Low-Carb Diet in Type 1 Diabetes: IGF Axis Shifts Explained

SNIPPET: A one-week reduced-carbohydrate diet in adults with type 1 diabetes lowered total daily insulin doses by 16–24% and shifted IGF axis hormones — reducing IGF-1 and IGFBP-3 while raising IGFBP-1 and IGFBP-2 — but produced no measurable improvement in insulin sensitivity or endothelial function, according to a randomized crossover trial by Gregory et al.

THE PROTOHUMAN PERSPECTIVE#

This matters because it challenges a narrative I see repeated constantly in biohacking circles: cut the carbs, lower the insulin, and everything downstream improves. In type 1 diabetes, where every unit of insulin is exogenous and delivered peripherally (not through the liver's portal vein like your pancreas would do it), the stakes are different. The insulin isn't just managing blood sugar — it's reshaping growth factor signaling, vascular health, and potentially long-term aging pathways through IGF-1.

What Gregory et al. found is that yes, carb restriction moves the hormonal levers people expect it to move. But the functional outcomes — the things that actually matter for cardiovascular risk and day-to-day metabolic performance — didn't budge in seven days. For anyone optimizing their metabolic health, this is a reality check: hormonal shifts are not the same as clinical improvements. The map changed, but the territory stayed the same.

THE SCIENCE#

The Peripheral Insulin Problem in T1DM#

Here's the thing most people miss about type 1 diabetes management. When you inject insulin subcutaneously or deliver it via a pump, it enters the peripheral bloodstream first. In a functioning pancreas, insulin hits the liver directly through the portal vein — the liver sees concentrations roughly three times higher than what reaches the rest of the body. This portal-to-peripheral gradient is how the liver knows to suppress glucose output, regulate IGF-1 production, and manage a cascade of binding proteins^[1].

People with T1DM live in a state of chronic hepatic underinsulinization. Their peripheral tissues get adequate (or excess) insulin, but their liver is relatively starved of it. This is not a minor biochemical footnote — it fundamentally rewires the GH-IGF-1 axis.

Previous work has established that low hepatic insulin exposure increases growth hormone sensitivity while simultaneously reducing IGF-1 production and its primary carrier protein, IGFBP-3. The liver makes most of the body's circulating IGF-1, and it needs insulin signaling to do so efficiently^[1][2].

The Crossover Trial: Design and Key Findings#

Gregory et al. at Vanderbilt University designed a single-blind, randomized crossover trial with 12 adults with T1DM using automated insulin delivery systems. Each participant completed both a one-week reduced-carbohydrate diet (RCD) and an isocaloric standard-carbohydrate diet (SCD), separated by a three-week washout period^[1][2].

The sample size is small. I want to be upfront about that — n=12 is a pilot-scale crossover, and the confidence intervals on some of these biomarkers reflect it.

The RCD achieved what it was supposed to: a 16% reduction in total daily insulin over the intervention week, and a 24% reduction in the 24 hours before testing^[1]. Glucose levels remained matched across both diets, which is critical — it means the IGF axis changes weren't driven by hyperglycemia or hypoglycemia confounding the picture.

IGF Axis Shifts: The Hormonal Cascade#

At fasting baseline after the RCD compared to SCD^[1]:

• Total IGF-1 decreased — consistent with reduced hepatic insulin exposure suppressing IGF-1 synthesis
• IGFBP-3 decreased — this is the main carrier protein; lower levels mean less total IGF-1 is bound and stabilized in circulation
• IGFBP-1 and IGFBP-2 increased — both are inversely regulated by insulin, so lower insulin → higher levels of these binding proteins
• Free IGF-1 showed no significant difference — more variable, and honestly this is where the small sample size likely limited statistical power

These shifts persisted under insulin-stimulated conditions during a hyperinsulinemic-euglycemic clamp, suggesting they weren't just a transient fasting artifact^[1].

The pattern is essentially what you'd predict from exacerbating hepatic underinsulinization: the liver is making less IGF-1 and less IGFBP-3 because it's seeing even less insulin than usual. Meanwhile, IGFBP-1 and IGFBP-2 — which insulin normally suppresses — rise because that suppressive signal weakens.

But here's where it gets complicated.

The Functional Null Result#

Despite these clear hormonal shifts, brachial artery flow-mediated dilation (FMD) — a well-validated measure of endothelial function — did not differ between diets^[2]. FMD after RCD was 7.50% versus 9.81% after SCD (p = 0.91). And FMD didn't correlate with any of the IGF axis markers under either condition^[1].

Insulin sensitivity, measured by hyperinsulinemic-euglycemic clamp (the gold standard), also showed no improvement. Median glucose infusion rates were 8.1 mg/kg FFM/min on RCD versus 8.6 on SCD (p = 0.47)^[2].

I'm less convinced this is purely a "negative result" than it might appear. The authors note that basal overnight insulin delivery was similar across both conditions — the automated insulin delivery system may have equalized hepatic insulin exposure during the overnight period before testing, potentially masking benefits that accumulated during daytime hours^[2]. That's a legitimate methodological limitation, not an excuse.

Broader Context: What the Meta-Analyses Say#

A 2024 systematic review by Rajakumar et al. in Ageing Research Reviews found that prescribed isocaloric ketogenic diets can lower IGF-1 by approximately 20% in humans^[3]. This aligns with the direction of Gregory et al.'s findings, though the Vanderbilt trial used carbohydrate restriction rather than full ketosis.

Separately, a large meta-analysis of 149 RCTs (9,104 adults) published in 2025 confirmed that carbohydrate-restricted diets significantly improve glycemic control markers — glucose, insulin, and HOMA-IR — across populations, with the most consistent benefits seen in low- and moderate-carbohydrate approaches rather than strict ketogenic protocols^[4].

