Auto Ads by Adsense

Friday, March 09, 2012

Review: The Longevity Project

The Longevity Project is a statistical study of the so-called "Termites", about 1500 high-potential Californians chosen by Professor Terman to follow over their lifetimes. The authors draw conclusions based on personality and life-events versus longevity.

Based on the data, they conclude several items of interest:
  • Conscientious people live the longest. A lot of it is because conscientious types actually follow doctors' orders and take their pills when they're supposed to and so on.
  • Social types actually don't live longer, because the benefits of having a stronger social network is offset by picking up bad habits like smoking and drinking. That means engineers and scientists actually live longer than the sales types.
  • Mild worriers actually live longer than the happy types, because they'll be proactive about health problems rather than ignoring them.
  • Starting school early is predictive of a shorter life, while skipping grades has no effect on longevity. They speculate that the loss of unstructured play time is really harmful.
  • Parental divorce takes 5 years off your life. In fact, it's better that one of your parents died than that they got a divorce. The exception is that life at home is so bad that all the damage has already been done.
  • Maintaining or increasing your activity levels through midlife is predictive of a longer life. The authors note that if a you spent 2 years over your life time exercising and gain 2 years of longevity, you only broke even from all the exercise. So the best deal is if you mostly became a couch potato and only exercised just enough to get maximum benefits.
  • Being married gave you the longest life, but only if you didn't get a divorce. Divorce is so traumatic that it reduces your life span. Even getting remarried later doesn't help as much. For women, it's better to stay single than to get a divorce and then remarry.
  • Being a top dog and high achiever causes you to live longer.
  • Religion makes you live longer, but mostly because of the social connections and having an active social life, rather than the prayer and meditation.
All in all, this is pretty impressive. Unfortunately, the authors fail to point out many of the obvious flaws in the study:
  1. The study pretty much consists of middle class, white Californians. That homogeneous sample means that if you're Asian, Black, or other ethnicity, the results might or might not apply.
  2. The study shows correlation. The authors do a great job of trying to tease out the underlying cause, and in some cases, they're quite believable, for instance, with respect to religion. For other parts of the study, correlation does not mean causation and you'd have a really tough time figuring things out.
  3. The study was a longitudinal study covering many decades. However, during that time, technology and social norms evolved. It could very well be that conclusions based on people who were born at the beginning of the 20th century would not apply to people who are born now, or who were born in the middle of the 20th century. For instance, do the conclusions about marriage apply to gay marriage? Are no-fault divorces as devastating to the spouses? This study couldn't answer such questions.
  4. Is 1500 people enough of a sample to truly draw such conclusions? The authors don't actually go into sufficient technical detail about their statistical methods to make me feel comfortable with their conclusions.
I'm happy to recommend this book as food for thought, but take their conclusions with several tablespoons of salt: I'm fairly sure they're not as cut and dried as the authors claim they are.

No comments: