The comparison of demographic, clinical features and the serum biomarker levels of the tophi-free group (76.47%) and the tophi group (23.53%) have been summarized in Table 1. The prevalence of carotid plaque was 65% in tophi group and 45.9% in tophi-free group. Mean cIMT in tophi group (1.15 ± 0.16) was higher than in tophi-free group (1.02 ± 0.22). In addition, compared with patients with tophi-free, those with tophi had a higher proportion of male and longer gout duration, higher levels of BMI, TC, TG, GFR, CRP, serum urate, F-INS, HOMA-IR and Gutt index than the patients with tophi-free.
The correlation between cIMT and several clinical characteristics in the whole study population was listed by bivariate linear regression analysis in Table 2. Carotid IMT showed a positive correlation with gout duration, BMI, tophi, serum urate, FPG, F-INS, TC, TG, FFA, HOMA-IR and trends of negative correlation with HDL-C and Gutt index (P < 0.05).
Variables showing statistically significant correlations in linear regression analysis were included in univariate logistic regression model in order to define the predictors of carotid atherosclerosis. Then, significant indicators were included in multivariate regression models.Table 3 lists the results of the both univariate and multivariate analysis showing that long gout duration, tophi, high levels of serum urate, TC, HOMA-IR, and Gutt index were independent risk factors for abnormal cIMT (defined as IMT >0.9m) or atheroma plaques. Logistic regression model was performed to simultaneously show the results of different adjusted models to evaluate the independent association between tophi and abnormal cIMT or atheroma plaques. As shown in Table 4, after adjustment for age, gender, current smoking, alcohol consumption, BMI and some cardiovascular risk factors, patients with tophi were associated with a 3.943-fold (95% CI 2.56-6.011, P < 0.001) higher prevalence of abnormal IMT and a 3.655-fold (95% CI 1.482-9.01, P = 0.005) higher prevalence of atheroma plaques than patients with tophi-free.
An interaction analysis was performed to explore the potential interaction between tophi and IR for increased carotid atherosclerosis risk in Table 5. The interaction between tophi and HOMA-IR had a statistically significant effect on abnormal cIMT (OR = 1.444, 95% CI: 1.065-1.698) and atheroma plaques (OR = 2.493, 95% CI: 1.806-3.443). Statistically significant interaction was also observed between tophi and the reciprocals of Gutt index on abnormal cIMT (OR = 2.602, 95% CI: 1.694-3.58, P =0.01) and on atheroma plaques (OR = 4.005, 95% CI: 2.168-11.78, P =0.001).
As shown in Figure 1, P values for the interaction were not significant in subgroup analyses for sex, age, BMI, eGFR, serum urate, smoking, drinking and kidney stones (P=0.159, P=0.115, P=0.7, P=0.169, P=0.099, P=0.325, P=0.24 and P=0.54, respectively). In different subgroups, there was no difference in the risk of tophi on carotid atherosclerosis.