Seemingly trivial things that intrigue you

Table 2 is where most of the analysis can be found.

I'm not great on statistics, but "model 1" is age adjusted, and shows for the mere presence of a wood burner or fireplace, an HR of 1.05, with a 95% confidence of 0.88 (which would mean wood burners reduce risk of lung cancer) to 1.31, centred at 1.05.

Under model 2, they adjust for no fewer than 9 factors all at once and what do you know the presence of a wood burner or fireplace means this jumps to an HR of 1.42 (CI @95 1.11-1.81).

Okay, so far seems plausible. Then they specifically separate out whether that fireplace was actually used or not. The age adjusted model 1 data shows an HR of 1.25, but the model 2 HR is still the same - 1.43, yet 12000 of the cohort of 31000 have been removed and added to the "baseline" because they say the fireplace wasn't used.

Therefore, this seems to show that there is no difference, in their multi-adjusted model 2 data, between the HR regardless of whether the fireplace was used. Just having a lump of iron in the room increases the risk of lung cancer by 40%, under their model. So what is this showing? That people in old houses are more prone to lung cancer?

This suggests either the data is wrong (i.e. "Did the fireplace ever get used?" "Um... no?") or the various assumptions they used to compensate for (x9) are wrong somehow.

You an kind of see the problems in the next row down, where they split out no use, 1-29 days a year use and at least one month a year use of said fireplace. Model 1 (age adjusted only) shows an HR of 0.88 and model 2 1.14 but with a wide CI and both relatively small numbers (less than 100 cancer diagnoses).

Table 3 gets worse, because it seems to show that smoking has a protective effect on cancer risk from a wood burner. There's a whopping HR of 1.9 if you never smoke, but 0.99 if you smoke.

A lot of the rest seems to me to be such small subsets that conclusions are (even more) risky.

To summarise, I think they analysed the shit out of these data and drew conclusions based on these HR and CI values, but didn't step back to see whether they were consistent with one another or actually made sense. The fact they had a smattering of HR values of less than 1 should ring alarm bells and cast doubt on at least positive associations of the same magnitude above 1.

The problem seems to me to be the data itself. Smoking is down to never, or "more than once" and usage of a fireplace over a decade is broken down to never, rarely, regularly. And all of this is self-reported.

I thought processed foods was basically an epidemiological correlation they’ve found and they’re now trying to work out the causation.

They've moved on to ultra-processed food now, because people like me always pointed out that milk and flour were processed. But I'm annoying like that. Same goes with people who "cook from scratch" - they don't mill the grain into flour.

Turns out one of the MiL’s “most watched channels” is GB news FML

She’s got very friendly with a 65 year old EDL supporter, all tied to that, really.

Just shows who can support that kind of nonsense.

@First.Aspect she knows better than to talk politics with me.

Less conspiratorial more hardcore cliche boomer stuff.

You name a boomer cliche, she’ll have said it

Avocados?

Yeah. Avos, wokery, pitiful state pensions but too generous benefits.

boJo is a legend and the establishment stitched him up.

Kids nowadays are too soft. Can’t say anything these days, political correctness gone mad.

IHT is a moral travesty etc

Younger generations moan too much. My favourite is that my family doesn’t work hard enough to afford proper sized houses (she’s never had a paid job in her life and she lives on her own in a 5 bedroom house, in a cul de sac of similarly sized houses, all occupied by widowed or coupled retirees who’s children have all left a decade or two ago)

Tim Hortons only ever did did good doughnuts, mediocre bagels and bad coffee.

If the one local to me is anything to go by, they now do bad burgers, in addition to their bad doughnuts.

Start glueing yourself to things.

Nah I just think a “f@ck em, I deserve more” politics is all anyone else does.

what is it the lifetime pay in to the UK coffers for those born in the 50s is £950k and the total lifetime government spend on benefits is what, £1.2m

They’ve played a blinder so f@ck em.

Just f@ck old people. If they were bad at saving tough sh!t.

Cut pensions. Charge for GP visits.

Give em an option for a housing capital gains tax every year, starting with a backdated bill or massive IHT. Chose your poison.

Any adult.

Flat fee, same as GP visit.

So if you’re at death’s door, great value. Hilary going because she got lightheaded after going up the stairs 8 time in a row, maybe not so much.