Why the Same Low-Carb Diet Raises Cholesterol in Some People (Not Others)
What Happened
Two people can eat what they believe is the same low-carbohydrate diet and get opposite cholesterol results. New research from Stanford University School of Medicine suggests part of that difference is written in their genes. The findings were presented at NUTRITION 2026, the annual meeting of the American Society for Nutrition, by Alexa Barad, PhD, RDN, a postdoctoral scholar at Stanford, and her colleagues.
This is the first thing to know and the easiest to miss: a conference presentation is not a peer-reviewed publication. The work has been presented and reported, but it has not yet gone through journal review, and the researchers themselves describe the findings as early evidence that needs confirmation. We are covering it because it is directly relevant to a question our readers ask constantly — "why did my LDL go up when my friend's did not?" — and because the underlying dataset is a real randomized trial, not an observation from a diet forum.
How the Analysis Was Built
The researchers went back to the DIETFITS trial, a randomized study that assigned more than 600 adults to either a healthy low-carbohydrate or a healthy low-fat diet for a year. The original trial's headline result was that neither diet won on average weight loss, a finding we covered in detail in our DIETFITS study read.
For this new analysis, the team focused on 431 participants who had genetic data available. Rather than testing single genes — a method that produced conflicting results for years — they aggregated thousands of small genetic variations into a single polygenic score estimating each person's inherited tendency toward higher LDL cholesterol. They then compared that score with diet assignment, saturated fat intake, and changes in LDL cholesterol during the first six months of the trial.
The Numbers and What They Showed
The pattern was clearest in the low-carbohydrate group. Participants with a higher polygenic LDL score were more likely to see LDL cholesterol rise; the largest increases appeared in those who combined a high genetic score with higher saturated fat intake. The same relationship did not appear in the low-fat group, which is important — it suggests the finding is not simply that some people have genes for high cholesterol, but that the low-carbohydrate dietary context changes how strongly saturated fat moves LDL.
"Our findings suggest that some individuals may be more sensitive to the LDL cholesterol-raising effects of saturated fat, particularly in the context of a low-carbohydrate diet, because of their genetic background," Barad said. She also pushed back on both slogans the field defaults to — that low-carb diets always raise LDL, or that they are universally heart-healthy — describing each as incomplete.
Two comparison numbers give the finding its scale. Current dietary guidance generally recommends keeping saturated fat below 10% of daily calories, and the standard advice is to replace some saturated fat with unsaturated fat from foods like olive oil, nuts, seeds and avocado. This analysis does not challenge that guidance; if anything, it gives a mechanism for why the substitution matters more in some people than others.
What It Means If You Eat Low-Carb
The honest framing is that this is a hypothesis with a plausible mechanism, from a modest genetic subset of one trial. It does not mean you need a genetic test before choosing a low-carb diet, and it does not identify a gene you can look up. What it does suggest is a practice that costs nothing: test your cholesterol before a major dietary change and again a few months after, then adjust based on your own numbers rather than on somebody else's internet result.
It also sharpens the food-quality question that sits underneath the carb-counting question. A low-carb diet that replaces bread with nuts, seeds, fish and avocado is not the same dietary pattern as one that replaces bread with butter, fatty processed meat and cheese — even though both are low-carb on a label. Here is where we have to be careful about what our own numbers can and cannot say: our library prices carbohydrate, not fat quality. We do not have polygenic scores or saturated fat gram data in our verified dataset, so we cannot tell you which of these foods is better for your LDL. What we can do is price the carbohydrate cost of each swap so the decision is at least visible:
- Butter: 0.06 g net carbs per 100 g (FDC #173410) — the lowest possible carbohydrate cost, and the food the study's finding is most directly about.
- Cheddar cheese: 3.37 g net carbs per 100 g (FDC #173414); heavy cream: 2.84 g (FDC #170859); bacon: 1.35 g (FDC #167914).
- Avocado: 1.83 g net carbs per 100 g (FDC #171705) — an unsaturated-fat source at a comparable carbohydrate cost to cheese.
- Pumpkin seeds: 4.71 g per 100 g (FDC #170556); pecans: 4.26 g (FDC #170182); almonds: 10.11 g (FDC #170158, which is why a weighed 28 g handful at about 2.8 g is the relevant serving).
Read that list and the real tension becomes clear: the unsaturated-fat swaps are not carbohydrate-free, but in weighed servings the difference is small — a 28 g handful of almonds costs about 2.8 g net carbs. That is the arithmetic a low-carb eater is actually choosing between. Which fat is better for your arteries is a question our data cannot answer and your bloodwork can. Our keto versus low-carb explainer covers why the two patterns produce different metabolic profiles even at similar carb counts.
Source Box
- Presentation: Barad A, et al. "Genetic predisposition, saturated fat intake and LDL cholesterol response in the DIETFITS trial," presented at NUTRITION 2026, American Society for Nutrition (nutrition.org) — conference presentation, not yet peer reviewed
- Reporting and researcher quotes: Medical Daily, "Your Genes May Influence How Your LDL Cholesterol Responds to a Low-Carb Diet, Study Finds," September 2026 (medicaldaily.com)
- Underlying trial: Gardner CD, Trepanowski JF, Del Gobbo LC, et al. "Effect of Low-Fat vs Low-Carbohydrate Diet on 12-Month Weight Loss," JAMA 2018;319(7):667-679 (DOI link)
All findings and quotes above are drawn from the coverage and the presentation as reported; the polygenic analysis itself has not been published in a peer-reviewed journal at the time of writing, and we have labelled it accordingly. Evidence grade: secondary genetic analysis of a randomized trial, 431 participants with genetic data, six-month LDL outcomes, presented as a conference abstract. This article is a plain-language summary by a data editorial team — it is not medical advice, and cholesterol decisions belong with your clinician.
Does this mean low-carb diets raise cholesterol?
No, and the researchers say so directly. The analysis found that LDL response varied with a person's genetic predisposition and their saturated fat intake, particularly within the low-carbohydrate group of the DIETFITS trial. It does not show that cutting carbohydrates raises LDL on its own. The lead author's own framing rejects both blanket claims — that low-carb always raises LDL, and that it is universally heart-healthy — as incomplete. Source: NUTRITION 2026 presentation; Medical Daily coverage, September 2026.
Should I get a genetic test before starting low-carb?
Nothing in this research supports that step. The polygenic score used here was a research tool built from thousands of variants across a trial population; it is not a consumer test, and the findings are not yet peer reviewed. The practical takeaway the researchers emphasize is cheaper and more direct: check your cholesterol before a major dietary change and again a few months in, then act on your own results. Discuss any dietary change and any lipid concern with your clinician.
Which fats should replace carbohydrates on a low-carb diet?
Our data cannot answer that — we price carbohydrate content, not fat type or cardiovascular risk, and we will not imply otherwise. What the wider evidence and this analysis point to is that a low-carb pattern built on nuts, seeds, fish, avocado and unsaturated oils looks different from one built on butter, fatty processed meat and large amounts of cheese, even when both carry the same carb count. A 28 g handful of almonds, for reference, costs about 2.8 g net carbs (from FDC #170158), so the carbohydrate cost of choosing unsaturated sources is usually small. Source: American Society for Nutrition guidance on saturated fat; NUTRITION 2026 presentation.