Diet: Is nutrition science a more reliable source of advice than your grandmother?

Meat consumption is bad for your health. It gives you cancer, heart attacks, stroke, you name it. Says nutrition science. And they must know. After all, it’s a science. Is it, really?

A few years ago, Jonathan Schoenfeld and John Ioannidis took a standard cookbook and randomly selected 50 frequently occurring ingredients (sugar, coffee, salt, etc.). They then carried out a systematic literature search, asking whether there were epidemiologicial studies that had investigated the cancer risk of these ingredients. And they found what they were looking for. For 80% of the ingredients at least one study existed, for many even several. Of 264 of these studies, 103 found that the ingredient investigated increased the risk of cancer, while 88 reduced the risk! So  after all Joe Jackson was right: ‘Everything gives you cancer’! But wait a minute: Milk? Veal? Orange juice?

And there is more advice from nutrition science: 12 hazelnuts a day increase life expectancy by 12 years, that is one year per nut! Alternatively, you can drink 3 coffee cups a day, because this leads to a similar life extension. A tangerine per day is less effective: only 5 years added. Be careful with eggs: One egg a day and you might loose 6 years, while 2 slices of bacon subtract 10 years of your lifespan. Not even chain smoking is that dangerous! One arrives at such projections when accepting the analysis of nutritional epidemiology and taking their causal rhetoric serious.

Fortunately, you might argue, nutrition science has more plausible results in store. Like the fact that a mediterranean diet, which is rich in fish, olive oil, red wine etc., is good for your heart. Not only does it taste well, with it one also lives longer enjoying a better health. Thus were the implications of a number of epidemiological studies, and one large interventional randomized controlled trial (PREDIMED), published in the renowned New England journal! But did you know that this study had to be withdrawn? Because there had been serious protocol violations, and the data may have been manipulated. In addition, no Mediterranean diet was tested at all, rather a dietary supplementation.

The related ‘French Paradox’ has suffered a similar fate in recent years. You might recall that despite a higher consumption of saturated fats (cheese!), French people appear to have a relatively low risk of coronary heart disease, especially compared to British people. It must be the red wine that the French love so much! In media speak: red wine protects against coronary heart disease. Medical advice doesn’t get better than that! In the years following the publication of the paradox and further studies in its wake the consumption of red wine exploded, in particular in the US. Inspired by the good news basic researchers took to their labs: legions of mice were made drunk, and isolated arteries bathed in various alcoholic drinks or extracts from them. Thirty years after the first description and hundreds of studies later, nothing is left of the euphoria. It turns out that the paradox is probably an artefact of the different recording of heart diseases in France and the UK and a delayed change in eating habits in both countries. With the vanishing of the paradox talk of beneficial effects of red wine has almost disappeard. In scientific circles,  the French paradox has disappeared even from dinner conversations.

In essence, the purported protective effect appears to have been a statistical artefact, in particular caused by chosing the wrong groups for comparision, and insufficient correction of confounders. This year, Chinese researchers finally burried the story of the protective effect of moderate amounts of alcohol. A mega-study with 500,000 participants, 10 years of follow-up, genotyping, etc. showed that every drop of alcohol is just one too many from a health point of view. Remember: Results that sound too good to be true are often just that, not true!

Despite these blows, nutrition research is not at a loss for new health advice. Now omega-3 fatty acids are supposed to bring us healthy into old age! Let’s just wait a while before we embark on a cod-liver diet.

But what about the ill effects of red meat consumption mentioned above? Recently several very large meta-analyses were published, all in one issue of the Annals of Internal Medicine (here, here, here). The verdict: The influence of meat consumption on total mortality or cardiovascular outcomes, if it exists at all, is small!

