A healthy vegetarian diet
* This note was inspired by a query from someone with type 1 diabetes who works with other people with diabetes to help them to control their blood glucose levels. It looks at how easy (or difficult) it is for a vegetarian diet to meet recommended daily allowances (RDAs) for vitamins and minerals. It also examines how much carbohydrate recommended vegetarian diets comprise.
* The 2020-2025 Dietary Guidelines for Americans (DGAs) presented two dietary options for vegetarians – one for toddlers and one for those over two years old.
* I examined the US DGA recommended vegetarian diet for a female aged 31-50 (as an example) and found it to be deficient in several nutrients. The DGA report did say that nutrient supplementation would be required despite one of the overarching principles of the guidelines being “Focus on meeting food group needs with nutrient-dense foods.“
* The US DGA recommended vegetarian diet was also extremely high in carbohydrate (60% of total calorie intake).
* There is a way in which a vegetarian diet can meet most nutritional requirements and be below 100g carbohydrate each day. However, it involves consuming a diet vastly different to the government recommended one of whole grains, legumes, fruits and starchy vegetables.
Many thanks to Jon Furniss, PhD, for the suggestion for this week’s note. Jon is an ambassador for the Public Health Collaboration charity and thus works to help other people to “find a way out of metabolic disease.” While not vegetarian himself, he does try to help some people who are and wondered “what could be a healthy vegetarian protocol for the metabolically unhealthy?” Jon has observed other conditions in people who don’t eat meat or fish (e.g., depression, anxiety, and sleep disorders) and wondered if these were a) related and b) could be avoided.
Jon was diagnosed with type 1 diabetes in adulthood. Jon wears a Continuous Glucose Monitor and has found that, the less carbohydrate he consumes, the more stable his blood glucose level is. He thus consumes extremely little carbohydrate – somewhere between keto and carnivore levels.
The challenge with vegetarian diets in the low carb world is that vegetarians avoid the two highly nutritious carb-free foods: meat and fish. (Liver can have a carb content – the glycogen stored by the animal – but meat and fish are essentially carb free). Pure fats (oils such as olive oil, sunflower oil, coconut oil etc) also have no carbohydrate (they have no protein either) but they also have negligible nutrition. Eggs have a trace of carbohydrate. Dairy products have a higher level of carbohydrate – little more than a trace for hard cheese, but fluid dairy (e.g., milk) can be 5-10% carbohydrate. Low carbohydrate intake – which would be helpful for the management of metabolic disease – is more difficult as a vegetarian.
To explore Jon’s topic, let’s first look at a ‘healthy’ vegetarian diet to see how it fares nutritionally for micronutrients (vitamins and minerals) and macronutrient intakes (carb, fat and protein). Then let’s look at how a vegetarian diet could be adapted to be lower in carbohydrate.
A government recommended vegetarian diet
The 2020-2025 Dietary Guidelines for Americans (DGAs) were the most recent government guidelines issued (Ref 1). We did a post on this in January 2021 (Ref 2). Five dietary patterns were presented in the DGAs – two for toddlers and three for those over two years old. Two of the five diets were vegetarian (one for toddlers and one for those over two years old). Despite one of the overarching principles of the guidelines being “Focus on meeting food group needs with nutrient-dense foods“, the DGAs advised that supplements would need to be taken with the vegetarian diets.
We should caveat up front that we don’t know if micronutrient recommendations are adequate. The conclusion of a recent post on micronutrients was “While the RDAs might be wrong, they’re a useful guide and we should see them as minimums for optimal health” (Ref 3).
Table A1-2 is on page 133 of the 2020-2025 DGAs. This table reports the acceptable macronutrient distribution ranges (AMDRs) and the recommended dietary allowances (RDAs) for males and females aged 2 and older. I picked a female aged 31 to 50 as our example person to analyse. I built a spreadsheet with these AMDR and RDA amounts entered as targets. Our female has a calorie requirement of 1,800 calories a day (a ‘junk’ allowance of 150 calories is on top of this – see below).
Table A3-4 is on page 148 of the 2020-2025 DGAs. This table presents a “healthy vegetarian dietary pattern for ages 2 and older.” One column in that table represented a diet that delivered 1,800 calories. That was the one to be consumed by our female aged 31-50.
This diet recommended:
– Vegetables: 2.5 cup equivalents of vegetables per day. There were additional recommendations for the type of vegetables to be consumed per week. This enabled us to calculate that approximately 9% of vegetable intake should be greens, 31% should be red/orange vegetables, 9% should be beans/peas/lentils (I know they’re pulses, not vegetables, but just go with it), 29% should be starchy vegetables and 23% should be other vegetables.
