To address potentially confounding factors, we ran a multivariate model, and variables were selected after bivariate analysis using a Cox PH model for the variables described in Table 1.
For this reason, a mediation model among the variables described before was chosen.
In a multivariable model incorporating the variables described in the "Covariates" section of the Methods, the adjusted odds ratio was 1.21 (95% CI, 1.11 1.32).
If the model lacks variables describing the existing selection process estimates of the parameters will be biased (or alternatively speaking, affected by omitted variables that describes selection process).
It also found that, compared with these Quebec Task Force classified subgroups, a more detailed predictive model with variables describing pain location, pain intensity and number of neurological signs, had somewhat greater prognostic capacity, although that more detailed model still did not have sufficient prognostic accuracy for us to consider it clinically useful in this setting.
Other model variables not described above included race/ethnicity [based on two questions: "Are you Hispanic or (Latino/Latina)?" and "Which of the following would you say is your race? You may select more than one category".]; injured on 9/11 (based on the question: "On 9/11, did you have any of the following injuries as a result of World Trade Center terrorist attack?").
Table 5 reports the results for model 1 with the variables described here.
A subsequent model included all the variables described plus dietary and alcohol intake information.
To analyse the effect of body condition on swelling size and shape, the MUAF index was similarly introduced as a fixed factor into a model containing the control variables described above (and in the absence of genetic parameters).
Finally, the fully adjusted model included all the variables described, plus dietary variables (intake of meat, dairy, seafood and total protein), and alcohol intake information (intake of beer, wine and liquor) as these have been described to affect serum urate levels [ 22- 24].
