Tuesday, March 20, 2012

Control issues

Happiness yoga image, Facebook
We have listened to a recent BBC radio interview on The Life Scientific with Michael Marmot, the epidemiologist who has published many books and papers showing that social status is correlated with health.  This is more than just that poor people live in dirtier, more dangerous places, and have dirtier, more dangerous lives, and eat junkier diets.

Marmot is know for his innovative, and well-replicated, studies, the "Whitehall studies", about the role of stress in health and longevity.  Before his work on British civil servants, it had universally been assumed that the boss has the most stressful job and that the price you pay for status and the higher pay grade is a lower health grade: more coronaries-under-stress.  Better to lay low, and stay healthy!

Well, that's not what his research has shown.  Instead, and to the contrary, the high stress of the high job is accompanied by better health and longer life!  It isn't that the boss has less stress, because the boss typically has more.  But the boss is in control, and that is the difference.

The bottom line is that health in this context is not an absence of stress but an absence of control.  Frustrating constraints are killers!  The problem, in Marmot's mind, is that our society gives most people little control over their lives, and he would like to see society adopt economic and social policies and conditions to liberate a large fraction of society from health-destroying servitude.

This is a particular issue, but is highly relevant to many topics we deal with here on MT, and they relate to what we believe are very simplistic ideas about genetic causation and how to find it.  If stress of the kind we're discussing here has a major effect on physiology that is associated with disease--or even death!--then clearly it isn't a genotype making the major difference.  More importantly,  epidemiologists can (claim or try or hope to) measure salary, diet, marital status, exercise, smoking, drinking, drug use, and so on.  But stress related to control of one's life experiences is much harder to measure, even if you think to look for it, and clearly would be omitted in the massive-scale case-control studies that are making up our current commitment to blimped-out GWAS studies.

There is no reason to think that a given genotype would have similar effects in environments that, overall, make substantial difference to overall health.  Even in regard to stress issues would similar effects be expected of a genotype if the person were a machinist, government minister, religious minister, railroad employee or school teacher?

When something so profound, perhaps fundamental, and subtle as feeling of control--as distinct from normal measures of life experience like education, salary, neighborhood and the like--can make such a difference then we need to adopt much more circumspect claims, and develop much better study designs, than is the case nowadays.  Of course, such more responsible studies are not in the cards because they'd be too much trouble to do.....

No comments: