Laroxe said:
Virtually all the evidence around diet and health is drawn from the huge body of Observational studies known as epidemiology. Unfortunately by their nature observational studies are considered to be unreliable, they are wide open to biases, confounding variables, data manipulation and financial conflicts of interest. These studies tend to rely on self-reported data, often from memory and over significant periods of time, making it practically impossible to verify. Really, observational studies are best used to provide hypothetical links that can then be verified by randomised controlled trials, but truly randomised control trials in nutrition is virtually impossible.
One of the most significant study using these methods was conducted by Bradford Hill and Richard Doll in 1950, which identified the association between smoking and lung cancer, an association so strong that it was almost immediately accepted as causal. However, they themselves recognised the very real problems involved in these sorts of studies and left a set of guidelines which can help in evaluating such research. Normally called the Bradford Hill criteria, they remain very useful in making judgements about the validity of the findings.
Perhaps the most important application of such research has been in the widespread adoption of the Mediterranean diet. It was noticed that the inhabitants of a number of Mediterranean islands seemed to have significantly longer lifespans than those in other areas. A number of American researchers then set about creating a diet, in line with their own beliefs about healthy diets, claiming this as the reason for their longevity. It's suggested that this diet wasn't really are flection of what the islanders eat, and it's thought few of the islanders had an accurate measure of their ages. Despite this, and the fact that there are other countries with longer average lifespans, including Finland, Switzerland, Australia and Japan, all with very different diets, the Mediterranean diet continues to be promoted for a wide range of health problems. It was this and the follow-on studies that promoted the advantages of plant based foods, Olive oil, and nuts while advised against red meat and animal fats. It continues to be widely promoted despite the continues inconsistency in findings, weak statistical correlations, some financial irregularities and poor theoretical rationals in explaining its effects.
How fragile are Mediterranean diet interventions? A research-on-research study of randomised controlled trials. https://nutrition.bmj.com/content/4/1/115
Unfortunately the same issues plague the findings of the studies that attempt to compare different diets, if there is an effect its a small one. Its interesting that the Vegan diet was the product of a prophetess in the seventh day Adventist, as a way to reduce the sex drive.
A2011 analysis of 52 claims made by nutritional epidemiology tested in 12 well controlled trials found that not one of the 52 claims—0%–could be confirmed.
https://rss.onlinelibrary.wiley.com/doi/epdf/10.1111/j.1740-9713.2011.00506.x
Working through a lot of the recommendations using the Bradford Hill criteria is interesting and simply shows that the state of research in diet and health is a mess really. One novel finding is even the studies that claim benefits in some health conditions, this rarely seems to affect lifespan.
Most of the work on longevity seems to be working on the principle that we may already have reached an upper limit and that any improvement will need to focus on the process of ageing specifically.
https://hsph.harvard.edu/news/life-expectancy-may-be-reaching-upper-limits-for-now/
https://www.nature.com/articles/s43587-024-00702-3#Sec6
One of the problems with the main study validating the Mediterranean diet is that it involved Mediterranean people eating a Mediterranean diet that their ancestors had been eating for centuries, if not for thousands of years.
It is not only likely, but probable, that the Mediterranean people have experienced selective genetic adaptations in that time frame to make them particularly suited to the Mediterranean diet (or conversely that the diet has been adapted over time to their very specific genetic profile).
It doesn't necessarily follow that people from Finland or Japan or Nigeria whose ancestors ate very different foods for the past many centuries are best suited to the same diet. Maybe Finns do better with the traditional Finnish diet, Japanese people do better with the traditional Japanese diet, and people from a particular part of Nigeria do better with the traditional diet of their ancestors. It could be that the Mediterranean diet is better for everyone, but the evidence being relied upon to show that doesn't establish that fact.
We certain know of cases where one place's traditional diet is ill suited to people from somewhere else.
For example, lots of people globally, including a particularly large share of East Asians, are lactose intolerant. But, the traditional diets of Northern Europe have lots of milk.
The Mediterranean diet has lots of olive oil, something that was not part of the ancestral diet in Northern Europe, the Pre-Columbian Americas, Australasia, Oceania, much of Asia, and much of Africa. I wouldn't be at all surprised if olive oil did not have the beneficial effects it does for Mediterranean people in some populations from some of the regions where olive oil was not part of the ancestral diet.
As another example, food in tropical and subtropical areas tend to be more spicy, while food in temperate and colder areas tend to be more bland. And, it would hardly be surprising if spicy foods had different health consequences for people with ancestors from tropical and subtropical areas compared to the health consequences for people with ancestors from the "bland zone". The leading hypothesis for this ecological zone variation is that spicy foods are anti-microbial. So, they might interact in a genes x environment manner with genetic HLA immune system responses which are some of the most active sites of genetic adaptation during the Holocene era in modern humans and which are often highly population specific, based upon the pathogens that different populations' ancestors encountered.
Yes, it's all speculation, but it is plausible speculation and it ought to be rigorously tested in many populations before it is recommended for everyone, the way that the Mediterranean diet is.
Also, actual genes from people's own DNA aren't the only possible way that regional affinities and dis-affinities for diets could arise. Regional or population specific variations in gut bacteria is another. So are immune system response to common non-deadly pathogens in a region that most people in that region encounter as children and develop an immunity to, which is why, for example, tourists visiting Mexico for the first time often get "Montezuma's revenge", while locals are unaffected.
Simple culturally learned preferences for certain foods also matter. Diets only work if you can keep following them consistently over time. This is easier if you are eating comfort foods that you've had and made for yourself routinely for your entire life and is harder if you've never made these foods, they aren't familiar to you, you don't even know how they are supposed to taste when made properly, and you may not have access to precisely the same ingredients.
As an example of that, Southern biscuits which are easy to make and ubiquitous in the American South, are notoriously difficult to make elsewhere, because the flour used to make them in the South comes from a different subtype of wheat grain than the baking flour available in grocery stores almost everywhere else, even though the labels on the flour in the South and elsewhere respectively, don't indicate anywhere that they are made from different kinds of wheat. The labels just say that one of the ingredients is "wheat" without further specification. Now, this isn't to say that Southern biscuits, in particular, are more healthy. Indeed, the opposite might be true. But the lesson that subtle differences in the kinds of ingredients that are only available locally in one region (which are invisible to all but the most sophisticated consumers) might impact how well recipes from that region work when you are cooking them, has broad applicability.
