A food-based, low-energy, low-carbohydrate diet for people with type 2 diabetes in primary care: A randomized controlled feasibility trial

A 12‑week nurse‑led, low‑energy, low‑carb food plan helped people with type 2 diabetes lose significant weight and improve blood sugar.

What this study found

This small, carefully run trial tested a simple idea: could practice nurses help people with type 2 diabetes follow a real‑food, low‑energy, low‑carbohydrate plan for 12 weeks—and would it work?

• People on the plan lost an average of 9.5 kg (20.9 lb), versus 2 kg (4.4 lb) with usual care; the difference between groups was 7.5 kg (16.5 lb).
• Average long‑term blood sugar (HbA1c) fell by 16.3 mmol/mol (~1.49%) with the plan, versus 0.7 mmol/mol (~0.06%) with usual care.
• Several measures of insulin resistance and sensitivity improved more in the intervention group, as did systolic blood pressure (by an adjusted −14.4 mmHg) and triglycerides (adjusted −0.58 mmol/L / −51 mg/dL).
• Many participants reduced medications: 7 stopped at least one diabetes drug and 7 stopped at least one blood‑pressure drug; there were no such changes in the control group.
• An exploratory measure of 10‑year cardiovascular risk (QRISK3) improved by −3.6% in the intervention group.

Importantly, the program was acceptable and doable: nurses delivered it with high fidelity, participants engaged and finished, and focus groups described it as positive and motivating.

What the diet actually was

For 8 weeks, participants aimed for 800–1000 kcal/day, <26% of energy from carbohydrates, and at least 60 g protein/day, using ordinary foods—not shakes or bars.

• Guidance emphasized excluding sugary and starchy foods (keeping dairy and limited fruit), strict portion control, and minimal added fats/oils.
• Suggested foods: fresh vegetables/salads, small portions of lean meat and fish.
• Weeks 9–12 focused on stabilizing weight by incrementally increasing portions or adding one high‑fiber carbohydrate serving, while keeping carbs lower overall.

Support included four short nurse appointments (baseline, weeks 2, 4, 8), a self‑help booklet with menus/recipes, and a brief GP check to adjust medications safely.

How strong is the evidence?

Study design and rigor

• Type of study: Randomized controlled feasibility trial (RCT), individually randomized, 2:1 (intervention:control), stratified by practice; not blinded.
• Participants: 33 enrolled (21 intervention, 12 control), average age 67 years, BMI 35.4 kg/m² (lbs/ft² not applicable), HbA1c 61 mmol/mol (~7.7%); 12 weeks total.
• Primary aim: Feasibility—can people be recruited; can nurses deliver with fidelity; will participants complete follow‑up? All criteria were met (near‑universal engagement and completion, and full delivery of core elements in recorded sessions).
• Secondary outcomes: Changes in weight, HbA1c, glucose/insulin/HOMA, blood pressure, lipids, liver tests, medications, diagnostic HbA1c category, and QRISK3 (exploratory).

Limitations (in everyday terms)

• Small and short: Only 33 people, just 12 weeks; results show promising short‑term effects but cannot prove long‑term remission or durability.
• Not blinded: Participants and clinicians knew the assignment; some in the control group reported eating lower carb, which could dilute differences.
• Feasibility focus: The trial was not powered to confirm efficacy; it shows what’s possible, not definitive clinical outcomes over time.
• Generalizability: Most participants were white British adults in UK primary care; broader populations need testing.
• Timing matters: HbA1c was measured soon after the energy‑restriction phase; effects at 12 months may differ.

Funding and interests to note

• Funding: NIHR Oxford Biomedical Research Centre and School for Primary Care Research (public research funders).
• Competing interests: Two authors have received research funding (no personal pay) from commercial weight‑loss companies; one author (Clare Bailey) writes recipe books and is part‑owner of thefast800.com. The paper states no company had interests in this specific program.
• Safety: No unexpected serious adverse events; one unrelated death in the control group.

What participants said (briefly)

Focus groups described the program as acceptable, motivating, and practical. People valued simple rules, recipes, shopping tips, label reading, and regular nurse support. Seeing quick improvements (glucose, blood pressure, weight) boosted confidence. Social support at home helped. Nurses found follow‑up contact motivating too.

Practical relevance for people with type 2 diabetes

Within ordinary primary care, a nurse‑supported, real‑food, low‑energy, low‑carb approach was feasible and linked to meaningful short‑term improvements in weight, blood sugar, and some cardiometabolic markers, with reduced medications in several cases—over just 12 weeks.

What this study does not claim:

• It does not claim long‑term remission, nor address specific macronutrient controversies beyond its described low‑carb, low‑energy framework.
• It does not assert that any single nutrient (like saturated fat) is beneficial or harmful; diet quality concerns (e.g., fiber) are noted as a future focus for longer trials.

Key numbers at a glance

• Weight change: −9.5 kg (−20.9 lb) vs −2.0 kg (−4.4 lb); adjusted difference −7.5 kg (−16.5 lb).
• HbA1c change: −16.3 mmol/mol (~−1.49%) vs −0.7 mmol/mol (~−0.06%); adjusted difference −15.7 mmol/mol (~−1.44%).
• Systolic BP: adjusted −14.4 mmHg.
• Triglycerides: adjusted −0.58 mmol/L (−51 mg/dL).
• QRISK3: −3.6% (exploratory).
• Medications: 7 reduced diabetes meds; 7 reduced BP meds (none in control).

Study details

• Setting: UK primary care across three practices.
• Duration: 12 weeks (8 weeks low‑energy, 4 weeks maintenance).
• Intervention delivery: Four nurse visits (15–20 minutes each) + one GP medication review.
• Diet: 800–1000 kcal/day, <26% energy from carbs, ≥60 g protein/day; vegetables/salads, small lean meat/fish portions; exclude sugary/starchy foods; minimal fats/oils; gradual energy increase to stabilize weight in weeks 9–12.
• Outcomes measured: Weight, BP, fasting labs (glucose/insulin/HOMA, lipids, liver tests), HbA1c, medications, QRISK3 (exploratory), and qualitative feedback.