Science can sometimes be a real buzzkill. Everything that is good for you one week turns out to be bad for you the next week, according to many science headlines. This article addresses how over-simplified study interpretations can make seemingly healthy habits appear detrimental. It emphasizes the importance of understanding the difference between correlation and causation. We will dissect examples like intermittent fasting and highlight the crucial role of critical thinking when interpreting scientific findings, empowering you to make informed health decisions.
The Case of Intermittent Fasting
For example, a headline announced that individuals on time-restricted diets, commonly known as ‘intermittent fasting’, were found to have a 91% higher chance of dying from a heart attack.
NBC News: Intermittent fasting linked to higher risk of cardiovascular death, research suggests
Isn’t intermittent fasting supposed to be healthy? Why would it increase heart attack chances?
It turns out that this is yet another example of the phrase that everyone needs to learn whenever they see an incendiary science headline: “Correlation is not causation.”
The Invisible “Third Factor” at Work
The human brain is drawn to correlations, and there’s a natural urge to try and find causation from correlation. If I push a switch, and a light turns on, the light must be triggered by the switch.
This approach worked great for the natural world, for millions of years, but it falls apart with many of today’s complex systems. Ice cream consumption, for example, is correlated with shark attacks. The more ice cream you’ve eaten recently, the more likely you are to be attacked by a shark.
Of course, sharks aren’t hunting down people with delicious ice cream, although it would be an understandable response. Instead, there’s a third factor at play: we eat more ice cream during the summer, and also spend more time at the beach where shark attacks can occur.
We see a similar correlation between flossing our teeth and heart disease. People who don’t floss their teeth are more likely to suffer heart attacks. Are oral bacteria, if not regularly cleaned, causing heart attacks?
More likely, people who have poor overall health or don’t have sufficient healthcare access don’t do preventative tooth care and are also more prone to heart attacks.
A third factor drives the other factors. Lack of health care (factor 3) drives poor gum health (factor 1) and heart attack risk (factor 2). Factor 1 and 2 are correlated, but not causative.
Another example: shoveling snow is sometimes rated as ‘one of the most dangerous activities you can perform,’ with a significant number of people having heart attacks while snow shoveling. But shoveling snow is not inherently bad; it just happens to be a vigorous aerobic activity that a lot of out-of-shape people perform.
Intermittent Fasting: Choice or Circumstance?
Let’s get back to our headline study, which points out how intermittent fasting is correlated with a higher risk of heart attacks. Excellent Medium author Dr. Jason Fung gives a great write-up of the limitations of this study and how it’s leading to spurious causality claims in headlines.
This study looked at data collected from patients between 2003 and 2018. About 20,000 adults recorded what they ate as part of the Centers for Disease Control and Prevention (CDC)’s National Health and Nutrition Examination Survey. It found that individuals who consumed their daily food intake over 8 hours or less were much more likely to die during that 15-year period from a heart-related event.
But was this intermittent fasting, a healthy lifestyle choice? Or is this a lack of access to food?
The concept of intermittent fasting started gaining popularity in 2012, and grew over the next decade. But this introduces our first flaw with the study’s findings: many of these individuals probably weren’t on an intermittent fasting diet because it wasn’t a popular diet in 2003.
Why were these individuals not eating for 16 hours per day? The authors of this finding haven’t yet published their data or a peer-reviewed paper about the details of their study yet, but there are a few alternative possibilities:
- These individuals were low-income and weren’t able to regularly afford multiple meals.
- These individuals had other significant health issues that prevented them from easily preparing food for themselves.
- These individuals may be elderly and not have constant food access (think: assisted-care nursing home residents).
The authors of these findings also noted that their data did not describe how many calories were ingested per day or whether the individuals ate nutritious, healthy food or not.
Overall, there’s no evidence for causation here. There certainly seems to be a strong correlation, but it’s almost guaranteed that alternate factors are driving both of these outcomes (restricted feeding and heart attacks).
Thinking Critically About Attention-Grabbing Science Headlines
Whenever a new study gets published, it’s usually shared by the media department of the organization that published it. Universities, for example, have PR teams that create press releases about findings from on-campus research groups and labs.
Those press releases focus on the most attention-grabbing, impactful aspects of the studies — not necessarily on the most well-supported and proven points.
Try to ask yourself, “is this study specifically able to demonstrate causation? Are the researchers running a controlled trial, where they give ingredient X to see if it causes outcome Y?”
Next, think critically! Ask yourself, “can I think of another factor that would drive both the measured statistic, and the outcome?”
A study isn’t worthless if it only demonstrates correlation, not causation, but there are myriad examples of where we have assigned spurious causation to links that are correlation-only, where we assume that factor A causes outcome B, instead of considering how factor C may be the driving force that leads to both A and B.
Other great examples:
- Screen time in young children causes worse educational outcomes: probably linked to income, not the screen itself.
- Social media and depression in teens: depressed teens are more likely to spend more time on social media, but social media itself doesn’t cause depression.
Remember, a real scientist is critical and insists upon clear evidence for everything — including results found by other scientists!
Conclusion
In conclusion, it is vital to approach sensational science headlines with a healthy dose of skepticism. Always consider whether a study demonstrates causation or merely correlation. By identifying potential ‘third factors’ that might be driving observed relationships, we can avoid making misguided health choices based on oversimplified interpretations. Encouraging a critical mindset ensures we are well-informed and proactive in managing our health.