* A study has been published, which reported the results of a survey of over 2,000 people who followed a carnivore diet for at least 6 months.
* The study makes no claims – it merely reported the characteristics, eating patterns, health perceptions/measurements and level of satisfaction of those following a diet based on animal foods.
* People were primarily motivated to follow a carnivore diet for health/body weight reasons.
* Most people consumed red meat at least daily, if not at each meal. The high majority of people reported never consuming legumes, vegetables (even non-starchy ones), fruit or grains. They also rarely consumed alcohol, which could have been a confounder in the health results.
* Changes in health were self-reported; virtually every condition surveyed was resolved or improved. As has been found in previous studies of very low-carb diets, lipid profiles worsened in some people (although more people reported resolution or improvement in lipid profiles than worsening).
* Participants in the survey had very high levels of satisfaction with the diet and the health improvements were undeniably impressive, even if self-reported.
* The significant caveat is that only people who enjoyed and/or felt benefit from undertaking such a diet would have continued to follow it for at least six months. This research tells us, therefore, if the diet suits you, this is what you might expect to experience.
A friend of mine called Emma sent me the paper for this week’s topic. The paper was called “Behavioral characteristics and self-reported health status among 2029 adults consuming a “Carnivore Diet” (Ref 1). The senior author was Professor David Ludwig. We have covered his work a few times – not least the recent review of the Carbohydrate Insulin Model of obesity. This paper was, as the title suggests, a review of the type of person most likely to be consuming a carnivore diet and details of their health status before and after following the diet. The paper didn’t make claims (e.g., ‘the carnivore diet reduces the risk of heart disease’). Rather, it reported the results from a broad survey of people following a carnivore diet. This week’s note doesn’t need an examination of claims, therefore. Instead, I have summarised the results and picked out what I thought were the most interesting findings, while noting a significant caveat and some confounders.
The significant caveat is that this was a review of people who chose to follow a carnivore diet and did so for at least six months. Anyone who started a carnivore diet and then stopped, for whatever reason, was not eligible for this study. This research will inevitably be more positive about the carnivore diet because of this. Only people who enjoy and/or feel benefit from undertaking such a diet will continue. The research is still valid and valuable, but we need to approach it with this caveat in mind. We will discover – of those people who started and then stuck with a carnivore diet – who were they, why did they do it and what did they gain from it?
The carnivore diet was defined as one based on animal foods, which excluded most or all plant foods. Part of the review examined how much plant produce was consumed, so we’ll cover that shortly.
The study was conducted via social media, which was novel and is likely to be an avenue for future research. The research was undertaken between 30 March and 24 June 2020 among adults self-identifying as consuming a carnivore diet for six months or longer. The survey questions were developed in consultation with members of the carnivore diet community. (Dr Shawn Baker was specifically thanked towards the end of the paper). The questions were designed to understand motivation, diet intake, symptoms suggestive of nutritional deficiencies or other adverse effects, satisfaction with this way of eating, prior and current health conditions, anthropometrics (the practice of taking measurements of the human body), and laboratory data. The full questionnaire can be seen in the supplemental file. The questions covered four main areas: 1) current intake and frequency of consumption of different foods; 2) chronic medical conditions and medication use, anthropometric and laboratory data, perceived health and well-being, and perceived symptoms of nutritional deficiencies or other adverse effects – including in the present and before starting the diet; 3) diet satisfaction and social support; and 4) characteristics of the participants.
The people (Table 1)
The paper was well structured, making it easy for the reader to understand the key factors of interest – who is most likely to be consuming a carnivore diet? What are they eating? Why did they start? What have they found?
Table 1, as usual, presented the characteristics of the people doing the diet enabling us to answer the “who?” question up front. The study was large. A total of 2,209 respondents were included in the research. The average age was 44 years and 67% of respondents were male. They had consumed a carnivore diet for 14 months on average. Their average (median) BMI was 24.3. The participants were primarily from the US and Canada (64%) but 11% were from the UK and Europe, 8% from Australia and 17% from other countries and so this was a global review. Participants were highly educated. Fewer than 1% had primary education only. Most (64%) had tertiary education and 26% had post-secondary education.
The “why?” was also addressed in Table 1. The vast majority of people were motivated to follow a carnivore diet primarily for health/body weight reasons (93% of participants). Next (41%) was “aiming for ketosis” followed by food preferences (33%). Health reasons were broken down further with the top reason being body weight/composition (84%), followed by focus/energy (74%), and then allergies/skin/autoimmunity (60%). Many conditions were reported as health reasons for following a carnivore diet. Digestive health was ticked by 52%, mental health by 45% and diabetes by 12%. One got the impression that there were many drivers for consuming a carnivore diet. We’ll see soon if the diet lived up to the hopes captured in the reasons for starting.
