Diet Soda Correlation Does Not Prove Weight Gain
Norton contrasts observational correlations with randomized substitution trials. People who drink more diet soda may also have higher obesity rates, but trials replacing regular soda with diet soda show weight loss, indicating that the simple correlation does not establish causation.
- Observational traits do not establish what caused them
- Obesity and diet-soda use can correlate without diet soda causing obesity
- Replacing regular soda with diet soda reduced body weight in trials
“Well, that's like saying basketball is causing people to be tall,”
“if we actually look at the randomized control trials where they substitute diet sodas instead of regular soda, they see significant loss of body weight.”