Variables with a p-value <0.2 in univariate analysis were included in an initial step for multivariate modelling, followed by stepwise backward removal of all variables with a p-value >0.05 to obtain the final model.
This paper seeks to overcome these problems by using a multivariate VAR model followed by the Toda and Yamamoto (1995) and Yamada and Toda (1998) method which is expected to improve the standard F-statistics in the causality test procedure.
In the multivariate models, we followed the recommendation of Ledikwe and colleagues to include a variety of calculation methods as well as an EDF-only calculation (Ledikwe et al, 2005).
Expressions of these genes were evaluated for their effect on patient survival first in univariate models, followed by multivariate models with TP53 status as a covariate.
The associations between the risk of overweight plus obesity and the following factors in multivariate models were as follows: (1) the risk increased by approximately 1.2-fold for both genders with family history of obesity.
Generalized mixed effect models followed by a model averaging procedure were applied on weed metrics and permutational multivariate analyses of variance were applied on weed species composition.
Multivariate model included: type of follow-up, regimen, age (categories), patient referral reason, and place of birth.
In a multivariate model, only duration of follow-up was a statistically significant predictive factor (OR 1.30, 95%CI 1.12 to 1.52, P < 0.001).
After simultaneous control for risk factors, BMI was the primary contributor of the variation in multivariate models (P < 0.001), followed by age and WBC count (P < 0.001), and family history of diabetes and triglyceride levels (P = 0.12).
In a multivariate model including age and BMI at follow-up and renal function after 1 year, a higher Pt AUC1 3 years was associated with an increased risk for persisting paraesthesia [odds ratio (OR) = 1.07 (95% CI 1.00 1.13), P = 0.043; supplementary Table S5, available at Annals of Oncology online].