– Fruits: 1.5 cup equivalents per day.
– Grains: 6.5oz per day – 3.5oz whole grains and 3 oz refined grains.
– Dairy: 3 cup equivalents per day.
– “Protein foods”: 3oz per day (yes – I know – protein is in everything other than oils and sucrose – keep going with it).
– Oils: 24 grams per day (the only one in metric).
– 150 other calories: On p20, this was explained as “Calories up to the specified limit can be used for added sugars, saturated fat, and/or alcohol, or to eat more than the recommended amount of food in a food group.” This means that the US government allows 150 extra calories without nutrients essentially (saturated fat shouldn’t be in there of course – it’s a component part of virtually every other food).
I kept the foods used as simple as possible. Carrots were used for red/orange and starchy vegetables, broccoli was used for greens, lentils were used for beans/peas/lentils and cauliflower was used for other vegetables. Apples were the chosen fruit. Flour (whole and refined) was used for grains. Milk was used for dairy. Eggs were used for protein foods and olive oil was used for oils.
I put these into a spreadsheet and used the United States Department of Agriculture (USDA) all-foods database to capture micronutrients and macronutrients per 100g of each of these products (each food converts differently from cups to grams.) I then used the actual grams of each product (e.g., a fifth of a cup of broccoli is 18g) to work out how much of each nutrient was actually consumed and then added these all together.
Reassuringly the total for our female came to 1,730 calories and 1,784 by two different methods (which was well within known margins of error for the USDA data) (Ref 4). (The table presenting the foods in the DGA vegetarian diet and their nutritional content is at the end of this note).
The diet did not meet the RDAs for retinol (Ref 5), B3 (niacin), B6, vitamin C (surprisingly), vitamin D, vitamin E and Vitamin K (and that’s before getting into D3 vs D2 or K2 vs K1 and what the body actually needs). The diet fared better for minerals (these come from the ground, so you would expect that in a plant-based diet) but it was still woefully deficient in iron – delivering 11.5mg out of a target of 18mg. And much of the iron was reported as coming from plant foods and is thus not as bioavailable as the iron provided by the eggs. Table A1-2 did not include essential fatty acids and there was only one mention of EPA/DHA in the report (this is the form in which we need omega-3 essential fatty acids). It is unlikely that the diet met EPA/DHA requirements and would likely be high in omega-6 to omega-3 as a ratio of essential fatty acids.
The macronutrient composition was 267g carbohydrate, 77g protein and 45g fat. The targets set in the AMDR table were 130g for carbohydrate, 46g for protein and no grams were given for fat, but a percentage range of 20-35% was given instead. The fat percentage turned out to be 23%, so the actual intake from our example diet was at the low end of the AMDR. The carbohydrate intake was a whopping 60% and protein made up the 17% remainder.
There are limitations of this exercise. I have chosen common foods for each food type e.g., carrots for red/orange vegetables. A different selection of vegetables, fruit or grains could be more nutritious. I suspect it would be a bit swings and roundabouts – a food better for one nutrient is often worse for another. I chose low-fat milk, as that was the DGA instruction. Choosing normal milk (whole milk is only 3.5% fat – it’s hardly a high fat food) would have been more nutritious, but it would have taken the diet out of the calorie range and it would have gone against the guidelines. I chose eggs for the “protein foods.” In a vegetarian diet I could have chosen nuts, more pulses, soy products and/or nuts and seeds. Eggs were the lowest carbohydrate choice of these and I am confident that they are the most nutritious option. Almost half the choline intake, for example, was delivered by the 3oz/97g of eggs.
A ‘box out’ on page 116 of the 2020-2025 DGAs entitled “Vegetarian or Vegan Dietary Patterns During Pregnancy and Lactation” cautioned that: “Women following a vegetarian or vegan dietary pattern during these life stages may need to take special care to ensure nutrient adequacy.” The nutrients of concern were iron (because plant source foods only contain non-heme iron, which is less bioavailable than heme iron), B12 (because it is present only in animal source foods), choline, zinc, iodine, and EPA/DHA. The box out advised that such women “should consult with a healthcare provider to determine whether supplementation is necessary and if so, the appropriate levels to meet their unique needs.” As we have seen, this doesn’t just apply to pregnant or lactating women – any vegetarian can be suffering from nutritional deficiencies.
Can a vegetarian diet meet requirements?
