Walking, Fast and Slow and Looking for Lurking Variables

Two years back there was a report explaining longevity of Hong Kong residents. The study looked at food, hobby and work habits of those who lived past 70 and 80 years and declared

  1. It is the food
  2. It is their love for Mahjong
  3. It is because they like to travel
  4. It is because they don’t retire at 60 and continue to work

Fast forward to present day, we have a study from Berkeley that found this about walking fast and slow,

It turned out that nearly 2,000 of the walkers had died. More telling, these deaths disproportionately were clustered among the slowest walkers. Those in Category 4 were about 18 percent more likely to have died from any cause than those in the other three categories and were particularly vulnerable to deaths from heart disease and dementia.

And that is all everyone needed to declare walking fast is better than walking slow or even go a mile further to say, “walking fast leads to longer life”.

What do you see as common aspects between the Hong Kong study and the Berkeley study on longevity?

No one seems to be asking reasons for older people traveling and working  and slow walkers walking slow. Had they asked those questions they likely would have found that

  • Older people walk more, play mahjong and travel because they have time and need entertainment
  • Older people work late into 80s because they either need the money or excitement
  • Slow walkers walk slow because of some underlying conditions

You will see these studies raise these caveats near the end of their paper but will go no further to actually go look into the reasons. And media does not even bother reading those cursory exceptions, they are happy they can report on, “scientifically proven ways to live longer”.

Any study that makes causation claim or even hints at causation claim is more likely to get page views than that stops at reporting correlation. (Now you can parse the causation claim in the last statement).

If such is the case of peer reviewed papers what can you say about the works of social media scientists and their retweet advice?

One thought on “Walking, Fast and Slow and Looking for Lurking Variables

Comments are closed.