Emma Lawson
You’ve probably seen the headlines this spring bouncing in different directions. One week, keto’s fixing your pancreas. The next, it’s destroying your liver. If you’ve been low-carb for a while, or you’re thinking about it for blood sugar control, the whiplash is real. I’ve had clients text me photos of contradictory articles asking which one’s true, and the honest answer is: both kind of, and that’s exactly the problem.
Here’s what actually happened. Within the first few months of 2026, three significant studies landed fast, and they don’t agree. That’s not a scandal, that’s science. But it matters how you read them, because one involves mice and one involves humans, and those are very different things.
The Human Data Is Genuinely Encouraging, With Real Limits
Start here. On April 21, 2026, the Journal of the Endocrine Society published findings that three months of a ketogenic diet improved beta-cell function in people with type 2 diabetes. The measurement that matters is the proinsulin-to-C-peptide ratio, which reflects how hard your pancreas is working to produce usable insulin. Lower ratio means less pancreatic stress. The keto group moved the right direction.
This matters because beta-cell burnout is part of what makes type 2 diabetes get worse over time. If diet can reduce that strain, that’s clinically meaningful. I’ve seen clients improve their fasting glucose and cut medication on low-carb diets, and this study offers one plausible reason why.
But three months is three months. The study doesn’t tell you what happens at year two or year five, and that’s where the next research gets uncomfortable.
The Mouse Data Is Alarming, and You Shouldn’t Dismiss It
The University of Utah Health study, published in Science Advances and widely circulating since January 2026, is the one that should grab anyone doing long-term keto. Researchers fed mice a ketogenic diet for an extended period and found hyperlipidemia, liver dysfunction, and impaired insulin secretion. Male mice showed the most severe liver damage. The lead researcher didn’t hedge: the results “suggest that more studies are warranted in humans to establish the cardiometabolic benefits and risks of prolonged [keto diet] feeding.”
Yes, it’s mice. Rodent metabolism isn’t human metabolism. But mouse models have predicted real human metabolic outcomes often enough that dismissing them entirely is intellectually dishonest. What’s worth paying attention to is when the damage showed up. It was especially pronounced after carbohydrates were reintroduced. That speaks directly to keto cycling, a practice that’s become popular in communities as a supposedly safer approach.
The implication isn’t settled. It’s a serious hypothesis. Repeated metabolic shifts, particularly the carb reintroduction phase, may create stress that neither a stable keto diet nor a standard diet would produce alone. If you’re cycling specifically because you think it’s safer than sustained keto, this study gives you reason to question that assumption.
The Virginia Tech Study Adds a Third, Genuinely Useful Data Point
Nutritionist Shares Meal Plan For Doing Keto Over The Long Term · TODAY on YouTube
In February 2026, researchers at Virginia Tech published findings in Nature Communications showing that a ketogenic diet lowered blood sugar and restored exercise adaptation in hyperglycemic mice. This one got less attention than the Utah study, but it’s relevant because it points to a specific population where keto’s metabolic effects might be most protective: people whose blood sugar is already elevated enough to blunt normal exercise response.
The exercise adaptation piece is interesting. Hyperglycemia interferes with how muscles respond to training, and this study suggests keto may partially restore that response. For someone with prediabetes or poorly controlled type 2 who’s also trying to use exercise as a tool, that’s a meaningful signal. Again in mice, again needing human replication.
Taken together, all three studies paint a picture that experienced clinicians have been describing for years: keto offers real short-term metabolic benefits, the long-term picture is murky, and the effects depend heavily on your baseline health.
Why This Conflict Keeps Happening
| Study | Publication | Date | Model | Key Finding | Duration |
|---|---|---|---|---|---|
| Journal of the Endocrine Society | Journal of the Endocrine Society | April 21, 2026 | Human (type 2 diabetes) | Improved beta-cell function; lower proinsulin-to-C-peptide ratio | 3 months |
| University of Utah Health | Science Advances | January 2026 | Mouse | Hyperlipidemia, liver dysfunction, impaired insulin secretion (especially males) | Extended period |
| Virginia Tech | Nature Communications | February 2026 | Mouse (hyperglycemic) | Lowered blood sugar, restored exercise adaptation | Not specified |
A JAMA paper titled “The Ketogenic Diet for Obesity and Diabetes: Enthusiasm Outpaces Evidence” made exactly this argument years ago, and it’s still being cited in fresh 2026 research because the underlying tension hasn’t resolved. We have a diet that produces measurable short-term improvements in blood sugar, insulin sensitivity, and sometimes weight, paired with a long-term evidence base that’s thin, mixed, and largely extrapolated from animal models or short human trials.
Part of the problem is that “ketogenic diet” isn’t one thing. Fat sources vary. Protein levels vary. Whether someone’s eating whole foods or processed keto products matters. Baseline metabolic health matters. Duration matters. A three-month trial in people with type 2 diabetes and a multi-month mouse study using standardized high-fat feed are measuring genuinely different things and calling them both “keto.”
I’ve seen people treat keto as a permanent state and others use it as a short reset. The research doesn’t clearly vindicate either approach for everyone.
What This Means If You’re Deciding Right Now
If you have type 2 diabetes and you’re considering a ketogenic diet under medical supervision, the April 2026 human data is legitimately encouraging for short-term beta-cell outcomes. That’s a real finding from real people. Talk to your doctor or a registered dietitian who knows your full picture, including your liver enzymes and lipid panel, before and during.
If you’ve been doing keto for more than a year, the Utah findings are a reason to get bloodwork done, not a reason to panic. Liver function tests and a full lipid panel are reasonable monitoring tools. If your numbers are trending the wrong direction, that’s information worth having.
If you’re doing keto cycling because it seems more sustainable, be honest with yourself about whether the cycling is evidence-based or just easier to stick to. There’s nothing wrong with easier to stick to, but the Utah study raises enough questions about carb reintroduction effects that it’s worth discussing with a clinician rather than assuming cycling is automatically lower-risk.
The reality is that we don’t have the long-term human randomized controlled trial data that would settle most of these questions. That study hasn’t been done at scale. Until it is, you’re making decisions with incomplete information, which is true of almost everything in nutrition science. Go in with realistic expectations, monitor your actual health markers, and don’t let either the enthusiasm or the alarm set your course. Let your labs do that.
Sources
- Keto diet may improve beta cell function in people with type 2 diabetes (April 21, 2026)
- Keto diet weight loss may come with a hidden cost (January 27, 2026)
- A long-term ketogenic diet causes hyperlipidemia, liver dysfunction, and glucose intolerance (2025, widely circulated January 2026)
- Keto diet could unlock the effects of exercise for people with high blood sugar (February 27, 2026)
Photo: RDNE Stock project via Pexels
This article is for general informational purposes only and does not constitute medical or dietary advice. Always consult a licensed healthcare provider or registered dietitian before making significant changes to your diet, especially if you have a medical condition.
