Is coffee good for you?
This question, while seemingly innocuous is something that truly highlights the importance of scientific literacy in society. Within scientific circles, meanings of words are sometimes different in the weight they hold when compared to everyday use. Case in point, “theory” holds a lot of weight in science but much less in regular use.
So whenever news organizations write articles on the benefits or detriments of (insert product here), the truth is usually lost behind an attention grabbing headline like this one from the Washington Post:
Let’s start by talking about scientific studies.
Not all studies are created equal. There is a definite hierarchy on the strength of a study design even before getting to the actual content of a study.
The gold standard of scientific studies is the Randomized Double-Blind Controlled Studies. Usually this study design compares the effectiveness of an intervention between two groups. One group receives the intervention, and one group receives a placebo. Neither the participants nor the researchers know who is getting which. The idea is that the results at the end will most accurate demonstrate the effect of the intervention (if any).
Unfortunately, this kind of study design is not always possible. Consider for a moment that a surgeon wants to test doing a specific procedure is better than giving a medication. It’s hardly possible to ensure participants are unable to tell that they have had surgery.
Therefore, there are a number of additional study designs that can introduce more bias, but are a better way to test a theory. Now the question arises: Well if these studies aren’t as good, how can we believe them?
This is where Meta-Analysis comes in. A meta-analysis is a special study design that doesn’t actually investigate anything directly. Instead it is a kind of statistical analysis that compares the results of multiple studies that have already been conducted. This way you can get a better picture of the overall effect of an intervention. To oversimplify this concept, imagine that you have 10 scientists studying coffee. 5 say coffee increases your risk of cancer by 10x, 3 say it increases by 2x and 2 say it decreases cancer by 4x. A meta-analysis would look at this data and say that overall, coffee increases the risk of cancer by 4.8x (note: these are all made-up numbers).
Is there anything better than this you ask? Why yes, there is. A meta-analysis of multiple meta-analyses can really dig down into the data and give you the clearest picture of cause and effect.
One more thing to know. Numbers matter. What I mean by this is that the number of studies showing an effect increases the likelihood that the effect is true. When talking heads on cable news say things like “just find a study that fits your beliefs” (yes I have actually heard this), that is BIG issue. The reason we consider something to be true in science is that many people have tested it and have all come to the same conclusion. Just because a study exists that argues the opposite doesn’t make it gospel. Studies like these are usually the first to be heavily criticized and if they do not hold up to scrutiny (case in point, Andrew Wakefield’s anti-vaccine study) they are thrown out.
Let’s go back to coffee then. Is it good or bad? Does it cause cancer or not?
Starting with cancer. On March 29, 2018 a judge in California ruled that Starbucks and about 90 other companies must label their products with a warning about carcinogens in their products. This is because the substance, acrylamide, is produced when coffee beans are being roasted. Acrylamide is found in a number of places including cigarette smoke, fried starchy foods and some adhesives.
There are two types of studies examining the risk of cancer (different types) due to acylamide exposure. The first is in animals, specifically rats and mice. Researchers exposed these animals to 1,000-10,000 times the amount people would normally find in food and found it increased their risk of cancer.
The second is in humans. The evidence supporting acrylamide causing cancer is woefully poor in this setting. these studies rely heavily on questionnaires exploring peoples’ diets over years of follow-up. Human memory, while miraculous, is anything but infallible. So currently, acrylamide is classified as “probably causing cancer” based on the animal studies.
Well what does that mean for your cup of joe? Really, nothing. Remember, it takes extremely high doses for a prolonged period of time to cause cancer in rats. Human physiology for one is different from rats and we are not ingesting that quantity of acrylamide on a regular basis to warrant a cancer scare.
So coffee is good?
Coffee is coffee. Nothing is good in excess. However an umbrella review of meta-analyses on the subject published in 2017 demonstrated that coffee drinking was associated more with benefits than harms. This study is incredibly comprehensive and gives the strongest credence to the conclusion that overall coffee drinking is generally safe.
Coffee drinkers (3-4 cups per day) are more likely to have lower overall mortality, lower risk of cardiovascular disease, lower risk of cardiovascular-related mortality, multiple kinds of cancer, liver disease and neuropathy. It is important to know however that all of these medical conditions have a multitude of genetic and environmental risk factors and while coffee may not increase your risk, it doesn’t magically protect you from it all.
Bottom line? It’s important to try to get the source information as much as possible. Be critical of both scientific studies but especially on websites that report on it (even this one).