PLAY PODCASTS
Alcohol: Are happy hours good for your heart?
Episode 6

Alcohol: Are happy hours good for your heart?

Normal Curves: Sexy Science, Serious Statistics · Regina Nuzzo and Kristin Sainani

April 21, 20251h 6m

Audio is streamed directly from the publisher (op3.dev) as published in their RSS feed. Play Podcasts does not host this file. Rights-holders can request removal through the copyright & takedown page.

Show Notes

Does a daily glass of wine really keep the cardiologist away? It’s a claim we’ve all heard: light to moderate drinking is good for your heart. But is it science or just a convenient excuse for happy hour? In this episode, we dive into the history behind this claim, discuss the challenges of observational studies and statistical adjustment, and explore attempts at randomized trials and natural experiments to get to the bottom of this boozy debate. Grab your drink—or maybe don’t—and join us!


Statistical topics

  • Statistical Adjustment
  • Regression
  • Residual and Unmeasured Confounding
  • Randomized Trials
  • Multiple Testing
  • Outcome Switching
  • Mendelian Randomization 


Methodological morals
“Statistical adjustment cannot erase all confounding.”

“When you can’t experiment on people, let Nature experiment on people.”


Citations


Page with more details on the CASCADE trial


Kristin and Regina’s online courses: 


Program that we teach in:

Epidemiology and Clinical Research Graduate Certificate Program 


Find us on:

Kristin -  LinkedIn & Twitter/X

Regina - LinkedIn & ReginaNuzzo.com


  • (00:00) - Introduction
  • (03:00) - Drinking habits in America
  • (04:13) - New Canadian drinking guidelines
  • (07:51) - Definition of light-to-moderate drinking
  • (08:43) - Risks and benefits of light-to-moderate drinking
  • (11:37) - History of the heart health claim
  • (18:34) - Problems with observational studies
  • (23:37) - Statistical adjustment
  • (26:36) - Residual and unmeasured confounding
  • (32:16) - Overconfidence in observational studies
  • (36:13) - Randomized trials of alcohol
  • (37:29) - Canceled NIH randomized trial of alcohol
  • (42:39) - The CASCADE randomized trial of wine
  • (44:15) - The problem of multiple testing
  • (48:53) - Outcome switching
  • (50:29) - Mendelian randomization
  • (01:00:01) - Mendelian randomization studies of alcohol and heart disease
  • (01:04:06) - Wrap-up

Topics

Normal Curves podcastKristin SainaniRegina NuzzoStanford UniversityGallaudet Universityjournal club podcaststatistics podcastscience podcastfemale co-hosts podcastconversational science podcastpodcast for grad studentspodcast for medical studentspodcast for science postdocsbest science podcastsscience communicationmedical writingscience writingstatisticscritiquing scientific studiesdebunking psychology mythsthe science behindhow to read scientific papersMathmath podcastdata analysisscientific researchprobabilitystatistical literacydata interpretationscientific studiesstatistical methodsdata visualizationstatistical significancehypothesis testingbiostatisticsstatistical modelingquantitative analysisstatistical inferencedata sciencestatistics educationstatistics consultingstatistics humorsociety and culturedatingsexfitnessrelationshipsmedicineprofessorsfashionfemale professorsjournal clubNormal CurvesNormal Curves podcast alcohol episodealcoholhappy hoursmoderate drinkingis drinking good for your hearthealth benefits of alcoholhealth effects drinking podcastis wine good for your heartis alcohol good for your heartstatistical adjustmentregressionresidual confoundingunmeasured confoundingexamples of confoundingexplanation of confoundingexplanation of statistical adjustmentrandomized trials of alcoholrandomized trialsmultiple testing exampleoutcome switching examplemendelian randomization alcoholmendelian randomization example