What 2,029 Carnivore Dieters Reported, and What That Can't Tell Us

If you’ve spent five minutes in carnivore circles, someone has waved “the Harvard survey” at you like it settles the argument. The study is real. It’s interesting. And it’s worth understanding properly, which means being clear about what it found and what it was never built to show.

I read the whole thing, methods and all. Here’s the rundown.

The study in brief

In 2020, a team led by Belinda Lennerz and David Ludwig ran an online survey of adults who said they’d eaten a carnivore diet for at least six months. They recruited through social media, mostly carnivore communities on Facebook, Instagram, Reddit and Twitter. Data came in between March and June 2020, right in the first COVID-19 lockdowns. The results were published in Current Developments in Nutrition in 2021.¹

Who answered: 2,029 people. Median age 44. 67% men. They’d been eating carnivore for a median of 14 months, and 93% said they started for health reasons.¹

What they ate

This wasn’t some loose “low-carb-ish” crowd. These folks meant it:¹

• 85% ate red meat daily or more often.
• Under 10% ate vegetables, fruits or grains more than once a month.
• 42% ate organ meat weekly or more; 72% ate non-milk dairy (cheese, butter) weekly or more.
• Most ate twice a day (64%), and they liked their cuts fatty (61%) over lean (2%).
• 37% took no vitamin supplements at all.

What they reported

Respondents described few side effects. Reported rates of possible adverse symptoms ran from under 1% to 5.5%, with gastrointestinal complaints at the top (3.1–5.5%).¹ They also reported:

Measure	What respondents reported
Overall health improved	95%
Well-being improved (various measures)	66–91%
Median BMI	27.2 before → 24.3 now
Improvement in existing medical conditions	48–98%, depending on condition

People with diabetes reported a median BMI drop of 4.3 points, a median HbA1c reduction of 0.4 percentage points, and cutting back on diabetes medication.¹

Impressive on paper. Hold that thought.

The cholesterol picture

Among the subset who reported recent blood work, results were mixed. Median LDL cholesterol was markedly elevated at 172 mg/dL (about 4.4 mmol/L). HDL (68 mg/dL) and triglycerides (68 mg/dL) landed in ranges usually considered favourable.¹ The authors themselves called cardiovascular risk factors “variably affected” and said the long-term effects need more study.

If you take one thing from this article, take this: get a baseline lipid panel, repeat it, and go over it with your doctor. Don’t guess. Measure.

Why the evidence is rated “low”

This isn’t a knock on the study. It was explicitly descriptive. Almost no data existed, so the goal was to document what these people eat and how they feel. Fair enough. But its design means it can’t answer the question everyone actually wants answered: “will this work for me?” Here’s why.

1. Selection bias. They recruited from fan clubs. People who tried it, felt lousy and quit weren’t hanging around those groups, let alone filling out surveys.
2. Everything was self-reported. Health conditions, weight history, even lab values came from memory or people’s own records. Nobody checked.
3. No control group. Without a comparison group, you can’t separate the diet from the weight loss, ditching ultra-processed food and booze, paying more attention to health, or just time passing.
4. A snapshot. One moment, about a year in. Says nothing about year five.

A group of researchers made many of these points in a published response, flagging especially the unverified self-reported metabolic data and unvalidated questionnaires.² Fair criticism.

A footnote from 1928

Living on meat isn’t a TikTok invention. In 1928, Arctic explorers Vilhjalmur Stefansson and Karsten Andersen agreed to eat nothing but meat for a year under the watch of researchers at Bellevue Hospital in New York.^{3, 4} At the end, the investigators reported both men were “mentally alert, physically active” with no specific physical changes or vitamin deficiencies, mild ketosis, and no evidence of kidney damage on the tests available.³

Two details usually get left out. First, when Stefansson was put on a lean-meat diet early on, he got sick within days. Adding fat back fixed it.⁴ Their diet got most of its calories from fat.³ Second: it was two men, and the work was partly funded by the Institute of American Meat Packers.³ Great story. Not proof.

The bottom line

The survey shows a group of committed carnivore eaters were, by their own account, satisfied and feeling good, with few side effects, but with notably high LDL. It’s a solid starting point that justifies real clinical trials. It does not show the diet causes those benefits, or that it’s safe for everyone long term.

If you’re trying it anyway, the practical lessons are simple. Eat enough fat. Watch your bloodwork. Loop in your doctor, especially if you take medication for diabetes or blood pressure.