Not used in the calculation of the nutrient density scores were the following micronutrients:

• Sodium is an essential mineral we all need a little bit of. However, most people in industrialized nations consume too much sodium (in the form of sodium chloride, i.e., table salt), and consuming even more sodium would almost certainly not improve health, but instead increase the risk of cardiovascular diseases. Thus, the vast majority of people do not benefit from consuming more sodium.
• Phosphorus is a similar case. Phosphorus is fairly ubiquitous in unprocessed foods, and also often added to processed foods. Thus, a deficiency in phosphorus due to insufficient dietary intake is almost unheard of. Further, consuming phosphorus in excess is associated with potentially negative health consequences.
• Vitamin D is very much a critical vitamin, because deficiencies are common in many populations around the world. However, most vitamin D is made in our skin, or from supplements or fortified foods. Natural foods are poor sources of vitamin D, with very few exceptions. And for some of those foods that can contain vitamin D, specifics of the production process make a big difference. For example, UV-irradiated mushrooms can be a very rich source of vitamin D, while non-irradiated mushrooms may contain almost none. Because such differences would not be apparent to the consumer, and because so few natural foods contain vitamin D, such that most people meet their needs through endogenous synthesis or by taking supplements, we decided not to consider vitamin D in the calculation of the nutrient density scores. 
• Panthothenic acid is a B-vitamin (B5) for which deficiencies are extremely rare, and have only ever been observed in situations of severe and long-term malnutrition. Thus, the vast majority of people obtain sufficient pantothenic acid from their diet, even without paying attention to pantothenic acid-rich foods. Further, no health benefits have been described from additional pantothenic acid intake beyond the required amount.
• Fluoride similarly can be consumed with food; however, the content in any given food is mostly dependent on whether that food was produced with fluoridated drinking water, and is therefore highly variable. It is also important to consider that no benefit is expected from increased fluoride consumption beyond the recommended level, and negative health consequences are much more likely with excessive intake than benefits. Therefore, measures such as drinking water fluoridation and the use of fluoride in toothpaste and dental applications are better suited to ensure optimal dental health than trying to increase fluoride intake by consuming fluoride-rich foods.
• Iodine is a bit of a tricky case, and we did consider including iodine in the calculation of the nutrient density scores. Iodine deficiencies are common around the world, and with good justification, one could certainly call this a ‘critical’ micronutrient. After all, it has been estimated that ~30% of the world population is deficient in iodine. However, the content of iodine in foods depends heavily on the content of iodine in the soil (for plant foods) or the fodder (for animal foods). For processed foods, the iodine content is heavily dependent on whether the salt used was iodinated or not. Further, for adults who meet the required intakes, there is no benefit to increasing their iodine consumption. And for those who suffer from an iodine deficiency, targeted treatments by supplement or by substituting iodinized salt for regular salt are much better strategies than recommending regular consumption of the few foods that are good dietary sources of iodine.
• Manganese is an essential mineral that is fairly ubiquitous in the food supply. Outright manganese deficiency has not been observed in humans, except in those that were given diets specifically low in manganese. And while some chronic conditions (osteoporosis and type 2 diabetes, in some studies) are associated with lower manganese concentrations in blood or tissues, there is little evidence to suggest that a diet rich in manganese, or manganese supplements, would improve these conditions. Further, potential toxicity due to overexposure is a concern with manganese.
• Molybdenum is an essential trace element. Like manganese, dietary molybdenum deficiency has never been observed in humans, except for specific unusual circumstances. There is also no evidence suggesting that increased consumption of molybdenum offers any health benefits.
• Selenium is another essential trace element that we decided not to include in the calculation of the nutrient density scores. The reason is that on a mixed diet, without paying particular attention to the selenium content of foods, US-Americans, for example, already are getting almost twice as much selenium from their foods as they require. Even globally, selenium is rarely ever seen as a nutrient of concern. More important in our decision was that selenium can be toxic at high intakes, and therefore selenium-rich foods such as Brazil nuts should not be consumed in excess (no more than 2-3 Brazil nuts per day are recommended). 
• Chromium and biotin are two essential micronutrients that we would have liked to consider including in the calculation of our nutrient density scores, but unfortunately, the concentrations of both have not been measured in most foods, and incomplete data in the database we used for this calculation (the USDA Food Composition Database, April 2018) made it impossible to include either in our scoring system.