The list of associations of certain diets with health, disease, increased or decreased life expectancy is almost endless. Sometimes it is the same diet that results in opposite outcomes when studied by different teams. Almost always a causal relationship is inferred or suggested from the claimed association. Correlation of the consumption of food X with a certain outcome Y translates to consumption of X causes disease Y. Everyone knows how many factors influence our eating habits, many of them intrinsically connected and thus inseparable. In its totality, what we eat certainly has tremendous influence on our health. But individual ingredients usually play a minor role. On the other hand, the socio-economic status, in other words: How much someone has to live, is a major factor. A legendary study once examined the contents of shopping bags at the supermarket checkout: Not surprisingly, beer, vodka, canned meat and cigarettes clustered in some bags, while red wine, olive oil, salad and muesli in others. Did you know that in Germany, according to the official Federal Health Report, the difference in life expectancy between the lowest and highest income groups is 13.3 years for women and 14.3 years for men? Along the same lines, but even more dramatic: Between the final stops of a subway line in Chicago (Red Line), life expectancy gradually decreases by 30 years from north (where the wealthy live) to south (where ‘racial gettos’ abound)! It must be the olive oil!

Nutrition researchers cannot be blamed for living conditions, and the fact that these strongly interact with eating habits. And that lots of factors, including genetic ones, interact in complex ways. It is also not their fault (at least in most cases) if their results are exaggerated or distorted in the media. Almost every major nutritional study involving a common food item makes it into the lay press, often in a lurid way. Nor can nutrition science be blamed for the fact that in their field it is difficult to conduct randomly controlled, prospective interventions. But nutrition science is responsible for methodological shortcomings, some of which were already mentioned above. All this is further complicated by the presence of nutrition scientists in committees which often on the basis of weak evidence give far-reaching dietary recommendations. Conflicts of interest abound which affect the design, analysis as well as the interpretation of such studies. The influence of the food industry on medical science is at least as strong as that of the pharmaceutical industry. And that’s quite something!

If we forget about the hype, the mix-up of causality and correlation, the overestimation of effect sizes and the underestimation of interactions and confounders, what kind of dietary advice does nutrition science give: We should eat a varied and balanced diet. Fruit and salad, a little bit of meat, some but not too much fat – everything in moderation. An omnivore’s diet. Just what our grandmothers have told us. And because we are much less physically active than our ancestors a few hundred thousand years ago: Not too many calories, with a healthy dose of exercise. As Johann Wolfgang von Goethe put : ‘Only through moderation do we survive’.

But is that science?

 

A German version of this post has been published as part of my monthly column in the Laborjournalhttp://www.laborjournal-archiv.de/epaper/LJ_19_12/20/index.html

 

 

Additional references

Joe Jackson. Everything gives you cancer. On the LP: Night and Day 1982  https://www.youtube.com/watch?v=EM2MGN_w4sU

Resnick B and Belluz J. A top Cornell food researcher has had 15 studies retracted. That’s a lot. Brian Wansink is a cautionary tale in bad incentives in science. https://www.vox.com/science-and-health/2018/9/19/17879102/brian-wansink-cornell-food-brand-lab-retractions-jama

Kolata G. Eat Less Red Meat, Scientists Said. Now Some Believe That Was Bad Advice.New York Times, 30.9.2019 https://www.nytimes.com/2019/09/30/health/red-meat-heart-cancer.html

Schwingshackl L, Schwedhelm C, Hoffmann G, et al. et al. Food groups and risk of all-cause mortality: a systematic review and meta-analysis of prospective studies. Am J Clin Nutr. 2017;105(6): 1462-1473.

Poole R, Kennedy OJ, Roderick P, et al. Coffee consumption and health: umbrella review of meta-analyses of multiple health outcomes. BMJ. 2017;359:j5024.

Knott CS, Coombs N, Stamatakis E, Biddulph JP. All cause mortality and the case for age specific alcohol consumption guidelines: pooled analyses of up to 10 population based cohorts.  BMJ. 2015 Feb 10;350:h384. https://www.bmj.com/content/350/bmj.h384.long

Ioannidis JPA, Trepanowski JF. Disclosures in Nutrition Research: Why It Is Different. JAMA. 2018 Feb 13;319(6):547-548. https://jamanetwork.com/journals/jama/article-abstract/2666008

Ioannidis JPA. The Challenge of Reforming Nutritional Epidemiologic Research. JAMA. 2018 Sep 11;320(10):969-970. doi: 10.1001/jama.2018.11025. https://jamanetwork.com/journals/jama/article-abstract/2698337

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s