Which cancers does alcohol play a role in?
Alcohol increases the risk of several types of cancer, most strongly for mouth, throat, oesophageal and liver cancer. Even at moderate consumption levels the risk of breast and colorectal cancer can rise, so drinking less is always the safer choice.
Alcohol has the strongest association with cancers of the mouth, throat and oesophagus. The risk increases the more you drink, and it exists independently of smoking. Alcohol is regarded as one of the most important causal factors in oesophageal cancer. Liver cancer also falls into the strong category. Years of heavy drinking frequently leads first to liver cirrhosis, and that cirrhosis forms an intermediate step towards liver cancer. In this process, alcohol acts as a kind of accelerator of cell growth in an already damaged liver. For colorectal cancer (colon and rectum) and breast cancer the evidence is less strong, but there are clear indications that alcohol raises the risk. Notably, that risk can be present at relatively low levels of daily consumption. In breast cancer, a likely contributing factor is that alcohol raises levels of the female hormone oestradiol. Alcohol acts as a cancer promoter through several pathways simultaneously. The substance your liver produces when breaking down alcohol directly damages DNA and inhibits its repair. In addition, free radicals can damage cells, certain liver enzymes are activated so that the body absorbs carcinogenic substances more readily, and deficiencies in folate and vitamin A arise that weaken cellular protection. For pancreatic cancer there are weak, indirect indications, but the evidence is not convincing. For lung cancer, bladder cancer, prostate cancer, stomach cancer, ovarian cancer, uterine cancer and melanoma, epidemiological research shows no increased risk from alcohol.
This answer is based on several large review studies that summarise the epidemiological and experimental research on alcohol and cancer. The associations with mouth, throat, oesophageal and liver cancer are broadly supported and have been judged as causal by the researchers. The associations with colorectal and breast cancer are plausible but less firmly established. For pancreatic cancer the supporting evidence is thin. The researchers base their conclusions predominantly on population studies; where the question of cause and effect is not fully resolved, this is noted in the answer.