How to read a brain study: causality, reproducibility and hype
Peter Attia's journal clubs on epidemiology and statistics, Mark Gordon on research fraud, and placebo research on where effects stop all teach readers to question associations and small studies. It helps people judge health claims instead of chasing each headline.
What is reverse causality? Show answer
When the outcome drives the behavior that seems to cause it, as when depression keeps people indoors by day and awake at night.
What is Peter Attia's diet soda example? Show answer
Diet soda is linked to obesity, but obesity may be causing the choice of drink, not the other way around.
What can settle whether an association is causal? Show answer
Direct randomization or Mendelian randomization. Observational epidemiology cannot.
How do you read an odds ratio? Show answer
Subtract one to get the percent change. A ratio of 1.78 means 78 percent more risk, and 0.7 means 30 percent less.
Why is a big sample not automatically trustworthy? Show answer
Huge studies can be overpowered, making clinically irrelevant differences look statistically significant.
What went wrong in the 2014 metformin study? Show answer
Informative censoring. Users who stopped the drug, were lost or moved to stronger medication were dropped, making the group look healthier.
Why did Attia still accept most of the light and mood finding? Show answer
It showed a consistent dose-response, strong biological plausibility, and the habit is a low lift.
What are Huberman's four questions for a paper? Show answer
What question are they asking, what approach did they use, what did they find, and do the conclusions follow?
What did the Alzheimer's fraud case involve? Show answer
Apparently falsified and reused images flagged across 132 papers by one prominent neuroscientist.
Why does every health claim need a placebo comparison? Show answer
Expectation alone produces real changes in dopamine, hormones and pain, so only a controlled comparison isolates the treatment.
Where does expectation stop working? Show answer
At structural outcomes. Placebo eased asthma discomfort without improving breathing and does not act on tumors.
How should you match your response to the evidence? Show answer
Act on cheap, low-risk habits with consistent evidence, and wait for stronger evidence before changing drugs or supplements.