Coffee Isn’t Giving You Cancer
Why you shouldn’t worry about the California ruling
If you’ve been living on the Earth for the last week, you’ve probably heard of the controversial decision made by the California judge that coffee causes cancer. Specifically, that it fails to meet the threshold for safety that California sets in its legislation for things known to cause cancer, and needs a warning label.
As a health person with some interest in the whole coffee thing, I was surprised to hear this. Most of the evidence suggests that coffee is good for you, so it’s a bit odd to hear people arguing that it’s giving us cancer.
So I had a look into the maths behind this decision.
It turns out that you probably don’t have to worry about coffee and cancer at all.
The Nasty Chemical
All of this hubbub around coffee is because of something called acrylamide, a chemical created when you cook various starchy foods. There’s good evidence that high doses of acrylamide in rodents causes cancer in a whole variety of places, but so far we haven’t identified a risk in human studies, meaning it’s placed at 2A on the IARC cancer rating scale, meaning that it “probably” causes cancer.
So we know that acrylamide in sufficient doses is probably going to cause cancer in people. It might not be the same cancers as rats, but it’s likely that there’s some nasty stuff going on there.
But the devil is in the detail. In this case, what’s a safe dose?
Well, if you look at the original rodent studies, they put the amount of acrylamide known as the No Observed Adverse Effect Level (NOAEL) — exactly what it sounds like, the level of intake at which no adverse effects (cancer) were observed — at roughly 0.5mg/kg. This means that, in rats, 0.5 milligrams per kilogram of body weight per day is ‘safe’*. If we extrapolate to people, that would be (for an average-ish 75kg person):
75 x 0.5 = 37.5mg per day
So, to see the effect observed in the studies (which, remember, is conservative because it’s the highest dose that there was no observed cancer), you’d have to ingest roughly 37.5 milligrams of acrylamide per day.
Which begs the question: how much of it is in coffee?
The FDA looked at acrylamide pretty comprehensively in the early noughties, and they put the amount in your average cup of coffee at ~10 nanograms per brewed millilitre. To convert to milligrams, you divide by 1,000,000, so 10 nanograms is 0.00001 mg/ml
That is not a lot.
If we do the maths, to get your 37.5 milligrams of acrylamide per day from coffee:
37.5/0.00001 = 3,750,000 millilitres
So, to get the same effect as seen in the rats, you’d have to drink 3,750 LITRES of coffee. That is about 10,000 average US cups**. Per day. For your entire life.
That’s quite a bit.
Being Safe
Now, before you go out and drink swimming pools full of coffee, there’s an important caveat here. To determine the ‘safe’ amount of a chemical, you can’t just take the numbers from rodent studies and apply them willy-nilly. Rats aren’t the same as humans, so it’s important to put some safety factors in there to make sure that there is absolutely no chance that people are going to be harmed.
The common safety factor recommended by the FDA is ~100x, so you’d divide your 10,000 cups by 100, but since cancer is a special case — it can be seen even in very low doses, and we’re talking lifetime exposure here — you might be conservative and go for 1,000 or even 5,000.
Which brings us back to the original judgement. Because, you see, the State of California is very conservative in protecting people. They didn’t use a factor of 100, or 1,000, or even 10,000.
They used 100,000.
What this means is that your ‘safe’ level of coffee becomes:
10,000/100,000 = 0.1
So even a single cup of coffee puts you over the ‘safe’ level of acrylamide.
Is this fair? It depends. In a state with 40 million people, you could argue that it makes perfect sense. If everyone drinks, on average, one cup of coffee each day, you’d expect it to cause dozens of cancers, because once your sample gets big enough even rare things happen. If the chance is one in a hundred thousand, and you’ve got 400 hundred thousands (40,000,000), then 400 people are going to get cancer.
But there’s another issue.
You see, coffee isn’t the only thing that contains acrylamide.
Acrylamide And You
As I mentioned, the FDA did a lot of work in the early noughties on acrylamide and food. You can find the enormous tables here, and it’s fair to say that they’re pretty comprehensive.
You’ll also see that there’s acrylamide in everything.
Acrylamide forms when you cook things with starch in them, otherwise known as “vegetables”. So things that have more acrylamide than coffee include black olives, cooked potatoes, bread, prunes, breakfast cereal, baby food, cookies, and corn snacks.
In fact, one study estimated that the dietary intake of acrylamide from coffee is roughly 5%, compared with 11% for bread and a whopping 35% for potatoes (mostly fries, potato chips, and a little bit from baked).
So what does this all mean?
Well, firstly, coffee probably isn’t giving you cancer. The doses required are very high, so at an individual level the lifetime risk of you getting cancer from coffee intake is minuscule. More importantly, your average dietary intake of acrylamide is mostly made up from non-coffee things, so if you really want to cut down on the stuff you should start boiling your potatoes, eating your bread untoasted, and leaving your fries only French.
If you’re a public health organization worried about the impact of tiny risks on a huge population, you might want to put warning labels on coffee.
But for an individual? For you?
There’s nothing to worry about at all.
If you’re concerned about your cancer risk, here’s a list of FDA recommendations to reduce your intake. It’s mostly cooking certain things less or in a different way.
Just don’t worry about getting cancer from coffee.
You probably aren’t drinking enough for it to matter at all.
Follow me on your favourite platform!
*Note: THIS IS EXTREMELY SIMPLIFIED. I could put this every paragraph, but let’s just say that the California decision takes 40 pages to partially explain, so this blog represents a much easier to digest version of what you actually do to calculate cancer risk. The full calculations are here.
**Using the average ‘regular’ sized Starbucks coffee, which is ~500ml. This is a lot, but as you’ll see even a lot of coffee does very little to your acrylamide exposure
