Keto Diet, High Cholesterol, and the Battle Over Arterial Plaque

The world of nutrition science is no stranger to passionate debates, but a recent study focusing on the ketogenic diet and its relationship with cholesterol and heart health has escalated these discussions into a full-blown conflict. At the heart of the controversy is a paper published in JACC: Advances, which examined the effects of a high-fat, low-carbohydrate diet on arterial plaque accumulation in a specific group of individuals. The study's interpretation and subsequent fallout have ignited a firestorm on social media and within the research community, raising critical questions about scientific rigor, data transparency, and the dissemination of health information.

The ketogenic diet, often simply called 'keto,' is a dietary pattern characterized by very low carbohydrate intake, moderate protein, and high fat consumption. The goal is to shift the body's primary energy source from glucose (derived from carbohydrates) to ketones (produced from fat). This metabolic state, known as ketosis, is pursued by millions, primarily for weight loss, but also for potential benefits in managing conditions like epilepsy and type 2 diabetes. Despite its popularity, the long-term health implications of consistently consuming high amounts of fat, particularly saturated fat, remain a subject of intense debate, especially concerning cardiovascular health.

A cornerstone of modern cardiovascular medicine is the lipid hypothesis, which posits a causal link between elevated levels of low-density lipoprotein cholesterol (LDL-C), often dubbed "bad" cholesterol, and the development of atherosclerosis – the buildup of fatty plaque in the arteries. This plaque can narrow blood vessels, restrict blood flow, and increase the risk of heart attacks and strokes. Consequently, medical guidelines universally recommend keeping LDL-C levels low, often through diet, lifestyle changes, and statin medications.

However, the ketogenic diet can sometimes lead to significant increases in LDL-C levels in certain individuals. This phenomenon is particularly noted in a subset of people who maintain a lean physique and good metabolic health despite having unusually high cholesterol on a high-fat diet. These individuals have been termed "Lean Mass Hyper-Responders" (LMHRs).

The LMHR Study: Aims and Participants

The controversial study aimed to investigate whether this specific group of LMHRs, characterized by high LDL-C levels while following a ketogenic diet, were indeed developing arterial plaque. The trial recruited 100 participants who had been following a ketogenic diet for at least two years, were otherwise healthy, but exhibited high blood cholesterol. Participants were followed for one year, undergoing assessments of arterial plaque levels at the beginning and end of the study period.

One of the key figures associated with the study and the LMHR concept is Dave Feldman, a software engineer and entrepreneur. Feldman, who lacks formal medical or nutritional training, has become a prominent advocate for the keto diet and has publicly challenged the conventional understanding of cholesterol's role in heart disease, particularly in the context of LMHRs. He claims to have coined the term LMHR in 2017 and has previously conducted informal, non-institutional experiments to draw attention to this phenotype.

The recent *JACC: Advances* study was crowdfunded by Feldman's charity, the Citizen Science Foundation, and conducted through the Lundquist Institute, a research organization that provided institutional review board (IRB) oversight to ensure ethical standards were met.

Initial Claims and Widespread Criticism

Upon the paper's publication on April 7, Dave Feldman released a video on social media claiming that the study found no association between LDL cholesterol and plaque in the participants, and similarly, no association between apolipoprotein B (ApoB) and plaque. ApoB is a protein component of lipoproteins like LDL and is considered a strong predictor of cardiovascular risk, often seen as a more accurate measure than LDL-C alone because it counts the number of potentially harmful particles.

Feldman's interpretation suggested that, contrary to established science, high LDL-C and ApoB levels in LMHRs following a keto diet were not increasing their risk of arterial plaque. This interpretation was quickly amplified within online communities supportive of low-carb, high-fat diets, presented as potential evidence against the long-standing lipid hypothesis.

