Clinical Question: Do calcium supplements prevent advanced colon polyps in patients who already have polyps?

Setting: Outpatient (any)

Study Design: Randomized controlled trial (double-blinded)

Allocation to Groups: Uncertain

Synopsis: Patients recently diagnosed with colon polyps were assigned randomly to receive 1,200 mg of calcium carbonate (n = 454) or placebo (n = 459). The patients underwent additional colonoscopies at one and four years after randomization to determine whether the polyps had recurred. Although the authors did not describe whether the allocation was concealed or if the pathologists grading recurrent polyps were blinded to the intervention, the main outcome of polyp recurrence was assessed via intention to treat.

The patients who received calcium had fewer hyperplastic polyps, tubular adenomas, and advanced adenomas, although these findings were not statistically significant. Advanced neoplasms occurred in 44 patients (10 percent) who received calcium and in 68 patients (15 percent) who received placebo (number needed to treat, 20; 95 percent confidence interval [CI], 11 to 117). Advanced neoplasms were defined as tubulovillous or villous adenoma, carcinoma in situ, or invasive cancer.

Given the mixed protective effects, three different conclusions could be drawn from this study: (1) calcium carbonate protects against developing polyps (although the study lacked sufficient power to detect this); (2) calcium carbonate protects against development of high-risk neoplasms; or (3) calcium carbonate has no effect on polyps and the findings in this study were a random event. Because other studies also have demonstrated a protective effect, the third conclusion is less plausible.

Bottom Line: Calcium carbonate in a dosage of 1,200 mg daily, given to patients with colon polyps, protects against the development of histologically advanced polyps. On average, 20 patients would need to be treated (95 percent CI, 11 to 117) to prevent one advanced polyp. Although this study included no patient-oriented outcomes, the intervention is cheap and innocuous, causing only a little gas and bloating. (Level of Evidence: 2b)