Overall, they found that personality and social relations in childhood significantly predicted risk of mortality many years later. The findings were not what you may expect though.

Those who were the most cheerful and optimistic, on average, lived shorter lives than those who were less cheerful and joking. According to the researchers, optimistic people tended to take more risks overall: going to more parties, using more drugs and alcohol, and getting into more accidents. Friedman notes that "fun can be overrated."

Who lived the longest? Those who were the most conscientious and committed to their jobs, friends, and community lived the longest. In fact, those working the longest throughout their lives (even taking part time jobs after retirement) and working in the most stressful jobs lived the longest. Those working in low-status jobs were far more likely to die before the age of 60 than those working in higher status jobs. According to the researchers,

"It was the most prudent and persistent individuals who stayed healthiest and lived the longest."

They also found an effect of divorce. While early parental loss didn't have an effect on longevity, early parental divorce was a very strong predictor of mortality in adulthood. The authors note how traumatic and painful divorce was for the children.There were also important gender differences. Men who remarried improved their odds of a long life, whereas women who stayed single after divorce were just about as well off as if they had stayed single.

Exercise also played a role, but not the kind you may think. People who lived the longest weren't obsessed with health and exercise. They didn't have structured regimes, but just tried to live as active a life as they could.

Also counterintuitively, kids who started first grade at too early an age had problems later in life and lived shorter lives. Early school entry was associated with less educational attainment and worse midlife adjustment. Also, while early reading ability was associated with academic success, precocious reading was less associated with lifelong educational attainment and was hardly related to midlife adjustment at all. Parents may want to re-think whether they push their child to enter school too soon.

In sum, the authors argue that people who follow simple, straightforward patterns of behavior can lead a long life.

Of course, there are reasons to be skeptical of the study. The average IQ of the sample is extremely high-- a whopping 151! And as I mention in The Truth About the "Termites", Terman's group consisted of only a few minority students (to be precise, he included 4 Japanese students, 1 black child, 1 Indian child, and 1 Mexican child in a total sample of 168,000). Teachers at that time most certainly had a bias toward identifying white students with talent.

In reality though, I think the characteristics of their sample actually underestimate the true effect of traits like IQ and Conscientiousness on mortality. Their sample consists of a "restricted range", in that there is very little variation in IQ scores. This actually makes it all the more remarkable that they found such relationships within such a narrow band of IQ scores and demographics.

In fact, the rapidly advancing field of cognitive epedemiology is showing that across a much broader range of IQ levels and demographics, cognitive ability predicts mortality, even after controlling for a number of related variables such as education and socioeconomic status. This research is pointing to the inescapable conclusion that cognition is related to health and longevity. Of course, the causal path is still unclear, but research in the coming years will get us closer to understanding why there is such a strong relationship.

The relation between IQ and the personality trait of Conscientiousness isn't terribly high, but IQ is related to measures of self-control, including the ability to to delay immediate gratification for longer-term gains (called "delay discounting"). This is entirely consistent with Friedman and Martin's idea that those who live longer take less risks.

And these findings are also consistent with other research showing that those who live a fast life have lower levels of self-control, choose immediate rewards over longer-term benefits, take more risks, and ultimately die younger. The causes of living a fast life seem to include a combination of genetic and environmental risk factors, including living in harsh and unstable environments, father absence and parental discord. All of these risk factors make it more likely that a person will show reduced executive functions in his or her brain, areas that correlate with IQ (but are not synonymous with IQ) and contribute to long-term planning.

So what does this all mean? Perhaps the conclusion is not all that counterintuitive after all. In fact, the conclusion seems pretty banal: self-control and interpersonal stability leads to a long-life. The finding that optimistic and cheerful people died younger is surpising though-- and worrisome. Personally, I'd rather enjoy my life and cut off a few years than live a longer life devoid of mirth. All work and no play makes a dull existence (unless, of course, work is play).