WORLDVIEW: Scientists don’t know what a healthy diet looks like

By Felicity Duncan .I should start this off by saying that I am pro-science. The rigorous process of gathering observable evidence and using that evidence to try to nullify cherished hypotheses is not perfect, but it's the best system we have for advancing human health, wealth, and safety..Science works – you can tell by the way that humans have almost doubled their pre-scientific era lifespans and today live lives of unimaginable comfort (if you shower every day, have watched more than 50% of your children survive to adolescence, and travel by anything other than foot or horse, you are already enjoying a more comfortable life than, say, most of the historical reigning monarchs of England)..___STEADY_PAYWALL___.However, as wonderful as our science is, it has its limitations. And understanding the relationship between diet and health is one of those limitations..Consider salt. Governments around the world recommend that people, especially those with high blood pressure, should limit the amount of salt that they eat. Our own SA Heart and Stroke Foundation says "A high salt intake is linked to high blood pressure. Reduce your salt intake to no more than 5g (1 teaspoon) of salt, from all sources, a day.".Read also: Diet – the collusion/collision of science and big-business.The global consensus on salt was originally based on a mix of animal studies (feed rats enough salt and they'll get high blood pressure), observational evidence (a French doctor noticed that some of his patients with high blood pressure loved salty foods), and population studies (the Japanese eat a lot of salt and have more strokes than some other populations)..This is the typical kind of evidence used for other health recommendations, like the low-fat diet (for more on this, see Gary Taubes' fascinating book, The Diet Delusion). But it's not without its problems..For example, in the animal studies in question, rats were fed the human equivalent of 500g of salt a day. No actual humans are walking around eating half a kilo of salt a day, so this isn't helpful information. Rats are also, importantly, not humans. Humans and rats do not necessarily react exactly the same to medical interventions, as dementia research has repeatedly shown..Population studies also have their problems. For example, the landmark study that got the salt ball rolling was by Lewis Dahl. He compared Alaskan Inuit, who consumed about 4 grams of salt a day, to Japanese people who consumed between 14 and 26 grams a day (and to various other populations that fell in between)..He concluded that salt consumption accounted for the higher incidence of high blood pressure and stroke among the Japanese. However, the Japanese people in the sample were older than the Inuit, and some of the Japanese samples studied were exclusively male, while there was a mix of both genders in the Inuit sample, and he appears not to have controlled for these differences..He ignored the role of genetics. The Japanese and Inuit were at the time (and still are) very homogenous, genetically distinct populations, but he didn't compare within populations to account for genetic issues – did the Japanese men who ate the most salt have more hypertension than those who ate the least? He also ignored other factors – what about the incidence of smoking, for example, which is also associated with high blood pressure? These are typical limitations of population studies, especially the early ones that most of today's diet recommendations are based on..Read also: Live longer with less salt – top researchers re-affirm top advice.As for the observational evidence, that's roughly the equivalent of that one guy who says he thinks it's OK to smoke because his friend's wife's grandma smoked a pack a day and she lived to be 90. The point of scientific inquiry is to find replicable results, not to base beliefs on outlier cases..Despite the limitations of the evidence, it was considered good enough, in 1977, to carve in stone the idea that salt raises blood pressure and, conversely, that a low salt diet could lower blood pressure when it was too high..Since those days, however, numerous studies have shown a very weak relationship between salt and hypertension or no relationship at all. In fact, some studies show that eating more salt means lower blood pressure..What's more, even the most rigorous salt reduction diets tended to reduce blood pressure by just 1 mmHg (the difference between 120/80 and 119/79), and there's no evidence this is enough of a reduction to make a difference to heart disease..There also appears to be no consistent relationship between salt and mortality – again, some studies show that the more salt people eat the less likely they are to die from heart disease. And, of course, there's the fact that the Japanese, with a notoriously salty, soy sauce rich diet, have the world's longest life expectancy..In other words, the evidence on salt is mixed at best. Some people do well on a low salt diet, some people do well on a high salt diet, and we don't know which people are which..And as with salt, so with virtually all dietary advice. Advice to lower fat consumption, for example, is based on similar evidence to the salt recommendation, as is advice related to cholesterol..The reality is that population studies or epidemiology are never going to be the right tools for diet recommendations because there are simply far too many confounding factors, ranging from genetic differences to measurement issues. It's impossible to isolate one variable in a population study without introducing other biases..What's more, you can't do a double-blind, placebo-controlled dietary study, because it's not like people won't know if they're on the high-carb diet or not. Studying diet is just hard. Yet many government organisations have made major health recommendations based on pretty speculative research. It might be better to just admit we don't know yet and leave it at that..The best diet studies try to come close to a proper experimental design. Studies that do so tend to find something fairly prosaic – different diets work for different people. Human genetic variation is substantial, both across groups and within groups (in fact, on many measures, there is more difference within a group of women than there is between a group of men and a group of women – we really are all unique). Some people thrive on a low-fat diet, but some thrive on a high-fat diet. It's possible that there is no one healthy diet, but that we all must figure out what works for us..Unfortunately, the effect of diet takes a lifetime to manifest and we all only have one chance to get it right. Even more unfortunately, science can offer us only limited guidance on this question. Those who claim otherwise are selling something.

WORLDVIEW: Scientists don’t know what a healthy diet looks like

Related Stories