2. Conversion of Micronutrient Contents to a Meaningful Number

As mentioned above, foods contain micronutrients in a measurement such as ug or mg or IU (International Units). The problem is that for most people, those are not meaningful numbers. Even for healthcare professionals, it’s hard to grasp the information on all of the micronutrients in a serving of food and interpret it correctly. 

We, therefore, converted the micronutrient contents, in three steps to create one number that would be meaningful and could be easily interpreted:

• We first calculated how many ug, mg or IU of each critical micronutrient could be found in 2,000 kcal worth of each food. The rationale for this was that each of us can only eat a certain number of calories per day, unless we want to lose or gain weight. What exactly that number is differs for each of us, but for most adults, it’s somewhere around 2,000 kcal per day or higher. In fact, the nutrition facts label that you find on all packaged foods in the United States uses a 2,000 kcal per day diet as a reference point, and we feel this is sensible.  Now, let’s assume you were eating 2,000 kcal per day, you would need to get all of your essential micronutrients in 2,000 kcal worth of food, right? By expressing the micronutrient content relative to 2,000 kcal worth of a food, we can immediately see how meaningfully that food would contribute to providing us with essential micronutrients, relative to the total number of calories we are able to eat.
• We then converted the ug, mg, IU of each critical micronutrient into a percentage relative to the recommended dietary allowance (RDA), or – for some micronutrients – the adequate initiate (AI). The RDA is the daily dietary intake level for each micronutrient that is considered sufficient to meet the requirement for almost all adults. It is set by the United States Food and Nutrition Board of the Institute of Medicine. The AI can be interpreted similarly. Now we have a percentage relative to the required amount (RDA OR AI) for each of the critical micronutrients in 2,000 kcal worth of a food. If that percentage is at least 100% for a specific micronutrient, that would mean that even if we ate only that food, we would get enough of that micronutrient to meet our daily needs.
• Lastly, we took the average of these percentages across all of the critical micronutrients.

Thus, the nutrient density scores are the averages across all chosen critical micronutrients, expressed in percent of RDA (OR AI) per 2,000 kcal of each food. Thus, a score of 100 means that if you only at that particular food, you would meet, on average, 100% of the required amount of all of the critical micronutrients. 

Other Considerations

Let me say that the Nourished by Science nutrient density scores are not a scientifically established measure that has long been studied, and it is certainly not the only way a nutrient density score could be calculated. This is simply an attempt to illustrate the vastly different nutrient densities of the different foods we eat, and I believe this is more helpful to people who want to eat a nutrient-dense diet than complex tables containing all of the micronutrients in each food.

Now, what if someone came along and calculated this differently, for example by including different critical micronutrients (or all of them), or by relating the contents of essential micronutrients not in percent of RDA relative to 2,000 kcal, but relative to, say, a serving of that food. Would these scores look entirely different? Almost certainly not. The absolute numbers would be different, but it is extremely likely that there would be a strong correlation between these scores and our Nourished by Science nutrient density scores.

One additional reason why we feel so strongly that nutrient density should be one major factor in helping us figure out what to eat is that nutrient-dense foods tend to have other benefits, beyond providing essential micronutrients. Most nutrient-dense foods are also high in fiber and so-called phytochemicals, molecules within plants that are not essential, but may provide health benefits. Diets rich in nutrient-dense foods will also usually be characterized by a low energy density (relatively few calories per g of food eaten), a low glycemic index (meaning your blood sugar levels do not spike as much when you eat them), a low content of ultra-processed foods, and a low content of food additives. All of these characteristics may contribute to the health benefits of a diet rich in nutrient-dense foods.

