So far, we've explored three interlinked causes of insulin resistance: cellular energy excess, inflammation, and insulin resistance in the brain. In this post, I'll explore the effects on micronutrient status on insulin sensitivity.
There is a large body of literature on the effects of nutrient intake/status on insulin action, and it's not my field, so I don't intend this to be a comprehensive post. My intention is simply to demonstrate that it's important, and highlight a few major factors I'm aware of.
Magnesium status is a determinant of insulin sensitivity. Although the evidence is not totally consistent, several studies have shown that improving magnesium status can reduce insulin resistance in diabetics and non-diabetics (1, 2, 3). This may be partially related to the fact that diabetes depletes magnesium stores. A four week low-magnesium diet reduced insulin sensitivity by 25 percent in healthy volunteers (4), although I would like to see this replicated using a more rigorous study design.
Iron may be another important factor. Although intuitively most people have a "more is better" approach to micronutrients, the evidence suggests that for many nutrients, the healthiest intake is a happy medium-- not too low and not too high. This seems to be true for iron. A useful measure of body iron stores is the blood concentration of ferritin, an iron storage protein. In general, the higher a person's ferritin level, the lower his or her insulin sensitivity (5, 6). This association is confounded by the fact that ferritin increases in response to inflammation, so it needs to be viewed with caution. However, studies have shown that reducing iron stores can increase insulin sensitivity in people with high ferritin (7, 8). This is typically done by repeated phlebotomy, or blood drawing.
One study showed that people who donate blood have a higher insulin sensitivity than people who don't (9), although it's difficult to be certain that this difference was not due to other factors. Personally, I give blood about once a year, because I want to help people in need of a transfusion, but also because it may have some side benefits to my health (as long as I don't dip too low-- low iron is a problem too). I recently received my "one gallon" pin, indicating that I've given blood eight times in the Seattle area since 2002.
It's best to get your ferritin checked before using phlebotomy to reduce iron stores. There's no use in reducing ferritin if it's already low. It's possible that our bodies are adapted to a constant loss of iron through intestinal parasites, injury and menstruation, and our body's evolved tendency to hoard iron doesn't serve us well in an environment where iron-rich foods are plentiful, we have few intestinal parasites, we don't often lose blood due to injuries, we have access to contraceptives that stop menstruation, and most women live well into their post-menopausal years*.
In recent decades, nutrition science has begun to explore the fascinating world of non-essential nutrients. As it turns out, these can have a powerful impact on health, including insulin sensitivity. Polyphenols are one class of non-essential nutrients that appears to be important. Numerous studies have shown that polyphenols and polyphenol-rich foods can increase insulin sensitivity, or attenuate the development of insulin resistance, in animals and humans (10, 11, 12, 13, 14). Some of the foods that have been commonly studied in this regard are chocolate, green and black tea, coffee, berries, citrus fruits, red wine and a variety of spices. These may act by increasing the defenses of cells against oxidants and inflammation by a process called hormesis (15, 16, 17).
There are certainly other dietary factors involved, some of which are unknown to science, and some of which are simply unknown to me. The broader point is that eating a diet composed of natural, minimally refined foods will offer an appropriate range of nutrients for most people.
* Interestingly, there is an iron storage disorder called hemochromatosis in which the body absorbs much more iron than is normal or desirable, leading to a variety of health complications. A mutation responsible for hemochromatosis (HFE C282Y) originated ~2,500 years ago in Europe and has undergone positive selection, rapidly increasing in prevalence to ~9 percent of Caucasians (18, 19). This suggests that this mutation, or something linked to it, may have conferred a strong survival advantage to our (my) ancestors a few thousand years ago. Why might this be? While people who are homozygous (carry two copies) for C282Y are at risk of serious medical complications, people who are heterozygous (carry one copy) for C282Y have a lower risk of iron deficiency (20). This would not have been important in the context of a hunter-gatherer diet rich in available iron from animal foods (which may be why this mutation was not selected for in the pre-agricultural environment), but in the context of an agricultural diet focused around grains, an iron-hoarding phenotype would have been advantageous. The prevalence of cribra orbitalia and porotic hyperostosis, both skeletal indicators of iron deficiency, increased greatly in early agriculturalists compared to hunter-gatherers, suggesting a high prevalence of iron deficiency (Paleopathology at the Origins of Agriculture, edited by Cohen and Armelagos). This suggests a strong selective pressure to evolve strategies to improve iron status. Our genome reacted, producing an effective, if crude, adaptation to a grain-heavy diet. In our current iron-rich environment, this adaptation is a liability because it increases the risk of iron overload. Carriers of the C282Y mutation are probably better off limiting red meat and liver, or at least monitoring ferritin status and giving blood appropriately.