Most of the sporadic colorectal cancer (CRC )develop from colorectal adenoma (CRA), patients with CRA have a high risk of recurrence and development of metachronous CRA or CRC after removal, therefore, the investigators conducted this clinical trial to explore the chemoprevetion effect of metformin for CRA recurrence after removal.

Colorectal adenomas are well-known to be precancerous lesions that develop into colorectal cancers on the basis of the adenoma-carcinoma sequence. The effects of screening for colorectal adenomas and removing precancerous lesions on the prevention of colorectal cancer have been established. Because of the high recurrence rates of colorectal adenomas in patients who have undergone polypectomy, the potential chemopreventive agents that may reduce the risk of colorectal adenoma recurrence need to be investigated.

Metformin is a widely used diabetes medicine. In recent years, anticancer activity of metformin has been explored. The aim of this study is to investigate the effect of metformin on the recurrence of colorectal adenomas by conducting a randomized, placebo-controlled, prospective clinical.

Inclusion Criteria:

1. Aged 40-80 years without diabetes;
2. CRAs removed without recurrence before recruitment;
3. Must sign the consent form after being fully informed and understanding the purpose and procedure of this study.

Exclusion Criteria:

1. Familial adenomatous polyposis (FAP) or hereditary non-polyposis colorectal cancer (HNPCC, Lynch syndrome);
2. Regularly taking aspirin, nonsteroidal anti-inflammatory drugs (NSAIDs), cyclo-oxygenase 2 (COX2) inhibitors, calcium or vitamin D;
3. Have a history of gastrointestinal surgery;
4. Have severe heart, liver or kidney disease ;
5. Have cancer history;
6. Women with pregnant, during breast-feeding period, or with expect pregnancy;
7. Diabetes（taking diabetes medicines or HbA1c>6.5%);
8. Inflammatory bowel disease;
9. Mental illness;
10. Intolerant to metformin;
11. Cannot tolerate colonoscopy;
12. Staffs in this clinical trial;
13. Poor bowel preparation for colonoscopy or the examination time is shorter than 6 minutes;
14. Unsuitable for inclusion by the investigator

The participant, investigator, data analyst, endoscopist and pathologist involved in this study are all blinded of which treatment the participants are receiving. The pharmaceutical company provided blinded bottles of study pills with the label 'A' , 'B' or 'C'. Unblinding will be advanced only in case of emergency.