However, this interpretation was met with immediate and strong pushback from numerous doctors and researchers in the nutrition and cardiology fields. Many who reviewed the published paper reached the opposite conclusion from Feldman's initial claims. Critics voiced their concerns publicly, particularly on social media platforms, highlighting several significant issues with the study and its presentation:

• Selective Reporting: A major point of contention was the apparent masking of the study's original primary outcome. The study was initially designed to measure the percentage change in non-calcified plaque volume (NCPV) – soft plaque that is considered particularly vulnerable – over the one-year period. While a graph showing NCPV changes was included, the specific measurements and their statistical significance were not explicitly provided or discussed in the main findings.
• Shifted Focus: Instead of focusing on the pre-specified NCPV outcome, the paper seemed to pivot to an exploratory analysis suggesting no association between ApoB and plaque progression. Critics argued that this analysis was not the study's original purpose and that the data collected might not be suitable or sufficient to support such a conclusion robustly. Shifting the focus after data collection without clear pre-registration of the new hypothesis is considered poor scientific practice, potentially allowing researchers to find and highlight statistically significant results by chance or through data manipulation (known as p-hacking or HARKing - Hypothesizing After the Results are Known).
• Lack of Comparator Group: The study lacked a control group (e.g., individuals with high LDL on a standard diet, or LMHRs who lowered their LDL). This makes it difficult to definitively attribute observed changes (or lack thereof, based on the initial interpretation) solely to the keto diet or the LMHR phenotype, or to compare the risk profile of LMHRs to other populations.
• Statistical Modeling Concerns: Experts questioned the validity and appropriateness of the statistical models used in the paper to support the claim of no association between ApoB and plaque progression.
• Short Timeframe: A one-year follow-up period was considered relatively short for assessing the progression of atherosclerosis, which is typically a slow, chronic process. Critics argued that significant plaque progression might not be detectable in such a limited timeframe, potentially leading to a false sense of security.

Spencer Nadolsky, a physician specializing in obesity medicine and lipidology, was particularly vocal in his criticism, going so far as to call for the paper's retraction. Nadolsky had a unique perspective, as he was initially involved in the study's design phase.

Problems from the Beginning: An Insider's View

The collaboration between Dave Feldman and Spencer Nadolsky stemmed from years of online debate. Feldman, promoting his alternative "lipid energy model" – an unproven theory suggesting that high LDL in LMHRs is benign because their bodies are efficient at transporting cholesterol while burning fat – sought scientific validation. Nadolsky, while adhering to the consensus lipid hypothesis, was interested in gathering data on LMHRs and saw the potential for a study to bridge the gap between the two camps.

However, designing a study to test Feldman's hypothesis faced ethical hurdles. An IRB would likely reject a trial that required individuals with extremely high LDL-C (known to be risky) to remain untreated. The workaround was to recruit LMHRs who were already following a keto diet, had high LDL-C, and were choosing not to take lipid-lowering medications. This observational approach, while ethically permissible as it didn't involve withholding standard treatment, had inherent limitations compared to a controlled trial.

Nadolsky's concerns grew during the recruitment and promotion phase of the study. Recruitment efforts were heavily reliant on social media platforms like X (formerly Twitter) and Feldman's LMHR Facebook group, which also served as fundraising channels. Nadolsky observed Feldman presenting preliminary data at a low-carb conference, suggesting that most participants showed no plaque at baseline, implying the LMHR phenotype was benign. Nadolsky felt this was an inappropriate presentation of unvetted data to recruit subjects and donations, essentially presenting supposed findings before the study was completed or properly analyzed.

Consulting with external scientists, Nadolsky was advised to distance himself from the project due to concerns about potential data spin, regardless of the actual results. He filed a complaint with the Lundquist Institute's IRB regarding ethical concerns, but according to Soto-Mata, the IRB reviewed the matter and allowed the study to proceed, finding no ethical transgressions. Nadolsky ultimately left the study team while recruitment was still underway.

Entrenched Positions and Interpretative Disagreements

Kevin Klatt, an assistant research scientist at UC Berkeley specializing in nutrition, has closely followed the study and the ensuing controversy. He has written extensively about the topic on his personal platform and shares many of Nadolsky's concerns regarding the study's design, execution, and the potential for bias, particularly given Dave Feldman's vested interest and lack of biomedical training.

Klatt also raised his concerns about undisclosed biases to the Lundquist Institute, but his email reportedly went unanswered. He views the situation as highly unethical, especially the pre-publication presentation of preliminary data used for recruitment and fundraising.

While some researchers are using the study to reinforce warnings about the potential adverse effects of keto, Klatt takes a more cautious stance on interpreting its initial presentation. He notes that the debate often involves two entrenched camps: those who firmly believe in the traditional lipid hypothesis and those who are open to or advocate for alternative theories like the lipid energy model. Klatt places himself in a third camp, questioning the value of drawing strong conclusions from a study with so many apparent methodological and reporting issues. As an editor for a nutrition journal, he stated that he would likely have rejected the paper outright due to its obvious problems.

