I look marvelous, but I don’t feel marvelous. Which is hokie-dokie for me, because, as you know, my credo is “It is better to look good than to feel good.”
In 1948 some investigators from Boston University decided to take on conventional wisdom. They went to a nearby town (Framingham, a 20 minute train ride away) and talked everyone in the town into giving a little of their personal data, blood, and health habits. They did this because conventional wisdom was that heart disease was brought on by personality (“Type A”) and plain bad luck. What they found was not only was personality non-contributory in and of itself, but they discovered a whole host of factors that actually did contribute to heart disease and were modifiable. The modern treatment of hypertension, diabetes, and high cholesterol, as well as the importance of physical activity and smoking cessation to the prevention of heart disease, all came form this study. The identification of the cause of 1/3 of strokes, atrial fibrillation, was made as a result of these serial observations as well. It has been estimated that the number of lives impacted as a result of these findings is in the millions and the amount of money saved in the billions.
These types of observational studies are important. Health care providers, stuck in the forest as we are, need someone to point out the trees. These types of observational studies on large populations are an important aspect of the forest-tree dialogue. Unfortunately the are also very expensive (current costs are several million dollars annually). In addition, the timeline is measured in decades making it an academic career killer for the first 20 years or so.
It appears that some folks in Britain have taken up the longitudinal study baton. The UK Biobank has taken information on 500,000 folks (with their permission) and committed to tracking these people over time. The first major report from this data bank came out in the Lancet this week. Having almost a decade of data, they looked at who died and what was associated with death. More usefully, they have put the data on a website so you can see whether or not to purchase a long-playing record.
They only enrolled folks between 37 and 70 and of the 500,000 only 2% died. Because of the length of the study, the model only works for 5 years out (Framingham has good data for 10 years out for most things they look at and longer for some). For guys, they found that if a guy said his health was excellent, it was. That was the single best predictor of 5 year mortality. For women, a history of cancer was associated with an increased risk of death, For both men and women, a slow self reported walking pace was associated with an increased risk of death. A history of smoking? Still bad news for both sexes and current smoker even worse. Things that didn’t have an effect at 5 years? Blood pressure, Average monthly beers. Cell phone use. Beef intake.
I fully expect more to come out of this effort. As of today, though, my UBBER age is 40 and I feel marvelous. What’s yours?