Colin Angus

Age-period-cohort effects are hard. Just ask a demographer.

tl;dr cohort studies are extremely cool, but tell you very little about period trends (at least, without external corroboration)

Drug use trends in young people may or may not be good, or bad (I've no idea), but knowing that 23 year olds are more likely to have ever tried drugs than they had when they were 17 isn't really very informative at all about that question.

As an aside, getting worried about the apparent increase in drug use is also not entirely sensible. If you ask people about whether they have ever done something, then asking older people will always get you more yeses, simply by virtue of the fact that older people have had more time to do stuff.

Or less. WE HAVE LITERALLY NO IDEA.

So while this is an interesting bit of research/data, it doesn't really show what people seem to be interpreting it as showing.

It tells us nothing whatsoever about whether things are getting better or worse overall.

In order to know whether we should be concerned or not about these findings, we need to know what the same trends look like in previous generations. It is entirely possible that rates of binge drinking rose far *more* between the ages of 17 and 23 in previous, less abstemious cohorts.

Longitudinal data like this is really cool and useful (and a massive faff to collect), but we can only interpret their findings as evidence in shifts over the lifecourse of the participants in the study.

It isn't surprising at all that people drank more when they were 23 than when they were 17.

Binge drinking rises sharply among gen Z in their early 20s Research challenges idea of ‘generation sensible’ as alcohol and drug use increase after teenage years

Was surprised to see reports today of rising binge drinking among young adults. But the comparison is current 23 year-olds with themselves when they were 17, not with any previous generation. So we've no idea if binge drinking *in 23 year-olds* is rising or not.

www.theguardian.com/society/2026...

This is interesting stuff on the timing and persistence of the life-expectancy impacts of the pandemic.

Notably for the UK, England & Wales and Scotland are in the 'first wave peak + rebound to pre-pandemic levels' group, while Northern Ireland is in the 'prolonged depression' group.

Today's episode of adventures in UI excellence. Who on Earth designed this and what is wrong with them?

This is an astounding photo.

Took a bit of work to track down as it has been reposted so much on Facebook without credit, but this is taken by David L Batcheller.

b) different alcohol policies are supported by different levels of evidence and can have very different levels and distribution of benefits. Modelling studies can help us to understand these differences, and help policy-makers to chose the policies that make the most sense for their contexts.

In the commentary I argue that:
a) modelling evidence like this is critical for understanding the potential effects that alcohol policies can have on chronic conditions like cancer where the impacts aren't immediate and can be tricky to identify in evaluation studies
...

Finally, I wrote a commentary for @thelancetph.bsky.social in response to an excellent new modelling paper looking at the potential impact of Minimum Unit Pricing and health warning labels on alcohol-related cancer harms.

Original paper: doi.org/10.1016/S246...
Commentary:
doi.org/10.1016/S246...

Using survey data from Jan-Mar 2023-25 we found that those experiencing financial difficulties were considerably more likely to cite cost as a factor in motivating an attempt to reduce their alcohol intake, however, they were no more likely overall to have tried to cut down (for any reason).

The second paper is led by @drsarahejackson.bsky.social as part of our project evaluating tax reforms and the cost-of-living crisis sarg-sheffield.ac.uk/projects/ref...

We looked at associations between cost as a motivator to reduce drinking and financial hardship.

doi.org/10.1016/j.dr...

We found that many of the venues were at low occupancy and/or closed early, which challenges the idea that later opening hours can help 'revitalise' the night-time economy.

We also found routine sales of alcohol to people who were already intoxicated (unsurprising, but a breach of licensing laws)

The first paper is led by Carol Emslie and comes from our ELEPHANT project evaluating extensions to late-night alcohol licensing hours in Aberdeen and Glasgow.

This paper reports on venue observations during these extra opening hours

doi.org/10.1016/j.dr...

It was a busy old week last week, 3 new papers published and I spent Friday in Liverpool speaking at the RCGP conference on Managing Addictions in Primary Care about our alcohol policy research and recent developments in UK alcohol policy (both good and bad)...

Oh my, this is incredible. I want one.

Oh, good point! Age-standardised deaths per 100,000 population.

Interestingly, Scotland still has higher rates of alcohol-specific deaths among men, but Northern Ireland has the highest rates among women.

R code for these plots is here: github.com/VictimOfMath...

NISRA have just released new data on alcohol-specific deaths in Northern Ireland in 2024. This confirms the sad fact that Northern Ireland now has the highest rates of alcohol-specific deaths of any UK nation.

Well this is beautiful and extremely cool

New research shows that spending less on tobacco, gambling and sweets is good for the UK economy - Sheffield Addictions Research Group A new study from the Sheffield Addictions Research Group (SARG) has found that public health policies which reduce spending on harmful products can actually provide a significant boost to UK jobs and ...

New SARG research published in @addictionjournal.bsky.social this week led by @damonm1989.bsky.social has found that public health policies which reduce spending on harmful products can actually provide a significant boost to UK jobs and the wider economy. #PublicHealth #HealthEconomics #UKEconomy

Welsh Government have just voted to retain the Minimum Unit Price for alcohol and to increase its level from 50p to 65p per unit, bringing it into line with Scotland. The increase will take effect from 1st October this year.

So obviously there is a correlation between pub sales and alcohol volume sales, but we can't infer anything directly about changes in alcohol volume sales.

This is total turnover (I think), so includes food and other sales, not just alcohol. LFL is basically adjusting for changes in the composition of what people are buying. E.g. if sales were 30% food, 70% alcohol in 2024, but 40%/60% in 2025, the LFL comparison adjusts for this difference.

Turnover in the on-trade in the UK *rose* by 2.9% on a like-for-like basis (and 6.2% overall) in December 2025 compared to December 2024, with the biggest growth in pubs.

So much for the 'government policy on alcohol taxes is killing pubs' argument.

SARG researcher Dr Parvati Perman-Howe secures NIHR Career Development Award - Sheffield Addictions Research Group We are delighted to announce that Dr Parvati Perman-Howe, a Research Associate within the Sheffield Addictions Research Group (SARG), has secured a highly sought-after Career Development Award.

🎉 We are delighted to announce that SARG’s @ppermanhowe.bsky.social has secured an NIHR Career Development Award to investigate the public health impact of mid-strength drinks. Read our news story to find out what her 12-month project will involve 👇 @fohsheffield.bsky.social @nihrsphr.bsky.social