« Multi-Cultural Math - Teaching the Teachers | Gene Expression Front Page | Much ado about women & Larry Summers »
January 22, 2005

Reaction Time & Mortality

New article out in the latest PS by I. Deary showing that individual differences in performance on chronometric tasks are highly related to mortality even after controlling for smoking, education, and social class.

Related (Thanks to Scott)

New study showing that High IQ mediates liklihood of suicide.

Alex's Comment

It has been known for decades that g mediates the relationship between [name your (psycho)pathology] and outcomes, but the Environmentalists in Psychology (or, name your own discipline) have long inisisted that it is the SES that (often) accompanies g that mediates the outcome. I hope this line of inquiry (i.e., g-->health outcome, inspite fo SES) continues.

This important area of research requires an open mind regarding potential explanatory variables, mechanisms, and direction of causation. It is tempting to posit that cognitive ability in old age relates prospectively to mortality because the brain sensitively reflects deteriorations in the state of the body. In this common-cause hypothesis, the association is more correlation than causation: A deteriorating brain is a part of a body that is deteriorating generally . . . This hypothesis implies that reaction time might be able to pick up bodily deterioration earlier than the terminal decline found in psychometric tests of cognition preceding death. However, this is not the whole story, because IQ of healthy 11-year-olds predicts survival almost 70 years later just as well as IQ of 56-year-olds predicted survival in this 14-year prospective study. Thus, there is something traitlike as well as possibly statelike about cognitive ability that offers a clue to longevity.

Middle-aged subjects provide an interesting case. Here, we found that psychometric intelligence and reaction times were significantly related to survival over the next 14 years, with an effect size comparable to that of smoking status. But what is the direction of explanation? Social class, education, and smoking did not explain the association, so perhaps the present results replicate the 11-year-old effect, reflecting traitlike aspects of even healthy brains that correlate with survival. Or did the IQ test and reaction time indices sensitively pick up preclinical decline in physiological mechanisms that subserve good health and survival? Findings in children support the first explanation, and findings in older people support the second. But perhaps such a trait-state dichotomy is false. It might be the case that, even in children, lower IQ relates to earlier death partly because it is a reflection of a body with suboptimal physiological integrity. This possibility is consistent with the finding that the association between lower IQ and earlier death is especially strong in the lowest quartile of IQ scores . . .

I am sure that most of you will be shocked that performance on chronometric tasks is partially heritable, and that we might even have located a specific gene that is implicated in variance of chronometric performance. Moreover, I am sure you will be aghast at the fact that individual differences in g play a significant part in the variance in chronometric task performance (e.g., 1, 2).

I guess you can add this to the pile of data suggesting not only that intelligence/g has a biological component, but that it also confers a procreational advantage.

Posted by A. Beaujean at 09:56 PM