Splendid Isolation

I’m often castigated by Friends and Readers (some overlap) for reading the foul Brit Daily Mail rag, and my answer is generally the same:  yes, they’re awful, the articles are often dire, the celebrity-obsession is tiring, and all the rest of it.  I know all that.

What I appreciate is that unlike the other online news outlets I read (Breitbart, American Thinker, NewsMax etc., whose principle editorial slant is politicspoliticspolitics all the time politics), the DM occasionally runs articles that are not about politics, nor about which little-known “celebrity” is bonking another of their own ilk.

Here’s one, written by some doctor bloke:

For years I’ve been taking a daily omega-3 supplement because I don’t eat enough oily fish.

This matters – partly because of the possible heart and anti-inflammatory benefits, but for me, mainly because the disease I fear most is dementia and my hope is that omega 3 will help prevent it.

So when I saw the headline about a new study suggesting omega-3 supplements might not protect against dementia but could actually be linked with faster decline, I panicked – not just worried about myself, but what I advise others, too. The researchers compared 273 people who take omega-3 supplements daily, with 546 similar non-users – and found that those taking fish-oil pills appeared to decline faster on several cognitive scores.

I got worried, too.  I don’t know how long I’ve been popping fish oil tablets (in the absence of actual fish in my diet, don’t ask), but it runs into multiple decades ever since I was first advised to do so by my own doctor.

However, here’s where the article is refreshingly candid:

So am I worried? The study was observational – where researchers look at what people are already doing, and then search for associations; for example, between using omega 3 and cognitive decline.

This sort of study can be useful because it suggests avenues for further research, but – crucially – it cannot easily prove cause and effect.

The curse of observational nutrition research is that it can make almost anything look good or bad depending on how the research is conducted.

Coffee once looked harmful in observational studies, because coffee drinkers were more likely to smoke.

And the good doctor goes on to explain why the original scary headline was a load of old bollocks.

As a one-time statistician myself, I know that this kind of bullshit has been foisted on the public for too long, and it needs to stop.  And it’s not confined to nutritional research, either.

Here’s an old example of where observational research caused actual harm.

Anyone remember the time a government study found a link between elevated cancer risk and a house’s proximity to electrical power-transmission lines?  Yes?  And do you remember that it set off a minor panic in the real estate market, with said properties losing as much as half their market value because who the hell wants to get cancer just by living close to a power line?

Of course, all that turned out to be total nonsense, because the original study had not been designed to measure cancer risk against power line proximity — that “link” was discovered by observation, not by the actual study itself.  In fact, the observation was purely coincidental, caused by sample distortion.  In other words, it just so happened that of the houses in the study, there was indeed a higher-than-average incidence of cancer occurrence.  But when the sample was expanded proportionately to include housing not located near to power lines — a much greater number, of course — it was discovered that the incidence of cancer was not especially higher in one house or another, regardless of any nearby power lines.  Higher incidence of cancer was linked, of course, to cigarette usage and genetics, not to whether your house was next door to a power line.

In the meantime, of course, untold millions of dollars were lost by those unfortunate homeowners whose houses had been branded as “cancer-causing”.

It was irresponsible reportage of the highest order — and by “reportage” I mean the publication of those observations by the so-called scientists who found the alleged linkage, not by the press (who were just reporting what they’d been told by the Gummint).  And yes, I know, the press should have investigated the numbers before making those “Avoid Buying These Houses!!!” headlines;  but journalists as a rule are not renowned for their statistical understanding at the best of times, as any fule kno.

The responsibility for publishing observational data lies completely with whoever compiled the data.  The problem, of course, is that people (scientists and doctors no less than anyone else) are obsessed with prevention of anything that has to do with public health.  That’s not altogether a Bad Thing, of course, but that obsession needs to tempered by reluctance to publish anything that wasn’t part of the original study’s stated goal:  tangential or even parallel conclusions, as we have seen, are at best faulty and at worst harmful.

In the mean time, as Dr. Rob Galloway suggests, you should keep taking those fish oil tablets if you’ve been advised to do so — but what you should really avoid is taking fish oil tablets which are past their expiration date, because those could actually be harmful (for the reason he gives in the article).

So avoid those bargain bins at the supermarket — invariably, they’re filled with old unsold stock, which is why the price has been massively reduced — and take only the stuff still on the shelves.  Saving a buck or two on the cheaper stuff may not be good for your health.

Caveat emptor.

Oh, and go and check your meds and such for any expired products.