Digesting the evidence
A version of this commentary appeared in the Vancouver Sun, Epoch Times and Kevin MD
Many organizations cook up recommendations for the number of fruit and vegetable servings you should eat every day. Well known examples include Canada’s Food Guide which says you need roughly 7-10 servings, the American Heart Association which recommends 8-10 and the Harvard School of Public Health suggests somewhere between five and 13 servings of fruit and vegetables. While recipes and recommendations vary somewhat based on the specific food item, one serving is roughly equivalent to ½ cup or 75 g.
So just how many servings of fruits and vegetables should we eat a day?
To add to the numbers game, the media has recently reported on two different studies examining the impact of fruit and vegetable consumption on mortality — with two quite different conclusions. The first, from the Journal of Epidemiology and Community Health was reported as suggesting we need more than seven servings of fruit and vegetables a day while the second study published in the British Medical Journal suggests five is enough. What gives?
First, I think it’s important to clarify how nutrition studies are done. Studies that look at the impact nutrition has on mortality, cancer and cardiovascular disease are almost always cohort studies. This means 1000s of people are asked to fill out a questionnaire or are interviewed about their daily food intake and then researchers estimate the number of servings per food type consumed on a daily basis. The same people are then followed for anywhere from 5-20 years and a comparison is made between the number of servings of specific food items and the outcomes mentioned above.
As I’m sure you can figure out, this type of study design could potentially lead to somewhat half-baked results because people are unlikely to be able to accurately recall their food intake and will often respond to questions in a way that makes them look as good as possible. Nutrition studies can also easily be influenced by confounding variables — people who eat more fruits and vegetables may also exercise more or smoke less. While researchers try to control for all of these factors, it is not possible to completely eliminate the problem.
Nonetheless, let’s go with the assumption that the two most recently reported nutrition studies truly reflect a cause and effect relation between the number of servings of fruit and vegetables and mortality. So what did they actually reveal?
Fortunately, the studies provide enough numbers to allow one to plot servings vs. mortality curves for both studies on a single graph. And here is where it gets interesting. While the media reported differing conclusions, surprisingly, the curves from the two studies when put on the same graph are almost superimposable — you might say as similar as garbanzo beans and chickpeas. In other words, the only real difference between the study findings appears to be how the researchers and the media chose to report them.
The graphs for both studies clearly show mortality steadily drops from a baseline of eating zero servings of fruits and vegetables down to five servings per day where a roughly 25-30% relative reduction in mortality is seen. Once you get to five servings a day, the curves for both studies are basically as flat as a pancake, or in other words, no additional relevant reduction in mortality is seen as you further increase the number of servings of fruits and vegetables.
Importantly, the data also clearly show you get a benefit even if you don’t eat five servings a day. One serving a day gives you very roughly a 10% relative mortality benefit, two servings, a 15% benefit, three servings, a 20% benefit, four servings, a 25% benefit — and then once you get to five servings, that is basically it.
As further confirmation, an earlier study in 2013 from the Netherlands also shows basically the same results. There are similar data for cardiovascular disease, and it is very unlikely that any single new studies will substantially change these findings.
So how many servings a day should you eat? Well if it was me, I would base it on the evidence, how much I like fruits and vegetables, and overall, on how eating fruit and vegetables make me feel. In other words, use common sense.
James McCormack is an expert advisor with EvidenceNetwork.ca and professor with the Faculty of Pharmaceutical Sciences at the University of British Columbia. With Dr. Mike Allan, he hosts a weekly show — Best Science Medicine Podcast — that is regularly rated one of the top medical podcasts from iTunes; also available at therapeuticseducation.org.
August 2014