Klatt's primary concern is the potential for a flawed study to be used by the public and online communities to prematurely declare the consensus on LDL cholesterol risks as "debunked," which he emphasizes is not the case.

Adrian Soto-Mota, the lead author of the study, defended the research, stating that interpretative disagreements are common in nutritional science. He maintained that all study limitations were acknowledged in the paper and that alternative statistical models suggested by critics still corroborated the paper's conclusions (referring to the initial, controversial conclusions).

Soto-Mata also disputed the claim that the focus was switched, asserting that changes in NCPV were shown in a graph and used in "almost all the analyses." He defended the analysis regarding ApoB and plaque as plausible and independently reviewed by a statistics expert during peer review.

Matthew Budoff, a professor of medicine at UCLA and an investigator at the Lundquist Institute, and a coauthor on the study, acknowledged the "incredible scrutiny" the paper received on social media. He mentioned that the research team was seeking to incorporate corrections into the paper, subject to the journal's discretion, and that a response to the Letter to the Editor would clarify some issues.

The Unveiling of the Data and the Aftermath

The most significant development came with the publication of the authors' reply to the Letter to the Editor. This response revealed crucial data that had been downplayed or omitted in the initial presentation. The authors stated that the "pooled median change" in NCPV – the very measure the study was initially designed to assess – was a substantial 42.8 percent over the one-year period. This indicates a significant progression of arterial plaque in the LMHR participants.

Furthermore, the authors' reply explicitly stated that the study's findings were "compatible with a causal role of ApoB in atherosclerosis," a position they acknowledged and supported in previous publications. They attributed the omission of the 42.8 percent NCPV increase in the original paper to a "sincere oversight, not intentional selective reporting."

This clarification dramatically shifted the interpretation of the study's results, bringing them back in line with the conventional understanding of cholesterol and ApoB's role in heart disease. The data, when fully revealed, did not support the initial claims that high LDL/ApoB in LMHRs was benign; instead, it showed significant plaque progression, consistent with the established lipid hypothesis.

However, the damage from the initial presentation had already been done. Dave Feldman's interpretation, suggesting high LDL in LMHRs was not linked to plaque, had spread rapidly through online keto and low-carb communities. Given the immense popularity of the keto diet, information from the study, filtered through the lens of advocacy, reached a wide audience before the scientific community's criticisms or the authors' clarifying response gained traction. The lipid energy model and the idea that LMHRs are somehow protected from the risks of high cholesterol have already entered the public discourse, even appearing in summaries provided by AI models like ChatGPT when asked about LMHRs.

The controversy highlights the challenges of communicating complex scientific findings, especially in the age of social media where information can be rapidly disseminated and amplified outside traditional peer-reviewed channels. It also underscores the potential dangers of non-scientists leading or interpreting research without sufficient checks and balances, particularly when they have a strong pre-existing bias or theory they are seeking to validate.

The case of the LMHR study serves as a cautionary tale about the importance of rigorous study design, transparent data reporting, robust peer review, and responsible communication of scientific results. While the authors have since clarified the findings, revealing data that supports the conventional view on cholesterol and plaque progression, the initial misinterpretation has already fueled a narrative that could potentially mislead individuals about their cardiovascular risk while following high-fat diets.

Understanding the Science: LDL, ApoB, and Atherosclerosis

To fully grasp the significance of this debate, it's essential to understand the underlying science of cholesterol and heart disease. LDL particles are responsible for transporting cholesterol from the liver to the rest of the body. While cholesterol is necessary for cell function, high levels of LDL particles, particularly small, dense ones, can penetrate the arterial wall. Once inside the wall, these particles can become oxidized and trigger an inflammatory response.

Macrophages, a type of immune cell, engulf the modified LDL particles, becoming "foam cells." These foam cells accumulate, forming fatty streaks, the earliest signs of atherosclerosis. Over time, smooth muscle cells migrate to the area, a fibrous cap forms over the lipid core, and calcium can be deposited, leading to hardened plaque. This plaque narrows the artery and can rupture, leading to blood clot formation that blocks blood flow, causing a heart attack or stroke.

