Okay, it’s time to finally answer the question: is drinking booze good or bad? Is there really a “J-curve”, such that it’s bad to drink zero alcohol, good to drink a little, and then bad to drink any more than that? What exactly is the “safe level” of alcohol consumption, and why do the meta-analyses on this topic all seem to tell us entirely different things?

In this episode of The Studies Show, Tom and Stuart get very badly intoxicated—with statistics.

We’re sponsored by Works in Progress magazine. There’s no better place online to find essays on the topic of “Progress Studies”—the new field that digs deep into the data on how scientific and technological advances were made in the past, and tries to learn the lessons for the future. Check them out at worksinprogress.co.

Show notes

* Media reports say alcohol is good! Oh no wait, it’s bad. Oh, sorry, it’s actually good! No, wait, actually bad. And so on, ad infinitum

* The three conflicting meta-analyses:

* 2018 in The Lancet (“no safe level”)

* 2022 in The Lancet (the J-curve returns)

* 2023 in JAMA Network Open (using “occasional drinkers” as the comparison)

* Some of the press coverage about the J-curve age differences

* David Spiegelhalter’s piece comparing the two Lancet meta-analyses

* Tom’s piece on the idea of “safe drinking”

Credits

The Studies Show is produced by Julian Mayers at Yada Yada Productions. We’re very grateful to Sir David Spiegelhalter for talking to us about this episode (as ever, any errors are ours alone).