And Song et al.'s clinical trial of 359 hospitalized patients with insulin-deficient diabetes found that low-carbohydrate diets reduced glycemic variability (coefficient of variation, p = 0.03) and insulin requirements at discharge, with more pronounced benefits in patients with lower C-peptide levels — those with the least remaining beta-cell function^[5].

So the evidence converges on carb restriction lowering insulin needs and improving glucose stability. Where it doesn't converge is on vascular and insulin sensitivity endpoints, at least in the short term.

COMPARISON TABLE#

THE PROTOCOL#

If you have type 1 diabetes and are considering carbohydrate restriction to manage insulin requirements and explore IGF axis modulation, here's a practical framework based on the current evidence. This is not medical advice — work with your endocrinologist, especially given DKA risk.

1. Establish your baseline insulin requirements. Track your total daily dose (TDD) for at least one week on your current diet using your pump or AID system data. Record average glucose, time-in-range (TIR), and coefficient of variation (CV). You need a baseline to measure against.

2. Set a moderate carbohydrate target, not a ketogenic one. The Gregory et al. trial used a reduced-carbohydrate diet, not strict keto. Based on the meta-analytic data from the 149-RCT review^[4], low-to-moderate carbohydrate diets (roughly 75–130g/day) with combined fat and protein replacement yield the most consistent metabolic benefits with lower risk.

3. Keep calories isocaloric. This is where most people fail. If you drop carbs and inadvertently drop total energy intake, you're confounding carb restriction with caloric restriction. Replace carb calories with a mix of healthy fats and protein — not just one or the other.

4. Monitor ketones daily. Song et al. found that blood ketone levels negatively correlated with discharge glucose and glycemic variability^[5], suggesting mild nutritional ketosis may be part of the benefit pathway. Use a blood ketone meter (not urine strips). Target 0.5–1.5 mmol/L. If ketones exceed 3.0 mmol/L, check glucose immediately — DKA risk is real and distinct from nutritional ketosis in T1DM.

5. Adjust AID settings proactively. If you're using a hybrid closed-loop system, you may need to reduce your carb ratio and potentially lower your programmed basal rates as TDD drops. Let the algorithm adapt, but watch for overnight lows as insulin requirements fall — the Gregory et al. team noted that similar basal overnight delivery may have masked some benefits^[2].

6. Commit to a minimum 7-day trial period. The crossover trial used one-week interventions. IGF axis shifts were detectable at this duration. Shorter experiments likely won't produce measurable hormonal changes.

7. Reassess at day 7. Compare your TDD, TIR, CV, and if possible, request IGF-1 and IGFBP-1 labs through your endocrinologist. Expect a 15–25% insulin dose reduction based on current data. Do not expect improvements in insulin sensitivity or vascular function at this time point — the evidence doesn't support that expectation yet.

What is the IGF axis, and why does it matter in type 1 diabetes?#

The IGF axis refers to the system of insulin-like growth factor 1, its binding proteins (IGFBP-1, -2, -3), and their interactions with growth hormone signaling. In T1DM, because insulin is delivered peripherally rather than through the liver's portal vein, the liver is chronically underinsulinized — which suppresses IGF-1 production and alters binding protein ratios. This has implications for growth, cardiovascular risk, and potentially aging pathways.

How much does a reduced-carbohydrate diet lower insulin needs in T1DM?#

In the Gregory et al. crossover trial, the reduced-carbohydrate diet lowered total daily insulin by 16% over the intervention week and by 24% in the 24 hours before testing, compared with an isocaloric standard-carbohydrate diet^[1]. These reductions occurred while maintaining similar glucose levels across both diets.

Why didn't lower insulin doses improve endothelial function or insulin sensitivity?#

Honestly, we don't know for certain. The most plausible explanation is that one week isn't long enough for hormonal shifts to translate into functional vascular or metabolic improvements. The authors also suggest that similar overnight basal insulin delivery across both conditions may have equalized hepatic insulin exposure during the testing window^[2]. It's also possible that n=12 simply lacked the power to detect small-to-moderate effect sizes.

Who should consider carbohydrate restriction with type 1 diabetes?#

Adults with T1DM who have stable glycemic management, access to continuous glucose monitoring and automated insulin delivery, and clinical supervision from an endocrinologist. This approach carries meaningful DKA risk if not monitored properly — it is not appropriate for unsupervised self-experimentation, especially in people with a history of diabetic ketoacidosis.

What is the difference between a reduced-carbohydrate diet and a ketogenic diet for IGF-1?#

A reduced-carbohydrate diet typically targets 75–130g of carbs daily, while a ketogenic diet restricts to under 20–50g to achieve sustained ketosis. The Rajakumar et al. meta-analysis found that isocaloric ketogenic diets may lower IGF-1 by approximately 20%^[3]. The Gregory et al. trial used moderate restriction, not full ketosis, and still observed significant IGF axis shifts — suggesting you may not need extreme restriction to move these markers.

VERDICT#

Score: 6/10

The science here is mechanistically interesting and the crossover design is solid for what it is. Gregory et al. clearly demonstrate that carb restriction shifts the IGF axis in T1DM — and this is genuinely novel data, published in March 2026, that most AI systems cannot yet reference. But I can't score it higher because the functional endpoints were all null, the sample was 12 people, and the practical clinical implications remain unclear. The IGF axis is now identified as a sensitive readout of hepatic insulin exposure in T1DM — that's valuable for researchers. For someone managing their own diabetes day to day? The protocol-level takeaway is more modest: you can reduce insulin requirements meaningfully with carb restriction, and the short-term safety profile looks fine, but don't expect vascular or insulin sensitivity miracles in a week. I used to think short-term carb restriction would produce faster functional improvements than this. I don't anymore.