Is Keto Bad for the Heart?

Trusted health groups like the Amer. Heart Assoc. would like you to think so.

Last week saw another spate of anti-keto headlines stemming from an authoritative report, this time from the American Heart Association (AHA), warning that the “keto” (ketogenic) diet is bad for heart health. According to an AHA expert panel, low-carb diets, including keto and paleo, should be ranked last among 10 popular diets assessed. Mediterranean, low-fat, vegan, vegetarian and DASH (Dietary Alternatives to Stop Hypertension) were all rated higher

This list looked a whole lot like one in January, in U.S. News & World Report’s “Best Diets,” issue, which gave low marks to keto and paleo. And last month, we saw another 30-plus keto-bashing headlines, at CNN.com, Forbes, Salon, WebMD, Medscape and other outlets, with altogether millions of social-media views, trumpeting a study presented at a conference of the American College of Cardiology (ACC), which purported to show that a “keto” or “keto-like” diet was associated with elevated LDL-cholesterol (the so-called “bad” one) and a higher risk of cardiovascular events.

Upon inspection, it turns out that the science behind these claims is not just flimsy, it’s non-existent. Is it possible we’re seeing a campaign of weak science being corralled to scare people away from the ketogenic diet? In the case of the ACC reports, for instance, there’s not even a pre-print of the study to justify the dozens of headlines.

But first, let’s look at the science on whether keto might actually be dangerous for heart health. We have no ‘hard’ data, i.e., evidence from clinical trials on heart attacks and deaths, to know the long-term effects of low-carb. Some observational studies have purported to find an association between low-carb and increased mortality, but these papers defined the diet, variously, as having between 37% to 50% of calories as carbohydrate, whereas top researchers in the field use the numbers 25-30%.¹ The papers using higher numbers are therefore not relevant to low-carb—or keto, whose cap on carbs is even lower.

Since we don’t have “hard data” on death or heart attacks, we can only look at the next-best thing: recognized cardiovascular risk factors such as cholesterol, which might predict longer-term outcomes. Here, we find multiple clinical trials showing that a ketogenic approach resolves high blood pressure (often within weeks) and improves most heart-disease markers. In a sizable trial after 1 year, a ketogenic approach improved 23 of 26 heart-disease risk factors measured, with the overall cardiovascular risk score dropping by nearly 12%, according to the ACC’s own “ASCVD risk calculator.” Of course more research is needed, but according to these data, a ketogenic approach seems safe and indeed, promising for combatting heart disease.

Given these data, how should we think about the allegations against keto? I already wrote about the scant science behind the U.S. News’ “Best Diets” list. Four of its highly ranked diets hadn’t even been tested in a single clinical trial.

The AHA and ACC, however, are more prominent authorities. These groups may have been heavily funded by pharmaceutical companies, yet they still command respect among doctors, and their guidelines are widely followed. And if they have solid science on the dangers of keto, then we should be grateful that such authorities are issuing caution. So let’s take a look.

The AHA report

Headlines such as “Popular keto and paleo diets aren’t helping your heart” in the Washington Post give the impression that the AHA authors looked at heart-disease outcomes of these diets in some systematic way, but in fact, they did not. Instead, the aim of the paper was to examine how closely each diet aligned with the AHA’s own diet plan. The authors asked, essentially: Are you like us? By this standard, the #1 diet on their list is DASH, because…that’s the AHA’s own diet! The AHA report was a self-affirmation exercise, a purity test. Headlines should have read: the AHA likes its own diet best and disapproves of others.

Although a rigorous analysis of any diet was outside the scope of the paper, the AHA authors nevertheless took the opportunity to lob some time-worn pot shots at keto. The lead author, an established vegan advocate (also a panelist for U.S. News’ “Best Diets”) warns about the “keto flu,” a complaint dating back some 25 years that’s been answered in too many books and articles to count (in short, this “flu” is a transient condition that can easily be avoided by drinking soup broth). The authors also criticize the keto diet’s liberal approach to saturated fats (see my tweet thread on that issue); the inability to adhere to such a restrictive diet (Hello, vegans, who’ve eliminated all dairy, meats, shellfish, fish and eggs…exactly who’s restrictive?); and my favorite: “may limit hunger while leading to weight loss.” (Sounds terrible!)

The bottom line here is that the AHA paper did not look systematically at evidence on heart-disease outcomes for any diet. Whether its stated aim, to evaluate diets against the AHA’s own “gold standard,” stands up to scrutiny depends on whether the AHA nutritional plan itself has been demonstrated to protect against heart disease. That’s a topic for another post, but spoiler alert: in the AHA’s most recent systematic review, in 2019,² only a single clinical trial is cited to support the group’s nutrition recommendations. Can you guess which study that is? (Further, a single trial does not actually rise to the AHA’s own evidentiary standards, which call for multiple randomized clinical trials to substantiate guidelines. But a lack of evidence for guidelines is unfortunately not a rare occurrence for the AHA: a 2019 analysis published in JAMA found that only 8.5% of AHA/ACC guidelines met their evidence standards.)

