A simple way of modeling this association is to regress Y i on all available X ij s (for j = 1, …, n) controlling for the mean vector of covariates, that is: Note that, since this model requires a complete set of predictors X ij, it will use only the subjects having all exposure data at all n visits, a major limitation of such an approach.
Researchers in different fields, during the last thirty years, have proposed approaches for modelling this type of association [ 3- 6].
Only number of hours worked per week was associated with frustration in the full model; this association had borderline statistical significance after adjustment for case-mix.
Furthermore, although a repeated-measures analysis of 62 cardiac rehabilitation patients showed that 120-hr averages of ambient BC were positively associated with resting DBP in single-pollutant models, this association was found to be confounded by PM2.5 (Zanobetti et al. 2004).
However, when age, which was strongly associated with %DFI, simultaneously was included in the model, this association was no longer significant (p = 0.28).
Although exposure to environmental smoking was associated with an increased antibiotic use in the crude model, this association was attenuated and turned non-significant after adjustment for potential confounders.
Although baseline levels of cTnT were associated with reversible myocardial ischaemia in a univariate logistic regression model, this association was attenuated and no longer significant in multivariate analyses.
In porphyry copper models, this association can be correlated with the sericite zone located between the silica-pyrophyllite zone and the potassic zone.
In multifactorial models this association was no longer significant.
After adjustment for all characteristics selected in the precedent models, this association remained significant.
