Figure 1. Interaction effect between vitamin D deficiency and dietary quercetin in predicting risk of prostate cancer.

A) Dietary quercetin was treated as a continuous variable. Lowess smooth curves were fitted and graphed by serum vitamin D deficiency (≤30ng/ml) and normal vitamin D (>30ng/ml) group. B) Multivariable analysis with dichotomized dietary quercetin (low vs high, cut at median 5.8 mg) by vitamin D status. Odds ratios (ORs) were age-adjusted.