Skip to main content

Advertisement

Table 3 Adjusted association analysis for incident gout by risk factor using three different approaches

From: Untangling the complex relationships between incident gout risk, serum urate, and its comorbidities

  Two-factors   Full   Stepwise regression    
  regressiona   regressionb   (AIC)c   (BIC)c  
  Est. (OR) p Value Est. (OR) p Value Est. (OR) p Value Est. (OR) p Value
Serum urate (mg/dl) - - .792 (2.21) <.001 .805 (2.24) <.001 .795 (2.21) <.001
Sex: Male -.121 (0.89) 0.424 .030 (1.03) 0.854 - - - -
Ethnicity: AA .674 (1.96) <.001 .647 (1.91) <.001 .622 (1.86) <.001 .675 (1.91) <.001
Age, years -.001 (1.00) 0.959 .000 (1.00) 0.974 - - - -
Glucose: High .080 (1.08) 0.750 -.055 (0.95) 0.830 - - - -
HDL: Low -.148 (0.86) 0.350 -.046 (0.95) 0.796 - - - -
LDL: High -.354 (0.70) 0.015 -.381 (0.68) 0.010 -.373 (0.69) 0.011 - -
Triglycerides: High -.057 (0.94) 0.710 .093 (1.10) 0.579 - - - -
SBP: Hypertensive .580 (1.79) 0.002 .510 (1.66) 0.006 .508 (1.66) 0.005 - -
BMI: Obese .054 (1.06) 0.723 -.044 (0.96) 0.786 - - - -
eGFR: Low .364 (1.44) 0.246 .042 (1.04) 0.200 - - - -
  1. aLogistic regression of gout on two predictors: serum urate plus one of the factors in rows
  2. bLogistic regression of gout all the factors listed in rows.
  3. cStepwise logistic regression. Rows with no results correspond to predictors that did not entered in the final model. p Values correspond to estimated coefficients. Bold indicates effect estimates that were statistically different from zero at 0.01 significance level