By Dorian Grayson
How many studies about coffee have you read about? As a social institution, science is trusted to be humanity’s foremost truth-seeking endeavor. When a new study comes out, outlets often lean on science’s credibility and take the conclusion of the study as a truth to report on, but that fundamentally misunderstands the purpose of these studies and the murkiness of scientific truth.
If you read a ‘scientifically backed’ productivity article about coffee, they will likely tell you that caffeine has a half-life of six hours, because that translates nicely into fractions of a 24 hour day. If you google caffeine’s half-life, Google’s automated card system will tell you, without a source, that it is five to six hours.
If you read “Health aspects of caffeine: benefits and risks,” an article from “Nursing Standard” about the general health effects of caffeine gleaned by over 40 studies, you’ll find that it is generally 2.5 to 4.5 hours, but has been known to be as extreme as one to 10 hours, depending on the person ingesting it. This evidence-based range gets ignored in a news cycle of sensationalizing individual studies for whole number “facts.”
I would be shocked if you haven’t seen an article with the words “coffee” and “cancer” in the title. These are usually clickbait articles from writers who only read the abstract and conclusion of a study. A better perspective of the scientific truth of the subject can be taken from “Coffee consumption and risk of cancers,” which examined 59 studies to find both the relationship of coffee to general cancer risk and to specific types of cancer.
For example, though coffee has been associated with reduced risk of multiple cancers, it has not been associated with reduced risk of stomach, lung, kidney or ovarian cancers. It has, however, been “associated with a reduced risk of bladder, breast, buccal, leukemic, pancreatic and prostate cancers,” among others. In general, the study found that an additional cup of coffee daily is associated with a three percent reduction in general cancer risk. Notably, these are only associations, as correlation is not causation.
An individual study can tell a very different story than a scientific literature review, especially if the study is poorly designed. A 2013 study published in “Applied Nursing Research” looked to test the sleep effects of coffee consumed within six hours of sleep. It found no benefits to avoiding coffee before bed.
A study that contradicts common wisdom and encourages more coffee consumption? Who doesn’t want to just take that as scientific truth and leave? Unfortunately, the details matter. The study involved 10 university students over two weeks, with the caffeine ingested before bed being one cup of coffee at dinner. The data set was laughably small and not sufficient to generalize the findings.
This study was cited in “Coffee, caffeine, and sleep,” a meta-analysis of 58 studies and trials. Within the meta-analysis, they compared different trials, including the 10 university students, and explained that, on top of the lack of data, the subjective measures they used to analyze sleep quality fail to capture the way caffeine affects sleep. The study found that caffeine generally results in a shallower sleep, postponing and shortening the “deeper” stages of sleep, when consumed within 16 hours of sleeping. No, that’s not a typo. If you’re going to bed at midnight, the science suggests you should stop drinking coffee after 8:00 a.m. if you want a full, deep rest.
That, as with all scientific truth, is subject to change. No amount of science or testing can give us The Truth, because that’s not the aim. It’s constructing the best approximate picture of general truth with a rigorous process. Rarely does science “prove” anything. Rather, it points us towards the truth, helping larger studies even if they’re wrong.