Thursday, October 2, 2014

Ignore this study!

A piece in the New York Times reports that a new study shows that working long hours causes type 2 diabetes, but only in people of lower socioeconomic status.  The study is a meta-analysis of 4 previous studies and analysis of 19 unpublished studies, and is published in The Lancet: Diabetes and Endocrinology ("Long working hours, socioeconomic status, and the risk of incident type 2 diabetes: a meta-analysis of published and unpublished data from 222 120 individuals," Kivimäki et al.).
Type 2 diabetes, characterised by hyperglycaemia and insulin resistance or insulin insufficiency, causes substantial disease burden. Globally, more than 285 million people have type 2 diabetes, and its prevalence is predicted to increase to 439 million by 2030. The findings from prospective cohort studies show that working long hours is associated with factors that contribute to diabetes, such as unhealthy lifestyle, work stress, sleep disturbances, and depressive symptoms. Working long hours is also associated with an increased risk of cardiovascular disease, which is one of the complications of type 2 diabetes. However, the direct association between long working hours and incident type 2 diabetes has been assessed in only a few studies.
And they found that "In this meta-analysis, the link between longer working hours and type 2 diabetes was apparent only in individuals in the low socioeconomic status groups."  People of low SES who worked more than 55 hours a week were at a 30% higher risk of developing t2d than those who worked 35-40 hours a week.


October 24, 1940: The Fair Labor Standards Act of 1938’s mandate of a 40-hour work week with time-and-a half overtime pay for hours of work beyond that goes into effect. The legislation was passed to eliminate “labor conditions detrimental to the maintenance of the minimum standard of living necessary for health, efficiency, and the general well-being of workers.”

Which means, of course, that working long hours doesn't in fact have a direct effect, but is correlated with something else that's associated with living life in the lower socioeconomic strata.  Or at best working long hours exacerbates the effect of that unidentified, confounding variable, or variables.  The authors note that they adjusted for age, sex, obesity and physical activity, and excluded shift workers, and that the effect of long working hours was independent of these variables.  But, while they wonder what it is about working long hours that might be causal, they do also note that long working hours may be mediating the effect of some unknown variable.  

Indeed, there are many differences between socioeconomic strata that might be associated with risk of T2D including ethnicity and genetic predisposition, diet, maternal health during pregnancy, type of job and thus pay scale, and other possible risk variables associated with poverty.  So, it's interesting that what looks to us like an inconclusive study that suggests but doesn't identify confounding variables associated with risk of type 2 diabetes, the NYT piece emphases the effect of long working hours only, though acknowledging that this is associated with depression, sleep deprivation, unhealthy lifestyle, which may be causal.  Though, they do not question why that would be only in poor people.

Curious that the lead author is quoted in the NYT on T2d prevention this way:
“My recommendation for people who wish to decrease the risk of Type 2 diabetes,” he said, “applies both to individuals who work long hours and those who work standard hours: Eat and drink healthfully, exercise, avoid overweight, keep blood glucose and lipid levels within the normal range, and do not smoke.”
So, basically, he's saying ignore this study.

And here's another study we might want to ignore
Steven Salzburg at Forbes alerts us to another work-related danger.  Standing up at our desks is said to be good for our health.  Indeed, it's said even to reduce risk of type 2 diabetes (and presumably the benefit increases the longer we stand and work, unless of course we're lower class, in which case the longer we stand, the more likely we'll get type 2 diabetes).  The mechanism has now been found -- you know how telomere length (the length of the ends of our chromosomes) is associated with longevity?  The longer they are, they longer we live, right?  Well, it seems that standing up lengthens telomeres!

Telomeres; Wikpedia

An RCT (random control trial) study in the British Journal of Sports Medicine ("Stand up for health—avoiding sedentary behaviour might lengthen your telomeres: secondary outcomes from a physical activity RCT in older people," Sjögren et al.) reports that the less time their subjects spent sitting, the greater the lengthening of their telomeres six months from the inception of the study. "Reduced sitting time was associated with telomere lengthening in blood cells in sedentary, overweight 68-year-old individuals participating in a 6-month physical activity intervention trial."

But, as Salzburg points out, this result is based on 12 individuals, and in fact only 2 of these people showed a marked effect, and probably drove the results.  And what do blood cell telomeres have to do with diabetes or longevity, unless this has to do with the immune system?  We ask, because red blood cells have no chromosomes, and white blood cells are basically in the immune system.  So are some discarded cells sloughed off from other tissues (and hence no longer being used by the body). Where is a plausible mechanism, unless it has to do with lifestyle correlates of those who choose stand-up careers?  The authors owe us at least some explanation. Or is it the journal that owes its readers and explanation for why it published such a paper?

Anything can get published
There are good studies and there are bad studies.  Even good (legitimate) findings can be falsely attributed to some measured putative cause without sufficient justification.  The publication and promotion of loose, over-interpreted, over-sold studies is one reason that we don't know which foods are good for us and which are bad.  But the problem is deeper than that -- reductionist science, which aims to identify single causal factors for complex diseases, no matter how well done the study and expert the analysis, is simply the wrong approach to understanding complexity.  It is systematically misleading.

Why these reports keep flowing is understandable in our news-cycle media culture.  But when the bottom line is basically that a reasonable, moderated, balanced lifestyle is the best and almost the only reliably known way to defer many chronic diseases, it's strange that scientists themselves can't see the relative nonsense they are purveying.

No comments: