5

A report on an Australian news site claims that:

Eight-year study finds heavy French fry eaters have ‘double’ the chance of death

The report quotes some of the details of the research:

On average, we each eat roughly about 14kg of hot fried potato chips every year. And that’s a problem.

Research published in the American Journal of Clinical Nutrition shows that those who eat fried potato two or more times each week have double the chance of dying when compared to those who eat none.

The numbers aren’t insignificant. Some 4440 people aged 45 to 79 were tracked over an eight-year period, during which 236 of them died.

The source is this American Journal of Clinical Nutrition paper.

Given the notorious unreliability of nutritional studies and the complexity of models adjusting for other risk factors, is this result credible or significant? And is it consistent with other studies?

Community
  • 1
matt_black
  • 56,186
  • 16
  • 175
  • 373
  • an aside ... "most epidemiology studies are wrong"... https://skeptics.stackexchange.com/q/1973/5125 – GEdgar Jul 16 '17 at 11:21
  • 3
    The claim 'those who eat fried potato two or more times each week have double the chance of dying when compared to those who eat none' is obviously wrong. Even people who are *not* eating french fries will with definitive certainty die. Doubling a 100% risk of something is not possible. – Tor-Einar Jarnbjo Jul 16 '17 at 12:04
  • 2
    @Tor-EinarJarnbjo: The newpaper article suggests in the body that this was over an 8 year period. – Oddthinking Jul 16 '17 at 12:47
  • Have you read the paper? – Nate Eldredge Jul 16 '17 at 14:02
  • 2
    @NateEldredge Only the abstract. There is a limit to how much science I can afford to pay for. – matt_black Jul 16 '17 at 14:03
  • @Oddthinking We should probably clarify the claim then: is it that the chance doubled over that 8 year period? Is even that a meaningful statement given the complexities of actuarial science? – IMSoP Jul 16 '17 at 14:22
  • @IMSoP the claim is reproduced as it appears in headlines. The right place to address ambiguity in what it means is in an *answer*. Otherwise we are not addressing the popular claim. – matt_black Jul 16 '17 at 14:33
  • @matt_black But as Tor Einar-Jarnbjo points out "double the chance of death" is meaningless, so trivially refutable by saying "everyone dies" without any reference to the study at all. At the other extreme we have "what did this study claim?" or "was this a well designed study?" which isn't really a skeptical enquiry, just scientific review. – IMSoP Jul 16 '17 at 14:43
  • @IMSoP My point is that we need the address claims in the way they are presented to the public or we won't be connected to the problem. A good skeptical analysis here would point out the meaninglessness of the apparent claim, but also point out what it should have said and then address the evidence relevant to that. Which won't be a restatement of what the paper says as it also deserves skepticism. – matt_black Jul 16 '17 at 14:56
  • The abstract includes the statement *"After adjustment for 14 potential baseline confounders, and taking those with the lowest consumption of potatoes as the reference group, participants with the highest consumption of potatoes did not show an increased risk of overall mortality (HR: 1.11; 95% CI: 0.65, 1.91)"* so it is apparently regular but not excessive consumption which showed increased mortality – Henry Jul 16 '17 at 16:19
  • 2
    I hate these claims. There's a million various things heavy french fry eaters might do that affects their longevity. –  Jul 17 '17 at 00:27
  • It's probably just some correlation but not necessarily a causality. People who consume more french fries than the average person probably eat generally more fast food than the average person. Fast food causes overweight and lack of exercise which can cause dying earlier. Of course I cannot prove this but that's my interpretarion on the data. – BlueWizard Jul 22 '17 at 21:00

1 Answers1

6

The result is not credible.

The abstract claims, somewhat disingenuously, that "this trial was registered at clinicaltrials.gov as NCT00080171." But the registration is actually titled Osteoarthritis Initiative (OAI): A Knee Health Study and sponsored by the National Institute of Arthritis and Musculoskeletal and Skin Diseases. It was not registered as a nutritional trial.

The abstract says: "Potato consumption was analyzed by using a Block Brief 2000 food-frequency questionnaire." But the registration is clear on the purpose of the questionnaire: "A questionnaire and physical examination at screening will assess for risk factors for the development and progression of knee OA."

The nutritional information was collected to facilitate the study of osteoarthritis, not the lifespan of the general population. The registration makes no mention of mortality at all. And participation was limited to those at high risk for developing osteoarthritis of the knee.

D Krueger
  • 261
  • 1
  • 2