Dependent variable | Predicting variable | Final model of multiple linear regression B (95% CI) | P value |
Halo count | Age | – | |
Sex | 1.109 (0.172 to 2.047)* | 0.021 | |
Ocular ischaemia | 1.103 (0.089 to 2.116)* | 0.034 | |
Polymyalgia | – | ||
Two or more systemic symptoms | – | ||
Temporal artery palpable changes | – | ||
Halo Score | Age | – | |
Sex | 2.902 (0.100 to 6.984)† | 0.041 | |
Ocular ischaemia | 3.488 (0.305 to 8.143)† | 0.028 | |
Polymyalgia | 2.813 (−0.053 to 7.080)† | 0.056 | |
Two or more systemic symptoms | – | ||
Temporal artery palpable changes | – |
Data are shown for baseline halo count and Halo Scores in patients with GCA (n=58). Multiple linear regression analysis was performed with backward exclusion of predicting variables. Since the Halo Score was not normally distributed, the Halo Score was transformed by square root. The probability of F for removal was 0.10. Results of the final model are shown. Age in years. Sex: 0=female, 1=male. Ocular ischaemia (ie, anterior ischaemic optic neuropathy, posterior ischaemic optic neuropathy and/or relative afferent pupillary defect), polymyalgia, two or more systemic symptoms (ie, anorexia, fever/night sweats, fatigue), temporal artery palpable changes (ie, thickening and/or loss of pulse): 0=absent, 1=present. (−) Variable removed due to backward exclusion.
*R2=0.157, F(2,55) = 5.138, p=0.009.
†R2=0.207, F(3,54) = 4.688, p=0.006.
GCA, giant cell arteritis.