My turn! I’m leaving academia shortly for an epidemiology role in industry. Exciting times. I won’t rant about push factors (you know them all). I will say I’ve been craving work that’s more practical, impactful, team-based, & yes, moneyed.

It's been a busy week for the group! 2 new preprints out on better understanding colorectal cancer risk https://t.co/FCsKIK5Sfc

Mendelian randomization studies of disease progression can speak directly to therapeutics, but there are many challenges. Ruth Mitchell, @april_hartley @venexia Apostolos Gkatzionis @jamesyrmsk @joshuaa_bell Amanda Chong @drpaternoster discuss them here

Delighted to see our work on the metabolic features of genetic liability to type 2 diabetes and coronary artery disease published in @eBioMedicine today! @joshuaa_bell @emmylooroll @caroline_bull @ECSanderson @mendel_random @mvholmes https://t.co/6cSlFrVRhI

To pre-empt: •Off-peak is £65 return, the only option I’d ever do. But prob can’t rock up at noon, & working on the train isn’t possible when squashed in the aisle •Single & advanced bookings make little difference •I don’t qualify for railcards Home-based it is!

So I now have a shiny new office in Paddington to work from if I wanted, < 2 hrs from home Only catch is it costs 100% of salary to get there 😂 I will never understand UK rail & why we put up with it

Bear in mind: not only are reviewers unpaid, but the editors seeking reviewers often aren’t paid either! It’s a voluntary side gig. Publishers cannot profit (massively) from this broken system yet take no responsibility for fixing it. PAY PEOPLE

And thanks @njoybetz @LifeAfterMyPhD & others with @AltAcChats. Haven’t met, & different fields, but your words have been helpful

I’ve not met anyone who thinks I’m making the wrong decision. I’ve worked with brilliant & kind people & I wish them the best. It’s just time for change. I’ll still exist! And will share about this transition. Very happy to chat with any epi folks considering the same.

•‘Disease proneness’ has a long reach & prob has something to do with lipids •Traditional academic publishing is due for implosion My main contribution to Mendelian randomization was getting the STROBE-MR checklist posted in Word format (tweeting works, you’re welcome)

SO, discoveries over the years: •Metabolically healthy obesity is a stupid idea •Obesity prob causes lots of metabolic problems & cancers •Delaying puberty prob won’t stop obesity or heart disease •Getting active & muscley is good but prob less so than losing some fat

Fellowships: I tried for 7, got 1. Soul destroying process of luck. Those rejections do feel personal & kill what interest & passion you have left. The time cost for each try is too high. Not worth it.

And money. I make no apologies. It’s useful for buying things, like a place to live. UK academia pays terribly. I’m first gen for the uni path, parents in manual jobs. I do actually need to make money & build my own life. It does feel like academia isn’t made for us.

I want more structure. A role with objectives clearer than ‘go away & discover things & win awards’. ‘My time is my own’ was great but got demotivating. I like deadlines & they don’t really exist in academia. I realised that’s because no one really needs what I’m doing

ICYMI, new preprint! 👇 (posted just before Xmas for sanity’s sake) We compared BMI-mortality among Covid & non-Covid respiratory patients admitted to ICU. Obesity looks harmful in Covid but protective in other respiratory diseases. What’s going on here?

My first attempt at cross-context comparison! Lots to unpack, & illustrates how causal inference is really a qualitative task. Thanks to @EBIBristol for grant support & stellar @ICNARC for hugely valuable data & input. And @mendel_random @ammhughes & other bias pros in @mrc_ieu

Looks like: 1. Only extreme obesity raises ICU mortality in Covid. This was mostly in early 2020. 2. Obesity consistently lowers ICU mortality in non-Covid, but I suspect reverse causation (which doesn’t vary by time/place). Non-Covid patients have more comorbidities.

We used UK national ICU data to compare BMI-mortality among severely ill Covid & non-Covid patients. We then asked how stable these effects are across times & places where biases differ. More stability: more confidence in causality.

2023 may bring dual surges of Covid & flu. Jan is busiest for ICUs. This makes for tricky obesity messaging… Obesity appears to shorten ICU survival in Covid, but lengthen it in other respiratory diseases. Causal? I doubt it. New preprint’s up: https://t.co/cOVpHEvzty

Dearest Students, I don't want to hear a single word of complaint from you EVER, about ANYTHING. #AcademicTwitter #AcademicChatter