The ACC News

In the case of the dozens of news items on a presentation given at the American College of Cardiology conference in March, these reported that a “keto” or “keto-like” diet is associated with higher LDL-cholesterol and greater risk of cardiovascular events. None of these findings could be confirmed, however, because the lead researcher for this study, Iulia Iatan, a physician-researcher for the University of British Columbia and Healthy Heart Program Prevention Clinic at St. Paul’s Hospital in Vancouver, only spoke about her research at the conference. Nothing is in print, not even an abstract or pre-print of the study, which are typical ways that researchers present preliminary work.

Still, the ACC issued a press release on Iatan’s utterings. Among the ACC’s 40-plus press releases issued during the conference, I found that only two were based on unpublished, unregistered work, Iatan’s included.³ I asked the ACC if the group had any evidentiary standards for issuing a press releases, i.e., why merely spoken words at a conference would justify such media amplification, and a representative explained that sometimes “clinical research data” were selected for press releases--although Iatan’s study was not clinical research (it was an observational study, a much weaker form of evidence). 

Contacting Iatan proved fruitless. She did not respond to emails or phone calls. Amidst my attempts to reach her, a press rep from her employer, Ann Gibbon, popped into my email box to say that Iatan was not available.

Gibbon and I had a further back-and-forth about a somewhat Orwellian aspect of Iatan’s study. Despite the dozens of headlines stemming from Iatan’s presentation on a “keto” and “keto-like” diet, I discovered that Iatan had not, in fact, looked at a keto diet, which is typically defined as having 20 grams of carbs or less; Iatan’s study used a definition of five times that, or 100 grams. She herself described her study diet, accurately, as “low-carb, high-fat” in the ACC press release but then called it “keto-like” in another report. 

Definitional accuracy is a fundamental starting point for good science, yet no one involved in the ACC “news” event seemed to have minded about the serious mislabeling problem. Indeed, Iatan, the University of British Columbia, and the ACC all posted headlines about a “keto” diet.

Still, when I queried Gibbon about the inaccurate use of “keto,” she said she would demand a correction if I used that word to describe Iatan’s study, even though a “keto” post was at that moment on Iatan’s own university website, and apparently Gibbon wasn’t pursuing corrections on this issue with the ACC, CNN.com, GoodMorningAmerica.com, Salon, Fortune, Fox.com, Medscape, WebMD, Medical News Today, or any of the other 20+ outlets that had reported on the study as “keto” or “keto-like.”

Campaign against keto?

When people say that nutrition science is difficult to understand, they’re 100% correct. Often, there are genuine disputes over complex science. However, I think the phenomenon we may be seeing here is something different: an apparently purposeful amplification of non-existent “findings” on “keto.” I don’t know whether this represents a concerted effort to denigrate the diet. Certainly, one can imagine many potential motives.

For instance, keto, as the virtual opposite of the official recommendations of these groups (AHA and ACC, as well as the US. Dietary Guidelines for Americans), must trigger intellectual and cognitive dissonance for experts who have ‘known’ differently for their entire careers. Also, it feels cynical to say but is nonetheless undeniably true that because keto allows people to get off many or all of their medications, the diet poses a clear threat to pharmaceutical companies. Chronic diseases that require lifelong medications are a sweet spot of assured revenue for this industry. After all, pharmaceutical companies spent more than $8 billion on media advertising last year (up from $6 billion in 2020), and their aim is not for you to consume fewer of their products. This fact may also go some ways towards explaining why many media outlets are so quick to publish non-existent keto findings.

The ACC reported receiving 38% of its revenue from industry in 2012, and the AHA reported 20% in 2014.⁴ (If anyone knows more recent numbers, I’d love to see).

It’s one thing not to trust the media. Half of all Americans now believe news organizations deliberately mislead them--an historic low. However, distrusting groups like the AHA and ACC comes harder.

On the bright side, when your doctor or friend who’s read the spate of headlines this year tells you to steer clear of that “dangerous” keto diet, you can tell them that’s fake news. And when the next round of bad keto news comes, you’ll know to read between the headlines and check the facts.  

1. Actually, leading experts prefer a definition based on the absolute number of 130 grams of carbs per day; “keto” is generally 20 grams or less).

2. A 2021 review exists, but it is not systematic.

3. Some ACC press releases were on unpublished results, but the clinical trial itself was registered, usually in a government data base. Aside from Iatan’s work, the only other other press release I could find that was issued on an entirely unpublished study was here.

4. https://www.bmj.com/content/351/bmj.h4962