Results of the logistic regression for the case-case studies.

<p>(A) Odds ratios and their associated P-value for all the selected cases in the univariate models.</p>*<p>Factors with P<0.05 are considered as significant.</p>†<p>Factors with a P<0.25 are entered in the multivariate model.</p><p>(B) Odds ratios and P-values for the final multivariate models. Previous steps, consisting in removing one by one the factors with the highest P-value at each step, are not shown. The program used to execute the analysis gave P = 0.002 for the overall model fit comparing 2002–06 with 1997–01 and P = <0.0001 comparing 2007–10 with 1997–01.</p>a<p>Humans are grouped into four age groups with the reference group being young children (0–4 years) (see <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0079331#pone.0079331.s002" target="_blank">Table S2</a>) and the odds ratio indicates the relative amount by which the odds of the outcome changes when the value of the predictor value is increased by 1.0 unit.</p>b<p>Reference group.