Coffee drinking and cancer risk
Today I’ve read an article about coffee drinking and cancer risk. The paper’s source can be found at the bottom of this article.
While looking at scientific papers related to coffee consumption and health, I was surprised by how many articles there are on the subject. Usually, studies focus on the relationship between coffee drinking and a specific health issue. But today’s paper is different.
Today’s article is an “Umbrella review”, which means, in the context of medical research, a review of reviews. Such an approach makes a lot of sense: we gather all existing studies related to a subject, we clean the dataset by removing the “bad” articles (in relation to what we are interested in), then we look at the different results and see if they all point to the same direction. Also, “Umbrella reviews” make it easier to dig into a complicated topic by summarizing the state-of-the-art into a sometimes short and concise article.
Today’s authors were interested in the relationships between coffee drinking and cancer risk. By this, they aimed at looking at the whole literature on the topic and see the literature was consistent in finding positive or negative correlations between coffee consumption and a variety of cancers including liver cancer, leukemia, lung cancer, etc.
The first thing to look at in this kind of paper is how did the authors manage to gather sources and how did they decide to filter some of them out. Hopefully for us, in our case, the methodology is clearly explained and associated with a nice and comprehensive figure.
They started by searching for all possible papers related to the topic from PubMed, Embase, Web of Science, and Cochrane Library. They ended up with 503 articles. They then removed duplicated titles and abstracts and went down to 237 articles. They then applied the following filters: excluding articles that were not meta-analysis or systematic reviews, excluding articles that were not specifically studying coffee intake and cancer incidence, excluding articles with no data synthesis or which were containing missing data. Thanks to these exhaustive filters, the authors identified 28 articles studying 36 coffee-cancer associations.
Then, they categorized each coffee-cancer association into 5 groups according to the evidence level: strong, highly suggestive, suggestive, weak, and no association. Each category is ruled by a certain number of cases, a p-value threshold, and others statistics variables focused on results reliability. They evaluated each coffee-cancer association by the sign of the risk: “Decreased risk” if the correlation was negative and “Increased risk” if the correlation was positive.
The authors found 1 coffee-cancer association exhibiting strong evidence: coffee drinking increases risks of acute lymphocytic leukemia.
They found 5 highly suggestive evidence: coffee drinking seems to decrease risks of endometrial cancer, liver cancer, melanoma, and oral/pharyngeal cancer and increases risks of bladder cancer. They also found 1 suggestive evidence: the coffee seems to increase the risks of lung cancer.
All other associations, wether they were negative of positively correlated showed weak evidence or no evidence, this includes breast cancer, colon cancer, colorectal cancer, esophageal cancer, nonmelanoma, leukemia, pancreatic cancer, ovarian cancer, acute myelogenous leukemia, rectal cancer, kidney cancer, laryngeal cancer, thyroid cancer, prostate cancer, gastric cancer, lymphoma, glioma, and biliary tract cancer.
What can we learn from this article? First, Umbrella reviews are an elegant way to get a better insight into the state-of-the-art of a specific topic. Second, there is strong and highly suggestive evidence of associations between coffee drinking and both decreased and increased risks of cancer. That being said, most cases point to a decreased risk of cancer for coffee drinkers.
Source: Zhao, L.G., Li, Z.Y., Feng, G.S., Ji, X.W., Tan, Y.T., Li, H.L., Gunter, M.J. and Xiang, Y.B., 2020. Coffee drinking and cancer risk: an umbrella review of meta-analyses of observational studies. BMC cancer, 20(1), p.101.
