Exercise Snacks: Can just four minutes of exercise a day improve blood sugar?
Summary
Hosts Kristen Sinani and Regina Nuzzo evaluate a study on "exercise snacks" for type 2 diabetes, finding the primary outcome failed while media highlighted significant secondary results. They discuss how researchers added unpre-registered analyses, leading to potential cherry-picking of data. The episode critiques science journalism for overselling single studies and emphasizes the importance of checking pre-registrations to assess research integrity.
Topics discussed
Intro: The problem of overselling single studies
Defining 'exercise snacks' and the 15-second threshold
The Washington Post headline and the study's premise
Why people resist exercise and the barrier to entry
Hosts' personal attempts at the exercise snack protocol
Reflections on protocol adherence and compliance
Study details: Journal, authors, and participant demographics
Study design: Case crossover trial and intervention specifics
Measuring outcomes: Blood glucose monitors and compliance logs
Intention-to-treat analysis and missing data
Continuous glucose monitors: Utility and controversy
Primary outcome results: No significant difference in glucose
Media coverage vs. actual study findings
Sponsor segment: That Aged Well podcast
Primary vs. secondary outcomes and statistical power
Reviewing the study's abstract and secondary claims
Analysis of secondary outcomes: Spikes, variability, and time in range
Consistency of results and the 'cold hands' analogy
Pre-registration, transparency, and potential p-hacking
Protocol changes and sample size adjustments
Discussion on effect sizes and practical exercise tolerance
Final verdict: Smooch rating and closing thoughts
Listen ad-free on Castria