In my 2010 book, The Obesity Epidemic: What caused it? How can we stop it? I noted that US and UK citizens were consuming, on average, 1,100 empty calories a day from sugar and flour. I undertook an exercise to see if I could replace those 1,100 empty calories with optimally nutritious calories to deliver all our vitamin and mineral RDAs (I didn’t look at essential fats). I discovered that this could be achieved in an omnivore diet by eating 1,077 calories comprising the following: 35g of porridge oats; 125g of whole milk; 75g of liver; 50g of broccoli; 200g of spinach; 25g of cocoa powder; 125g of sardines (oil based, bones included); 200g of eggs and 20g of sunflower seeds (these are invaluable for vitamin E).
I built on this exercise for this week’s note and discovered that a vegetarian could overcome the vitamin and mineral deficiencies in the US DGA diet by consuming five foods: 250g white raw mushrooms (grown in sunshine); 30g sunflower seeds; 450g eggs (just over 10 medium eggs); 1kg of whole milk and 50g spinach. This would provide approximately 1,600 calories leaving the person to consume more fat or carbohydrate as suited them and their metabolic needs.
The macronutrient composition of those five foods is 20% carbohydrate, 27% protein (high) and 53% fat. The five foods in those quantities provide 81g of carbohydrate, which would be way better for metabolic management than any government recommended diet. The majority of the carbohydrate grams (53g) came from the milk. Cheese would be lower in carbohydrate, but milk provides 1mcg of vitamin D for 60 calories and cheddar cheese provides 0.3mcg for 400 calories. Other parameters would fail if this swap were done.
In answer to Jon’s query about other health conditions – depression, anxiety, sleep disorders – without supplementation these would be likely on a nutrient deficient diet. I always think that the B vitamins should be called the M vitamins, as they are responsible for Mind, Mood, MoJo, Muscles, Memory and so on. The B6 deficiency alone, which is likely in vegetarians, could be causing issues.
The low carb vegetarian
This exercise helpfully illustrates just how high in carbohydrate a government recommended vegetarian diet is. Someone trying to optimise their management of diabetes/metabolic disease with a low or very low carbohydrate intake would view moderate carbohydrate intake (130g/day) (Ref 6) as an absolute maximum. If such people wore a continuous glucose monitor for even a couple of weeks, they would realise, as Jon did, that the lower the carb intake, the more stable the blood glucose. They might well realise that carb intake of 50g/day or even lower would be a sensible target.
For the vegetarian low carb eater, 50g/day of carbohydrate is a challenge. Grains, beans, peas, lentils etc. need to be avoided. Fruit needs to be berries, not apples or higher sugar fruits. Vegetables should be green, rather than starchy. As much of the diet as possible needs to be consumed in the form of eggs, hard cheese and lower carb dairy and even then omega-3 supplements, among others, should be considered.
There are many apps available today that will calculate the nutrient composition of one’s typical food intake. We all tend to eat similar things in a week – often similar things each day. We would all benefit from examining our own diets to see what we may be deficient in and considering dietary changes, or supplementation, to ensure that we are at least obtaining recommended dietary minimums. Vegetarians will likely have greater deficiency challenges than omnivores, but these can be overcome.
For optimal health, we may wish to consider higher nutrient intakes. The RDA stands for Recommended Dietary Allowance, after all. As Sally Fallon Morell once said, “why am I only allowed a certain amount of nutrition?”!
Until the next time
All the best – Zoë
The Table. You can view it full size here
Ref 1: U.S. Department of Agriculture and U.S. Department of Health and Human Services. Dietary Guidelines for Americans, 2020-2025. 9th Edition. December 2020. Available at https://www.dietaryguidelines.gov/
Ref 2: https://www.zoeharcombe.com/2021/01/dietary-guidelines-for-americans-2020-2025/
Ref 3: https://www.zoeharcombe.com/2020/08/the-carnivore-diet-micronutrients/
Ref 4: The first method simply took the calories per 100g of each food and then calculated what proportion of 100g (or multiples of 100g) were consumed and then totalled up the calories in this way. That came to 1,730 calories. The other method took the macronutrients consumed in the total diet and allocated 4 calories per gram of protein, 4 calories per gram of carbohydrate and 9 calories per gram of fat (this is known to have rounding errors). That came to 1,784 calories.
Ref 5: I did not calculate carotene equivalent because not everyone can convert carotene to retinol, conversion is inefficient, and conversion cannot be relied upon – as I found out after two (avoidable) eye operations!
Ref 6: Noakes TD, Windt J. Evidence that supports the prescription of low-carbohydrate high-fat diets: a narrative review. Br J Sports Med 2017.