Will This End Up Labelled As Bad Science?

[Hypognosis]

Counter intuitive and currently on my 'interesting if true' list.

http://www.dailymail.co.uk/news/article-2514102/How-turning-central-heating-slim-Study-uncovers-link-higher-temperatures-lower-levels-fat.html

However, the story makes no attempt to tackle the issue of cause and effect. It suggests that the researchers are claiming that a warmer household causes the inhabitants to avoid becoming obese.(*)

It would be more logical, to me, if they claimed that non-obese people are more prone to feel the cold and hence turn their heating up higher.

So, as per the thread title, do any of you think this will end up labelled as Bad Science? Do you expect a contrdictory piece of research to emerge within the next two or three years?


* or at least there are no claims that weight loss is achievable - that would have been even more headline-grabbing, especially in the face of the exhorbitant fuel pricing, of late.

[Hypognosis]

//I don't believe that the editorial staff of the Daily Mail read the journal "Obesity"! //

Neither do I, SIQ. It appears to be google-proof and a paper-only journal would require actual leg movements to be made, in the direction of the nearest academic-type library.

I reckon I'd have to drive 40 to 60 miles (each way), for the chance to read the publication myself. Not that I'd go to such lengths on an idle curiosity basis. ;-)

[solvitquick]

This is bad science!
In any comparison of two or more groups of people, accurate balancing of the groups is critical. Failure to do so will result in a difference but this is likely be a spurious difference.
The overall health of each of the subjects is critical in a study like this e.g. thyroid activity effects cold-tolerance and body weight; body weight is also affected by Type 2 diabetes, a growing health problem. Similarly the Basal Metabolic Rate of the subjects, at start at least, should have been measured.
In the study described it must have been impossible to ensure that overall average health etc., was balanced between the groups - 100,000 over 13 years! Wow.
In my research career, spending massive amounts to carry out an ill-designed study and hence a speculative result was just unacceptable.
The overall research/budgetary managers should have seen that the money was spent on clear-cut, well-designed research leading to a clear "yes", "no" conclusion.
SIQ.

[Hypognosis]

Thanks, SIQ.

I have to say that I was impressed by the magnitude of the study - it says 100,000 households were studied, of which 15,000 ran their CH on a thermostat setting of 23C or above. That stands on its own as a revelation.

The researchers are somehwat caught in a bind because, to perform it as 'proper science', it would have to be done, like you decribe, under controlled circumstances, using identikit houses and health-screened subjects.

However, human factor steps in: No-one would want to live like a lab rat for upwards of 10 years, let alone the hundreds or thousands (entire families) required to accumulate some robust statistics.

The cludgy alternative is what we have presented here - real-life houses, of diverse designs with the 'constant' being taken to be the thermostat setting.

Google search throws out numerous studies correlating the rise of central heating use in developed countries with increasing obesity (1950s onwards). Were it not for the fact that this study has been running for a decade, it would be tempting that they set out specifically to counter such findings, whereas it seems that the "central heating makes you fat" concept has only recently got into circulation.

(And, if I was speculating about such things 20 years ago but saying nothing because I couldn't produce proof then, I guess others did, as well).

[Cloverjo]

I'm sure there was something on QI in the last couple of years that disputed Pixie's statement about mini skirts and cellulite. It doesn't get cold enough in many countries for that to have an effect. Hardly any women wear miniskirts nowadays, and those that do usually wear tights, so I don't think you could prove the correlation. And anyway miniskirts usually cover the thighs, which is where cellulite tends to be. It sounds like an old wives' tale that mothers told their daughters in order to get them to wear something less revealing.

[Hypognosis]

//It sounds like an old wives' tale that mothers told their daughters in order to get them to wear something less revealing. //

Haha. Now that's a statement I'm willing to accept without corroborating data, Jo.
(Unscientific, perhaps, but pragmatic.) ;-)

[solvitquick]

Dear Hypo,
Point about impossible to cage-up families taken.
Stick with my "bad science" as answer though.
Additions to my initial argument:
(a) Best done by epidemiology.
E.g. parts of Canada against each other. Not all Canadian houses have central heating like Winnipeg but Vancouver Island has mainly open fires to heat relatively small timber-built houses in my experience.
(b) What are we meant to do about such results? Turning down the thermostat is the current mantra anyway - so do we reverse this advice?
Nope - the study, correct or not, has no practicable point!
(c) The location of the thermostat varies - should be in the hall but ours is in the lounge! Thermostats, unless expertly calibrated,, are not perfect analogue devices. So your setting of 20 is probably firing the boiler to provide a different heat than ours - and neither guaranteeing 20 deg. C.
(d) Volunteers in a study are commonly paid - "not as an incentive to join" but as an "honorarium" for "inconvenience and/or cost of taking part". But the effects are the same - one cannot trust the reason for volunteering and cannot bind them to obey the protocol.
Waste of research funds, brains and time.
Utter rubbish - from the intelligent Scots of all people.
SIQ.