Miracle Nuts

We’ve been treated this week to a remarkable assortment of headlines proclaiming that just half a handful of nuts can save us all from an early death. Nut marketers rejoice!

This joyous news comes from an observational study in the International Journal of Epidemiology that neglects to discuss the limitations of this study’s observational design and its use of self-reported dietary data. The study concludes that “nut intake was related to lower overall and cause-specific mortality.”

Taking this a step further, the authors issued a press release saying, “Nuts and peanuts, but not peanut butter, may protect against death from cancer, heart disease, respiratory disease and other major causes.”

The next step was inevitable: a flood of headlines about miracle nuts. With so many to chose from it’s hard to pick a favorite. But the Toronto Sun wins our acclaim for hyperbole with “A Handful of Nuts Protects Against Death.”

Two catches render these claims false. The first is that this study didn’t capture any real data on eating nuts. It only captured data on people saying they ate nuts. Gallons of ink have already been spilled explaining why it is that what people report eating is simply not a reliable measure of what they actually ate. It’s hopelessly distorted by feeble memories and feelings about what people think they’re supposed to report. If you want all the details, click here and here.

The second problem is a bit more obvious. An association between claiming to eat nuts and living longer proves nothing about cause and effect. Period.

Oh nuts! We can’t count on nuts to protect us from death.

Click here to read more in the International Journal of Epidemiology.

Nuts! Photograph © Paul Moody / flickr

Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.


4 Responses to “Miracle Nuts”

  1. June 15, 2015 at 8:04 am, Mary-Jo said:

    For the past couple years, I’ve been feeling very concerned about the hype about nuts. Imo, people are now eating too much/many nuts. And with all these misleading headlines, not to mention ‘studies’, no wonder? And human nature being what it is, if a ‘little’ is touted as being so miraculously good for you, well, ‘alot’ must be even better! Alot of nuts are NOT helpful for many people. They are SO high in fats (ok, the good fats, but still!) and calories, they can totally undermine someone who needs to watch dietary intake of these macronutrients. So, I’m totally on board with a re-examination on the whole nutty issue. I’m frequently helping clients understand the facts about nuts so they don’t over-snack and consume them ‘ad libitum’ because they think it’s helping them save their lives!

    • June 15, 2015 at 9:43 am, Ted said:

      I’m with you Mary-Jo. Nuts are neither manna nor poison. They’re just a food that was demonized for a while. Now it’s being hyped with excessive health claims. Middle ground, please.

  2. June 16, 2015 at 8:27 pm, Michael Guth said:

    Thank you for taking the time to point out the shortcomings in the Journal of Epidemiology study. It is baffling that the authors would undertake such sophisticated analysis to include “multivariate case-cohort analyses” yielding “restricted cubic splines” showing “nonlinear dose-response relationships with mortality.” But it never dawned on the principal investigators that nut consumption reported in a 1986 survey might not reflect actual nut consumption for 1986, and worse, might not and probably would not have accurately reflected nut consumption in years 1987 – 1996. In other words, the nut consumption would have been uniform across all those years only if the thousands of respondents had highly inelastic demand for nuts. Otherwise, they would have substituted other foods for nuts so that actual nut consumption would oscillate.

    • June 17, 2015 at 4:18 am, Ted said:

      Well-said, Michael. Journals routinely publish studies that describe self-reports of food consumption as if they represent actual food consumption. They don’t. Not at all.