In addition, we conducted a sensitivity analysis to examine the potential for confounding due to co-exposure to multiple phthalates by repeating regression analysis while adjusting for both molar sum DEHP and molar sum DBPs.
We repeated the univariate and multivariate logistic regression analysis, modelling an increase in serum P versus no change/decrease.
Repeated measures regression analysis evaluated a change in the Respiratory Assessment Change Score (RACS).
On repeated measure regression analysis, the cardiac index was on an average 0.50 L/min/m2 higher in the DN group than in the STH group at any time point (P = 0.002).
Linear repeated measures regression analysis was used to test the null hypothesis that there was no relationship between coral cover and time from 1968 to 2004.
We used repeated measures regression analysis based on the individual trajectories of subregions or reefs (for the analysis of the reef monitoring data) to quantitatively estimate the absolute net decline of coral cover.
To test whether malaria infection had any effect on anti-meningococcus antibody levels, we repeated the regression analysis with seropositivity to P. falciparum, defined as a detectable response to any of the three antigens tested, as a factor.
The annual loss based on repeated measures regression analysis of a subset of reefs that were monitored for multiple years from 1997 to 2004 was 0.72 % (n = 476 reefs, 95% CI: 0.36, 1.08).
Eleven infants seroconverted for EBV between 9 and 11 months, and six seroconverted for CMV though none seroconverted for both To establish whether the difference between EBV-infected and uninfected infants was due to the effect of infection between the time of vaccination and sampling, we repeated the regression analysis excluding any infants that seroconverted between the two sampling times.
In order to test whether the results could have been biased by failure to classify infants that were actually uninfected, we repeated the regression analysis with all unclassified infants as uninfected, and found that the associations still remained significant (data not shown).
