Findings from the meta-analysis:
Twelve studies with seventeen effect sizes evaluated soy isoflavones effects on serum CRP concentrations (Figure 2 ). The overall effect suggested a non-significant effect in serum CRP concentrations after supplementation with soy isoflavones compared with comparison group (WMD= 0.08 mg/L, 95% CI: -0.08, 0.24; p=0.302). There was a substantial heterogeneity between studies (Cochrane’s Q test, p < 0.001; I2= 87.0%). Sensitivity analysis showed that excluding each trials from the overall analysis did not reveal a significant change in overall effect size of soy isoflavones on CRP concentration. Subgroup analysis based on all studied variables could not diminish the heterogeneity between studies (Table 3 ). Subgroup analyses revealed no significant change in circulatory levels of CRP following soy isoflavones intake compared with comparison group in any of the subgroups.
The overall effect of soy isoflavones plus soy protein on serum CRP is shown in Figure 3 . Eleven studies with sixteen effect sizes assessed the effect of soy protein and soy isoflavones combination on serum CRP levels. The overall estimates indicated non-significant effect in serum levels of CRP following soy isoflavones plus soy protein intake in comparison with control group (WMD= -0.02 mg/L 95% CI: -0.12, 0.08; p=0.715) with high heterogeneity (Cochrane’s Q test, p <0.001, I2= 84.9%). Sensitivity analysis showed that excluding each trials from the overall analysis did not reveal a significant change in overall effect size of soy isoflavones plus soy protein on CRP concentration. The between-group heterogeneity was not significant in trials using soy isoflavones >100 mg/d (Cochrane’s Q test, p=0.600, I2=0.00%), in trials with parallel design (Cochrane’s Q test, p=0.074, I2=53.7%), in trials performed with study duration>56 day (Cochrane’s Q test, p=0.074, I2=47.9%) and in trials conducted in healthy people (Cochrane’s Q test, p=0.330, I2=13.2) (Table 4 ). In subgroup analysis based on BMI indicated that soy isoflavones plus soy protein could increase CRP levels in people with BMI≤26 (WMD= 0.36 mg/L 95% CI: 0.01, 0.72; p=0.045, I2=65.8%), but a non-significant reduction was observed in people with BMI>26 (WMD= -0.08 mg/L 95% CI: -0.18, 0.02; p=0.121, I2=91.3%) (Figure 4 ). However, in subgroup analysis revealed no significant change in circulatory levels of CRP following soy isoflavones plus soy protein intake compared with comparison group in other subgroups (Table 4 ).