adjustment for kinship, and ε i is the residual error.
A bicluster (I, J) follows an additive model if a ij = c+ α i + β j + η ij, where c is the typical value within the bicluster, α i is the adjustment for row i∈ I, β j is the adjustment for column j∈ J and η ij is the noise associated with the element.
Maternal spontaneous IFN-γ response was significantly associated with child cytokine response (OR 2.9 95%CI 1.3 6.6)), after adjustment for residence (OR 0.4 95%CI 0.2 0.9)) and educational level (OR 0.4 95%CI 0.2 0.9)).
Because the purpose of the analyses was to detect a potential relationship between changes in certain efficacy and safety variables and 3 different study drug doses, no type I error adjustments for multiplicity were performed.
Model I includes adjustments for child's sex; Model II includes adjustments for child's sex, gestational age, admission to neonatal ward during the first ten days, multiple pregnancy, mother's smoking during pregnancy, parental education, maternal age and maternal height.