ApoB is a protein found on the surface of several types of lipoproteins, including LDL, VLDL (very low-density lipoprotein), and IDL (intermediate-density lipoprotein). Each of these potentially atherogenic particles contains one ApoB molecule. Measuring ApoB levels essentially counts the number of these particles. Research has increasingly shown that ApoB is a better predictor of cardiovascular risk than LDL-C alone, especially in individuals with metabolic syndrome or diabetes, who may have a higher proportion of small, dense LDL particles.

The traditional lipid hypothesis, supported by decades of epidemiological studies, genetic research, and clinical trials (including those using statins and other lipid-lowering therapies), demonstrates a clear causal relationship between elevated LDL-C and ApoB levels and the risk of atherosclerosis and cardiovascular events. Lowering LDL-C and ApoB has consistently been shown to reduce this risk.

The lipid energy model, proposed by Dave Feldman and others, suggests that in individuals adapted to burning fat for energy (like those on a ketogenic diet), the dynamics of lipid transport change. They hypothesize that high LDL in LMHRs might represent increased transport capacity rather than increased risk, or that the particles themselves might be less atherogenic in this metabolic state. However, this model lacks robust scientific evidence from controlled studies demonstrating that high LDL/ApoB in LMHRs does not lead to increased plaque or cardiovascular events over the long term. The findings from the *JACC: Advances* study, particularly the revealed 42.8% increase in NCPV, directly contradict the core tenet of the lipid energy model as it relates to plaque progression.

The Role of Social Media and Citizen Science

The controversy also highlights the complex role of social media and the rise of citizen science in health research. Platforms like X allow for rapid dissemination of information and direct engagement between researchers, advocates, and the public. This can democratize access to information and foster community, but it also creates an environment where preliminary findings, personal anecdotes, and interpretations lacking scientific consensus can gain significant traction and influence public opinion before formal peer review and critical evaluation are complete.

Citizen science, where non-professional scientists participate in research, can be a valuable tool, particularly for data collection or hypothesis generation. However, when citizen scientists take on roles typically reserved for trained researchers – such as study design, data analysis, and interpretation – without adequate oversight or collaboration with experienced academics, it can lead to methodological flaws and biased conclusions, as critics allege occurred in this case.

The crowdfunding aspect also introduces potential biases. When research is funded by a community with a vested interest in a particular outcome (e.g., validating the safety of a diet they follow), there can be pressure, conscious or unconscious, to interpret data in a favorable light. While crowdfunding can enable research that might otherwise not be funded, it necessitates extra vigilance regarding transparency and potential conflicts of interest.

Looking Ahead: The Need for More Research and Clear Communication

Despite the controversy, the study did bring attention to the LMHR phenotype, which warrants further investigation. While the *JACC: Advances* study's data, when fully disclosed, aligns with the conventional understanding of atherosclerosis progression in the presence of high ApoB, it was a relatively small study with a short follow-up period, focused on a specific, self-selected population.

More rigorous, long-term studies are needed to fully understand the cardiovascular risks associated with diet-induced hypercholesterolemia in LMHRs. These studies should ideally include control groups, utilize robust methodologies for assessing plaque progression (like advanced imaging techniques), and be conducted by researchers with appropriate expertise, free from significant conflicts of interest.

In the meantime, the scientific community faces the challenge of correcting misinformation that has already spread. This requires clear, accessible communication about the study's actual findings, the limitations of the research, and the strong existing evidence supporting the link between high LDL/ApoB and cardiovascular disease. It also highlights the responsibility of journals to ensure the rigor of published research and the clarity of its presentation.

The "war" in the nutrition world over this study underscores the passionate beliefs surrounding dietary choices and health outcomes. While healthy skepticism and the questioning of established paradigms are vital for scientific progress, they must be grounded in sound methodology, transparent data, and objective interpretation. For individuals navigating conflicting health information, relying on the broad consensus of the scientific and medical community, based on the totality of evidence from multiple high-quality studies, remains the most prudent approach to protecting their health.

The debate surrounding the LMHR study serves as a stark reminder that even seemingly straightforward scientific findings can be subject to intense scrutiny and conflicting interpretations, particularly when they touch upon popular dietary trends and challenge long-held medical beliefs. The ultimate goal of nutrition science is to provide evidence-based guidance for improving public health, a goal that is best served by rigorous research, open data, and a commitment to scientific integrity above advocacy or personal theories.