Two reasons why confusion reigns about diets and human health
European sailors dying of scurvy in the ocean during long overseas voyages was very common up until the mid 1800's. Despite a long history rediscovering and forgetting the cure, it wasn’t until 1747, that James Lind supplied citric fruits to sailors in one boat out of 4 in his expedition (arguably carrying out the first controlled experiment in nutrition research) that a cure for the malady was formally found in the English speaking world (Spanish and Portuguese Navy had known and implemented this for at least 100 years). It wasn’t until the 1930’s that the mechanism was understood, Albert Szent-Györgyi isolated and identified ascorbic acid (vitamin C) and described the metabolic mechanism that enables its use within cells, laying the foundation of modern nutrition. This research earned him the 1937 Nobel Prize in Physiology.
Nowadays, the popularity of nutritional epidemiology is uncontested, to the point that almost any nutrient or food has contradictory peer reviewed evidence on its health implications and is covered extensively in the popular media, especially if it’s the passionate topic of eating meat. Recently there have been major publications looking at the global scale impacts on personal health and the environment of adopting alternative diets. Comparative risk assessments like The Global Burden of Disease (GBD) study place dietary risk factors as the primary driver of global deaths and morbidity, overtaking smoking and communicable diseases as leading risk factors. According to GBD, high sodium intake is the main culprit of poor diet related health. On the other hand, research by Tilman and Clark suggests that adopting a pescatarian or Mediterranean diet can reduce the likelihood of overall mortality while also reducing environmental impacts. The latest EAT-Lancet report goes as far as suggesting a global dietary pattern that is good for both personal and planetary health. But doesn’t the jump from contradictory evidence on health outcomes for every food or nutrient to the prescription of a global healthy diet seem somewhat incoherent?
There are in my opinion two culprits of the confusion (others will find more) — firstly, the quality of evidence in nutritional science, and secondly, the interpretation of that evidence as applied to individuals versus populations. Let’s start with the source of the evidence. Causality in nutrition science is hard to establish because the gold standard methodology –clinical randomized controlled trials (RCTs)– are almost impossible, and arguably unethical, to carry out long term studies on dietary patterns, or even particular foods (imagine getting approval for the sailor’s scurvy experiment today). The RCTs available usually look at intermediate markers (e.g. high blood pressure), which in turn are associated with later onset of disease (e.g. cardiovascular diseases), but these markers are not the disease itself. The second ‘best’ available long-term evidence to date –and likely to stay so into the future– relies on observational (prospective cohort) studies where problems abound: Infrequent and unreliable food consumption reporting, heterogeneity of the confounding factors over long study periods, high drop-out rates, biased study populations, etc. Still, if the signal were strong enough for a given nutrient or food type (think again of vitamin C), the significance and interpretation of the effects shouldn’t flip-flop as it does with many single food interventions.
So, simply put, the quality of evidence is ‘not ideal’, but it’s the only evidence we have. The path out of this conundrum remains unresolved. Is there anything with reliable evidence? The answer is somewhat. Over and under consumption in reference to your activity levels is bad; trans-fats are bad; and so are diets high in ultra processed foods containing high amounts of the main suspects of poor health such as trans-fats, added sugars, and sodium. On the positive side, higher dietary diversity is good for health, but exact amounts and proportions of individual items are very hard to pin down.
The interpretation of the evidence on the other hand, even if it were reliable and non-contradictory, is perhaps more paradoxical. Most single food item changes are estimated to have small relative risk reductions (see questions about relative vs absolute risk). At the individual level this means that changing your diet by that one item will have a smaller absolute risk reduction for incidence of disease or death, relative to other factors influencing your personal health. If the odds of seeing a health benefit are only 1 in a 1000, making a shift at the individual level may not make any sense for that individual. On the other hand, for a public health official, thinking about a population of, say, 10 million people, the same odds translate to 10,000 fewer people that need treatment for their disease. That would be a major benefit for public health and reduced health-care costs. Should you find yourselves in the official’s position, the decision seems clear-cut, and that’s why most national dietary recommendations follow the findings of nutritional epidemiology. However, dietary recommendations rely on the individual to make the ‘informed’ choice to realize that population level benefit. On the other hand, if policies go beyond just recommendations or guidelines they have a better chance to attain population level health benefits. However, they rarely do, so there is virtually no personal incentive to opt for the 1 in 1000 chance of being better off. Or is there? Some argue there is, through the improved collective well-being, decreased taxes and insurance premiums associated with the healthcare costs of an unhealthier population, as well as reduction in shared environmental costs of consuming foods bad for the planet.
What does all this mean for the impact of diets on the environment and the win-win scenarios we so desperately look for in the environmental science literature? It’s complicated. There is little to no disagreement on the fact that ruminant animal meat like beef and lamb are the most green house gas (GHG) and land-use intensive foods, but the ‘eat less meat’ argument for both your health and the planet seems to have way stronger support for the planet aspect than the health one (at least directly, since a worse off planet is bad for you too). Substituting read meat consumption for other protein rich foods (pork, poultry, or legumes) in regions currently above the world average, would reduce global agricultural GHG emissions by at least 15% (more scenarios and comparisons here). That means those of us reducing or eliminating red meat consumption for environmental reasons have great grounding and gain smaller changes in personal health outcomes, but those of us concerned primarily about our health, won’t be looking at red meats for the biggest win (unless we are eating a plate full of meat every day, but a plate full of just one thing every day is bad right?). Consequently, health driven diet shifts (other than reductions in overconsumption) won’t represent the largest wedge on diet related climate change mitigation. To stress this point further, health concerns always rank higher than environment in surveys of food choices in the general population and both tend to rank lower than price, taste, culture, and convenience.
In conclusion, nutrition research does matter, historically it has told us everything we know about the essential components needed in our diet, including curing diseases such as scurvy, but low hanging clear cut findings are now rare. New research might find very small marginal gains for already balanced diets. In general, the confusion generated by ‘bad quality’ evidence and the population vs. individual paradox, sound like bad news, and they are. Same for the environmental implications. It amounts to yet another call to steer away from studies looking at individual foods and for action to design better nutrition research focusing on whole diets. Equally important are novel policy ideas to reduce the intake of environmentally costly foods and diets. While the personal win-win of reduced red meat for health and environment may not be equally strong on both sides, it is still a really good idea to reduce red meat consumption for the sake of our planet and our society. Any reduction, regardless of size will bring us closer to environmental targets.
