Abstract
BACKGROUND: We aimed to evaluate the associations between calcium and various stages of colorectal carcinogenesis and whether these associations are modified by the calcium to magnesium (Ca:Mg) ratio. METHODS: We tested our hypotheses in the prostate lung, colorectal and ovarian cancer screening trial. RESULTS: Calcium intake did not show a dose-response association with incident adenoma of any size/stage (P-(trend) = 0.17), but followed an inverse trend when restricted to synchronous/advanced adenoma cases (P-(trend) = 0.05). This inverse trend was mainly in participants with Ca:Mg ratios between 1.7 and 2.5 (P-(trend) = 0.05). No significant associations were observed for metachronous adenoma. Calcium intake was inversely associated with CRC (P-(trend) = 0.03); the association was primarily present for distal CRC (P-(trend) = 0.01). The inverse association between calcium and distal CRC was further modified by the Ca:Mg ratio (P-(interaction) < 0.01); significant dose-response associations were found only in participants with a Ca:Mg ratio between 1.7 and 2.5 (P-(trend) = 0.04). No associations for calcium were found in the Ca:Mg ratio above 2.5 or below 1.7. CONCLUSION: Higher calcium intake may be related to reduced risks of incident advanced and/or synchronous adenoma and incident distal CRC among subjects with Ca:Mg intake ratios between 1.7 and 2.5.