A final note to add is regarding food fortification. As you may or may not be aware, several critical micronutrients including vitamin A, vitamin D, or vitamin B1 are added to many foods in the United States and other countries, in order to reduce the risk of nutrient deficiencies in the population. We decided not to consider food fortification, i.e., we used the non-fortified versions of all foods for scoring. While fortification of foods is an important public health measure, we do not believe that the artificially elevated nutrient density of fortified foods should be considered as a criterion to guide personal dietary decisions. The reason is that many fortified foods are those that otherwise would be poor nutritionally, such as refined grain products, and not just with regards to their micronutrient content, but also with regard to many of the other factors that often come with a high nutrient density, such as a high fiber or phytochemical content.

So, what should we do with the nutrient density scores? I personally feel that eating a nutrient-dense diet is one of the key factors for long-term health, and therefore make sure that most of my meals consist mostly of foods with nutrient density scores of at least 100. Those foods with nutrient density scores of less than 100 and particularly less than 50, I see as treats that I do enjoy from time to time, but only in small amounts. Using these two simple rules basically ensures an adequate intake of critical micronutrients while still allowing me the occasional indulgence.

Again, you can get an electronic copy of a poster with the Nourished by Science nutrient density scores of very nutrient-dense foods and foods with low nutrient densities (‘empty-calorie foods’) by signing up to our newsletter here.

It is critical to consider the following points:

• A nutrient density score of 100 means that 2,000 kcal worth of that food contains an average of 100% of the recommended amount for the critical micronutrients. This does not mean that it contains 100% of each. Thus, it is important to consume a variety of these nutrient-dense foods to ensure that micronutrient requirements are met for all vitamins, minerals, and trace elements. It also means that we may want to give preference to foods with nutrient density scores well above 100. For example, sunflower seeds (nutrient density score of 163) are a very good source of vitamin E and copper, but lack vitamins C and B12. Oranges (score of 317) are a rich source of vitamin C, but also lack vitamin B12 and also vitamin K. For vitamin B12, we need an animal food, such as dairy, meat, fish, or shellfish. For vitamin K, leafy green vegetables. Thus, the best approach would be to include a mixture of different kinds of nutrient-rich foods (i.e., vegetables vs. fruit vs. legumes vs. grains vs. nuts & seeds vs. meat/fish/dairy/eggs) in one’s diet every day.
• Keep in mind that micronutrient density is only one of many factors we may want to consider when choosing which foods to eat, and which not to eat. Some of the most nutrient-dense foods listed here can be low in, for example, essential fatty acids or essential amino acids (i.e., protein). Thus, the nutrient density scores are not to be used as the one and only basis for designing someone’s diet.

Disclaimers

This chart was prepared by Nourished by Science LLC. While we took great care in providing this information, we cannot accept any responsibility for the accuracy of this content. This is partly due to the fact that the micronutrient information in nutrition databases is incomplete for some foods, and subject to constant change as more data are generated and entered. We suggest using these scores as estimates, and to also consider that the nutrient content of foods is heavily dependent on growth conditions, the soil nutrient content, the specific breed of a plant or animal, and many other factors.

As always, this content does not constitute medical or nutritional advice, and we suggest speaking to a qualified healthcare professional prior to making changes to your diet.

Literature of interest

1. Higdon J and Drake VJ: An Evidence-Based Approach to Vitamins and Minerals. 2nd edition. Thieme 2012. ISBN-13: ‎978-3131324528.
2. Bailey RL, West RP, Black RE. The Epidemiology of Global Micronutrient Deficiencies. Ann. Nutr. Metab. 2015; 66 (Supplement 2): 22-33.
3. Fulgoni VL, Keast DR, Bailey RL, Dwyer J. Foods, Fortificants, and Supplements: Where Do Americans Get Their Nutrients? J. Nutr. 2011; 141: 1847-54.
4. Beal T, Massiot E, Arsenault JE, Smith MR, Hijmans RJ. Global Trends in Dietary Micronutrient Supplies and Estimated Prevalence of Inadequate Intakes. PLOS One 2017; 12(4): e0175554.