The diet (Table 2)
Table 2 reported the frequency of food intake. Dieticians look away now! Red meat was reported to be consumed at least daily, if not at every meal, by 85% of participants. Eggs were consumed at least daily, if not at every meal, by 44%. Non-milk dairy was consumed similarly (43%) but milk consumption was far less frequent with only 8% of people consuming it daily and less than 1% consuming milk at each meal. Conversely 65% of people studied never consumed milk. A similar number (64%) reported never consuming nuts. This makes me think that carbohydrate intake was being controlled by many people, as fluid dairy and nuts have a measurable carb content and are often restricted by those trying to achieve ketosis.
The highest “never eat” was for legumes (89%). The next highest was 79% of participants reported never eating grains, 74% never eating starchy vegetables, 69% not even eating non starchy vegetables and 66% never eating fruit. I suspect that the 34% who ate fruit occasionally consumed berries, which are the lowest carb fruits.
The food intake table also reported on supplements – 80% reported never taking multivitamins. Other vitamins were taken by 29% daily, 11% weekly but still 47% reported never taking other vitamins at all.
Alcohol was covered in the supplemental file rather than the main paper. 83% of people reported never consuming beer, 55% never consuming spirits and 57% never consuming wine. Those who did consume any alcohol were most likely to consume it less than monthly.
For those who like tables rather than words (me), here are the above numbers in a table. I’ve just taken the extremes of “each meal” and “at least daily” vs “ less than monthly” or “never”. (The difference between these numbers and 100% can be found in the middle four options, which were >weekly, weekly, > monthly, and monthly. I’ve also not taken all the lines from Table 2 – there were also lines for poultry, seafood, organ meat, processed meat and more food items):
The supplemental file also reported frequency of eating. The most common eating pattern (64% people) was twice a day. 17% of people ate once a day and 16% of people ate three times a day (the remaining couple of percent were four times a day or fewer than one time a day). The carnivore diet projects an image of ‘hunter’ style eating and supplemental Table 3 informed us that 76% of people consumed their meat medium rare, rare or raw. They didn’t worry about salt with nearly 80% of people having no specific salt intake or aiming for a high salt intake. 60% didn’t check for ketones.
Health changes – self reported (Table 3 & Figure 1)
Table 3 in the main paper was impressive. It reported which self-reported symptoms had changed when following the diet. People could answer that symptoms had resolved, improved, were unchanged, had worsened or had newly appeared. Obesity/overweight was reported as resolved in 52% of people and improved in 41%. Hypertension was reported as resolved in 61% of people and improved in 32%. The outcome for diabetes/insulin resistance was almost perfect – it was reported as resolved in 74% and improved in 24%. The following conditions scored greater than 80% (many nearer 90% or over) in combined resolved/improved: cardiovascular symptoms, gastrointestinal, endocrinologic, autoimmune, muscular skeletal, neurological, cognitive, psychiatric, respiratory, urologic, dermatologic, hematologic and other symptoms. If this were a medication, it would be a block buster. The highest condition that was unchanged was ophthalmologic at 51% unchanged – this condition still had 48% of issues reported as resolved or improved and barely 1% worsened or new.
The one condition that jumped out from the table as having any measurable worsening or newly acquired condition was lipid abnormalities. While these were resolved or improved in 27% and 29% respectively and unchanged in 18%, they worsened in 19% and became new issues in 8%. These figures are captured in the table below. (The numbers don’t add to 100 with rounding errors.)
This lipid issue has gained the most attention as an outcome of very low carbohydrate diets (which carnivore diets will be). Table 3 put this issue in context with the many significant improvements, which are not covered by critics of diets high in animal produce. We’ll come back to lipids in measured health changes in the next section.
Before moving on from health changes, we should note that the supplemental file contained even more detail. Supplemental Table 1 recorded perceived symptoms (all self-reported) on many varied conditions, from dry skin to muscle cramps to insomnia to concentration. Survey questions about these conditions were intended to detect nutrient deficiencies, which may have arisen as a result of the diet. The table compared the percentage of people self-reporting that the symptom was stable/improved vs new/worsened. Two symptoms recorded the same percentage incidence for stable/improved and new/worsened. These were a very small 0.7% for “weight-loss unintended” and an even smaller 0.1% for seizures. Every other perceived symptom in supplemental Table 1 reported higher stable/improved vs new/worsened. Some comparisons were multiple times better. Far from picking up nutrient deficiencies, the health questionnaire picked up further improvements in a variety of health measures.
