Problems with nutrition research…
“Healthy” Subjects
Most nutrition research is conducted on so-called "healthy individuals"—but in modern Western societies, “healthy” often just means the absence of diagnosed disease. It rarely reflects optimal metabolic function. Many people considered “normal” still have signs of low-grade inflammation, blood sugar dysregulation, hormonal imbalances, and subclinical deficiencies. What's typical in industrialized nations is not necessarily a healthy baseline.
To truly understand how food affects human biology, we should be studying metabolically resilient populations—like those in non-industrialized societies, where diets are based on whole foods, lifestyles involve natural movement and exposure to light, and metabolic dysfunction is far less common.
Short-Term Benefits Can Be Misleading
On of the biggest problem with nutrition research is that it often focuses on short-term effects without considering the impact on metabolism in the long run.
Adding almost any nutrient-dense food to a deficient diet will have some positive short-term effect. If someone is eating a highly processed, nutrient-poor diet and then starts eating more spinach, nuts, or fortified cereals, they may see improvements in certain markers—at least temporarily. But does that mean those foods are the best long-term choices over all? Not necessarily. Some might come with drawbacks, like high anti-nutrients or an imbalanced fatty acid profile, changing our cell membranes for the worse over time.
Context Matters, But Studies Often Ignore It
Most nutrition studies don't consider what else people are eating. If a study finds that people who eat more whole grains have lower inflammation, is it because grains are inherently anti-inflammatory? Or is it because those people are replacing ultra-processed junk food with something slightly better? Similarly, if a study links red meat to higher disease risk, is it because of the meat itself—or because it’s cooked in seed oil and combined with refined carbs?
Correlation vs. Causation
Epidemiological studies (which make up a huge portion of nutrition research) rely on food frequency questionnaires, where people recall what they ate—often inaccurately. Then researchers find correlations between certain foods and health outcomes, but correlation is not causation. A classic example is how studies used to link dietary cholesterol to heart disease, but later research showed that cholesterol in food has little impact on blood cholesterol for most people.
More Nutrients ≠ Optimal Health
Just because something contains vitamins and minerals doesn't mean it's the best source. For example:
Fortified plant milks contain calcium, but not in the most bioavailable form.
Nuts contain magnesium, but also high levels of omega-6, which can be inflammatory.
Spinach has iron, but it's non-heme iron, which is poorly absorbed compared to the iron in animal foods.
The presence of a nutrient doesn't tell you how well your body can use it or what else comes with it that might be harmful.
Long-Term Effects Are Hard to Study
Nutrition science struggles with long-term studies because:
People's diets change over time.
Peoples over all metabolism is not measures or controlled.
It's hard to isolate one food or nutrient over decades.
Many factors (sleep, stress, exercise, environment) affect health, making it difficult to pinpoint food's exact role.
This is why many diet trends look promising at first but fail long-term. Vegan diets, for example, often improve health markers short-term because they may cut out processed foods, but long-term, they are more prone to lead to deficiencies in key nutrients like B12, zinc and selenium.
Just because a food or nutrient shows benefits in the short term doesn't mean it's optimal for long term metabolism.
The best approach is to focus on nutrient-dense, bioavailable foods that humans have thrived on historically — rather than relying on trendy superfoods or isolated nutrients.
