Let’s Start with the Truth examines a study using the end of British sugar rationing in 1953 as a natural experiment. Among more than 64,000 UK Biobank participants, greater exposure to rationing during the first 1,000 days of life was associated decades later with substantially lower rates of five cancers, including a reported 69% lower hazard of liver and intrahepatic bile-duct cancer. The study describes these effects as causal, but the article argues that this conclusion depends heavily on assuming that nearby birth cohorts differed primarily in their early exposure to sugar rationing.

The reported 41% lower hazard of lung cancer provides a useful test of that assumption. Because lung cancer is strongly influenced by lifetime smoking and other environmental and social exposures, such a large association raises the possibility that the birth cohorts differed in important ways beyond childhood sugar intake. The article doesn’t argue that this disproves a protective effect of early sugar restriction; instead, it uses the finding to illustrate the limits of causal inference from natural experiments. The larger lesson is that especially large, appealing results deserve scrutiny for alternative explanations—even when they support something we already believe.

https://letsstartwiththetruth.substack.com/p/the-sugar-studys-lung-cancer-problem