The Bradford-Hill Criteria give you a scientific checklist that can help you evaluate whether wild health claims are actually true.
What is causality anyway?
Causality means that one thing directly makes another thing happen.
For example, flipping a light switch (A) causes the light to turn on (B), so A causes B.
But ice cream sales (C) and drownings (D) both increase in the summer – not because ice cream causes drowning, but because hot weather causes both, so C and D just happen to occur together. The Bradford-Hill criteria can help us figure out which is which!
🤓 Nerd Alert – Why are they called Bradford Hill criteria?
Sir Austin Bradford Hill was a British epidemiologist and statistician who first presented the criteria in 1965 (not THAT long ago!). The aim was to take something that looked statistically linked and think critically about whether there was truly a cause-and-effect link.
He was also famous for carrying out the first randomized controlled trials (now the gold standard in research), and he was the first person, together with his colleague Richard Doll, to link smoking to lung cancer.
What are those criteria then?
Bradford-Hill proposed 9 criteria, and here we’ll cover the most important and commonly used ones. Consider them as a checklist in your toolbox for critical thinking.
✅ Temporality: Does the cause come before the effect?
🦟You have to be bitten by a mosquito before you can get an itchy bump. You can’t get the itchy bump first and then get bitten afterward. The cause (mosquito bite) must always come before the effect (itchy bump) in time. This is the only criterion that is absolutely essential. IF this doesn’t apply, then there is no “causality”.
✅ Strength: How strong is the relationship between the cause and the effect?
🚬 Not only can smoking cigarettes lead to lung cancer, but data show that this cause-and-effect relationship is very strong. In fact, the link between smoking and lung cancer is so strong that it’s become a textbook example of causation in medicine – it’s not a small or subtle effect!
✅ Dose-Response relationship (aka Biological gradient): Does increasing/decreasing the exposure increase/decrease the effect? (So more exposure = more effect?)
🚿🏃 A popular TV personality claims in an interview that increasing the number of showers you take will improve your running speed. If this is true, then the more showers you take, the faster you should run, and the fewer showers you take, the slower you should run. But when researchers look at actual runners, there’s no pattern – people who shower once a day run just as fast as people who shower twice a day. No dose-response relationship = probably not causal!
✅ Consistency: Have there been similar results shown by other researchers?
💤☕You mention to a friend that you are having trouble sleeping, and they suggest you try drinking lettuce tea. Your friend claims there is scientific evidence behind this idea, but when you dig a little deeper, you find that there has only been one study done, and it wasn’t on humans! No other studies have been able to show that drinking lettuce water will aid sleep.
✅ Plausibility: Does the association make biological sense?
🍋⬇️ You read a claim on social media that adding lemon juice to your coffee will help you lose weight, but when you dig a little deeper, the recommendation is not found on any reliable website. You are also unable to find how lemon juice would help with weight loss (eg: Does it suppress your appetite? Does it help you burn calories?).
✅ Experiment: Has experimental evidence supported the relationship?
🖐️🧼Doctors in the 1800s noticed that fewer mothers died of infections when doctors washed their hands before delivering babies. But to really prove handwashing caused the difference, one doctor divided his clinic into two groups – one where doctors washed their hands, one where they didn’t. The handwashing group had far fewer deaths, providing experimental proof that clean hands save lives!
Bottom line: So, next time you hear a claim that something causes (or cures) a disease, think about these criteria and see how well the claim holds up!
Stay well. Stay curious.
Love,
Those Nerdy Girls
