Lizard tongue, eye of newt, and playing tennis lengthen life

A new study was published in Mayo Clinic Proceedings titled Various Leisure-Time Physical Activities Associated With Widely Divergent Life Expectancies: The Copenhagen City Heart Study. In this study, 8577 people completed a questionnaire asking them about exercise habits. They were then followed up to see what became of them up to 25 years later.

Compared to people who led sedentary lives, life expanse categorized by participation in various sports showed that people who played:

Tennis lived 9.7 years longer
Badminton lived 6.2 years longer
Soccer lived 4.7 years longer
Cycling lived 3.7 years longer
Swimming lived 3.4 years longer
Jogging lived 3.2 years longer
Calisthenics lived 3.1 years longer
Health club activities lived 1.5 years longer

No surprise, media issued reports like this from Forbes: Want to Live Longer? Take Up Tennis. Or this from the New York Times: The Best Sport for a Longer Life? Try Tennis.

Thankfully, the research group publishing the data stated outright that the data are observational and cannot therefore be used to determine cause and effect relationships: You can NOT say that playing tennis lengthens life by nearly 10 years; you can only say there may be an association of tennis and longevity that may or may not be causal.

If you’ve been following my Undoctored conversations lately, such as this commentary on the work of Dr. John Ioannidis and the hazards and misguided conclusions of observational studies, you will immediately recognize that the headlines are highly likely to not be true: It is highly unlikely that, by playing singles or doubles tennis, you will live an additional 10 years, nor will jogging add 3+ years to your life.

This is because of numerous confounding factors, i.e., factors besides sport-of-choice that could play a role, that make such associations unlikely:

• Tennis is often played by people with higher incomes and thereby have access to better food and are less likely to smoke cigarettes, take illicit drugs, or engage in other high-risk behaviors. Sure, it could be that tennis is more social, but it probably has little or nothing to do with that.
• If you play badminton, you likely don’t live in the inner city, don’t live in a trailer park, and probably don’t smoke cigarettes or eat twice a week at KFC.
• If you are well enough to play soccer into your 30s, 40s, or onwards, you likely don’t have rheumatoid arthritis, obesity, type 2 diabetes, or leg edema. In other words, the rigor of the sport deselects people who are unable to participate, making it appear that soccer per se is associated with better health. But, no, it just means that the rigorous demands of the sport put it out of reach for less healthy people. Likewise, jogging and swimming: If you are unhealthy for other reasons, you probably won’t run 3 miles or swim 100 laps.
• Join a health club–that costs money. It means that people who make health a priority and/or have more disposable income to spend on things apart from mortgage/rent, food, and basic life necessities join health clubs that typically cost $100 per month or more. Membership in a health club identifies you as being different from people who are not members.

Are you beginning to see the deeply flawed nature of observational data? This is why I say that observational data are about as good as no data at all. Yet observational data forms the basis of national government dietary advice. Observational data serve as the basis of the USDA’s food plate and pyramid, the American Heart Association’s TLC diet, the advice provided by your friendly neighborhood dietitian, and the stuff of media headlines—but it is all little better than fairy tales and parlor magic. Observational studies, such as those generated by Harvard researchers Dr. Frank Sacks and Walter Willet through the Nurses’ Health Study and Physicians’ Health Study are nonetheless embraced as fact, are used to determine what your child eats for lunch at school, the meals that the military are served, and yield headlines that attract attention.

Surely we need better science to divine how we manage diet. But the last source of dietary advice you should accept are the people who purport to understand diet based on the house of cards that come from observational data. The real study that would have to be performed to explore this question would involve randomly assigning people to playing tennis, soccer, cycling, or jogging every day for 10 or more years and seeing what becomes of them—a virtually impossible study. Would you accept random assignment to playing badminton every day?

Here’s a nugget of wisdom that may help you navigate through the mess that is modern dietary advice: If primitive humans, adhering to a genetic dietary script crafted over millions of years did it, you should probably do it, too. If they did not do it, you probably shouldn’t do it, either. It means that liver, fat, and coconut are on the menu of health, but Coca Cola, Hot Pockets, and Tostitos are not.