DiscussionResearch

Were Fat Studies Buried? The Data They Never Published | Mark Hyman

Mark Hyman discusses the Minnesota Coronary Experiment by Ansel Keys, examining whether it was buried or simply incomplete due to methodological flaws. Despite high dropout rates and control issues in that study, he notes that meta-analyses of higher-quality, longer-duration trials without trans fat confounding variables do show reduced cardiovascular disease risk.

Summary

The transcript discusses the Minnesota Coronary Experiment, an Ansel Keys study that has been subject to debate regarding whether its results were deliberately buried or if the study was simply incomplete and underpowered for publication. The speaker notes that 75% of subjects dropped out in the first year, which may have led the research team to question whether they had sufficient statistical power to detect meaningful differences or publish significant findings. The study was conducted in a psychiatric ward setting with a randomized design where participants were unaware of their fat source—receiving either saturated fats or margarine in a cafeteria setting. Additional control issues included participants being allowed to leave for multiple days at home before returning, which introduced confounding variables by modern standards. The speaker acknowledges that when examining meta-analyses of studies with longer duration, lower dropout rates, and better intervention design that avoids trans fat confounding variables, there is evidence of reduced cardiovascular event risk. The speaker emphasizes that meta-analysis quality depends heavily on the quality of included studies.

Key Insights

  • The Minnesota Coronary Experiment had a 75% dropout rate in the first year, which may have made researchers question whether they had sufficient statistical power to detect significant findings or justify publication
  • The study was conducted in a psychiatric ward where participants could leave and stay home for multiple days before returning, creating control problems that would be considered inadequate by modern research standards
  • Debate exists over whether the Minnesota Coronary Experiment was deliberately buried or simply remained unpublished due to being incomplete and underpowered
  • Meta-analyses of longer-duration trials with lower dropout rates and proper intervention design that exclude trans fat confounding show evidence of reduced cardiovascular event risk
  • The quality of meta-analysis conclusions depends heavily on the quality of the individual studies included in the analysis

Topics

Minnesota Coronary ExperimentStudy methodology and dropout ratesSaturated fats vs margarineMeta-analysis of cardiovascular trialsStatistical power and study design

Transcript

[0:00] Minnesota Coronary Experiment. And that was an Ansel Keys study. And there are some different opposing views on that, whether it was buried or whether it was incomplete. 75% of the subjects appear to dropped out in the first year. So, it may well have been that the research team were like, "This is completely underpowered for us to even publish and understand like do we have a significant finding here or do we have enough power to detect a difference?" This was in the setting of a psychiatric ward. So, the way they had it set up, it was randomized. People didn't know what fat they were getting. So, when they make the food in the kind of…

Full transcript available for MurmurCast members

Sign Up to Access

More from The Proof with Simon Hill

Get AI summaries like this delivered to your inbox daily

Get AI summaries delivered to your inbox

MurmurCast summarizes your YouTube channels, podcasts, and newsletters into one daily email digest.