Listen to this story



Drinking Coffee Still Won’t Save Your Life

What happens when the media gets a hold of observational research

Image for post
Image for post
Pictured: Tasty. Not lifesaving

Getting a coffee in the morning is something of a magical experience. Walking up to the machine, barista or (god forbid) instant, smelling that rich, dark brew, taking the first caffeine-filled sip, draining the last bittersweet drops from your mug. Finally managing to properly open your eyes as the last fog of sleep is chased away by the sultry brew.

Coffee is great.

And we love it. Worldwide, more than 2.25 billion cups are drunk each day, with the average person in the US drinking more than 100 litres each year.

That’s a lot of coffee

So it’s no surprise that we are interested in the health impacts of coffee. It’s even less surprising that every time a study comes out the suggests that coffee is good for us, we jump on the idea because it’s exciting to think that that morning cup of joe might not just be making our work better it might also be saving our lives.

Enter the newest study making headlines. If the media stories are to be believed, drinking coffee doubles the chances of survival for people who have HIV. That sounds pretty freaking amazing.

Sadly, it’s a bit more complicated than that.

The Study

The study that all of these headlines are referring to was an analysis of a large French dataset of patients infected with both HIV and the hepatitis C virus (HCV). Basically, the researchers looked at about 1,000 people, divided them up by how much coffee they drank, and found that people who drank an elevated amount of coffee — defined as more than 3 cups per day — had a 50% reduced risk of dying over the course of the 6 years of the study.

They also made this nifty graph to show the effect:

Image for post
Image for post
Pictured: nifty

Drink more coffee, die less. It sounds pretty rosy, right?

Sadly, it doesn’t quite work like that.

The Litany

The issues with the study’s reporting are all things I’ve mentioned before.

Sometimes writing a blog is like screaming into the endless abyss

For the sake of brevity, and your continued interest, I’m going to summarize and link to an article that explains it better for each point:

  1. Absolute risk decrease was small. Whilst the number that everyone reported — the relative risk decrease — was big at 50%, the absolute risk of death from drinking lots of coffee was more like 4.5%. That’s not completely worthless, but it’s certainly less enormous than the relative difference that everyone reported on.
  2. The study was observational. Death is a complicated thing, and it can be caused by a lot of factors. This research was not designed to answer the question of whether coffee causes better health, just whether it is correlated with better health. There are far too many factors involved in people dying that this study didn’t have the scope to examine that might have caused the results that were reported on — for example, people who drank coffee could’ve also had better access to medical professionals. We just don’t know.
  3. There weren’t many outcomes of interest. Basically, there were too few deaths to draw any definitive conclusions. Whilst the results were technically statistically valid — for the stats-y people their p-value was 0.032 — the confidence interval for the result went from 30% to 90%. What this means is that there’s a good chance that a study with more numbers would change the result: perhaps not to an insignificant one, but it’s entirely possible that the effect would be dramatically reduced.
  4. The coffee drinkers were pretty healthy. The thing about an observational study like this is that, if your groups start off really different, there’s a good chance that you’ll find results that are not necessarily true. This is because your really, really sick group will skew the results even if you try to control for other factors.

What Does This Mean?

Does any of this mean that the study was useless? No, not at all. As far as I can see, the researchers did a really good job looking into an issue that is really interesting. There’s even a reasonable biological basis for believing that coffee might help with HIV/HCV, because it has anti-inflammatory properties for the liver. As a guide for future research, it’s really interesting — I’m sure there’s an RCT being planned right now to answer the question of whether coffee helps people with HIV/HCV more definitively.

What it does mean, however, is that anyone reporting that coffee is saving lives is getting it wrong. Coffee is a great drink that does have some minor health benefits — my favorite one is that it helps a bit with asthma — but the jury is still very much out as to whether it has serious benefits for long-term health.

In this case, I don’t even really blame the journalists reporting on the study. If you check it out, the researchers essentially claim that this study proves that coffee is stopping people from dying. Not really helpful when the research doesn’t exactly support that conclusion.

As ever, the best health advice is to eat food, not too much, mostly plants. You’ll improve your health much more by quitting smoking and drinking than by doing anything about your coffee intake.

And if you’ve got HIV/HCV, talk to your doctor.

Coffee still isn’t the answer, no matter how delicious it is.

Gid M-K on twitter

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store