Health changes – measured (Table 4)
Table 3 was based on self-reported improvements. Table 4 was based on self-reported test results. While still self-reported, therefore, the results in this table were more objective and measurable. Median BMI dropped from 27.2 pre-diet to 24.3 currently. (Weight dropped from a median of 85kg before the diet to 76kg currently).
Table 4 covered the lipid results in more detail. Lab results for cholesterol (total, LDL and HDL) were reported before the diet and “currently” for approximately 250 people. For these ‘paired’ before and after results, the median total cholesterol was 209 mg/dL before and 256 mg/dL currently, the median LDL-cholesterol was 126 mg/dL before and 172 mg/dL currently, and the median HDL-cholesterol was 58 mg/dL before and 68 mg/dL currently. This reflects some common findings with cholesterol markers and very low carb diets. The paper noted that the HDL-cholesterol and triglycerides were “optimal” and suggested that the LDL-cholesterol elevation, when associated with low triglycerides, “may reflect large, buoyant lipoprotein particles, possibly comprising a relatively low-risk subtype. Indeed, the low ratio of TG to HDL-cholesterol is suggestive of high insulin sensitivity and good cardiometabolic health.” For those interested, the reference cited by the paper for the large buoyant theory was Ahmadi et al (Ref 2) and the reference cited for the low ratio of triglycerides to HDL-cholesterol being a good indicator of metabolic health was Quispe et al (Ref 3). This carnivore paper might be useful to take to healthcare practitioners, who are worried about any change in lipid profiles, to illustrate that benefits of this way of eating likely far outweigh concerns.
Satisfaction & Support
Supplemental Table 5 captured the participant satisfaction with the diet and the response from healthcare practitioners, family and friends. One section of the survey asked the following questions:
How satisfied are you with the carnivore way of eating?
How has your medical provider responded to you eating this way?
How have friends / family responded to you eating this way?
The answers were (rounding errors):
The patients were far more content than the practitioners, therefore.
The caveat noted at the start is significant. It would have been very useful to include in the paper (or do a follow-up paper) people who started a carnivore diet and dropped out and to see when and why and what the characteristics of those people were. This research tells us – of the people who stuck to a carnivore diet for at least six months, how did they get on?
There is also a healthy person confounder going on here. The participants consumed alcohol far less than people following more typical (non-carnivore) diets. If the alcohol cutbacks/cessation started at the same time as the carnivore diet, this would have confounded health improvements. Additionally, exercise was not mentioned in either the main file or the supplemental. Instead, the main paper noted as a limitation of the research that “Adults adhering to a carnivore diet and responding to this online survey represent a special subpopulation with high levels of motivation and other health-related behaviors (e.g., physical activity, consumption of relatively whole, unprocessed foods.”
The limitations also noted an omission that I observed “we did not obtain detailed information on diet and lifestyle habits before beginning a carnivore diet.” It would have been useful to know what diet the 2,029 people were following before adopting a carnivore diet. If they were junk food monsters, the health improvements would be less surprising. However, I suspect that not many people go from junk food monster to carnivore. I suspect that many people ‘go carnivore’ from a real food diet and probably a real food, low-carb diet, and then they take a final plunge into carnivore. That’s my opinion (based on meeting many people at conferences and talking about their diet journeys). If correct, it makes the health improvements even more impressive.
Despite the caveats and confounders, this research is undeniably impressive. The health improvements reported are striking. I can’t help but think that a survey of over 2,000 people who had gone vegan for an average of 14 months would not have achieved such impressive results. Especially if the people had gone vegan from, say, being vegetarian and thus the final step was not as dramatic as from junk-food diet to plant-based, whole-food diet. Practitioners should always be open minded to the power of diet to improve health and they need to be open minded to the power of this diet, the carnivore diet, to improve health.
Ref 1: Lennerz et al. Behavioral Characteristics and Self-Reported Health Status among 2029 Adults Consuming a “Carnivore Diet”. Current Developments in Nutrition. December 2021. https://academic.oup.com/cdn/article/5/12/nzab133/6415894
Ref 2: Ahmadi et al. The impact of low serum triglyceride on LDL-cholesterol estimation. Arch Iran Med. 2008.
Ref 3: Quispe et al. Relationship of the triglyceride to high-density lipoprotein cholesterol (TG/HDL-C) ratio to the remainder of the lipid profile: the Very Large Database of Lipids-4 (VLDL-4) study. Atherosclerosis. 2015.