Meal Replacement Evidence: Nutrition, Satiety and When It Helps

Meal replacement evidence splits in two. Clinic programmes with formula food and follow-up have trials. A lone desk shake does not inherit those results. This is a source-reviewed guide, not a hands-on lab test.

Adverts mix three jobs. One sitting replaces a sandwich. Two formula meals leave one cooked plate. Total diet replacement replaces almost all food for a set block. Meal replacement evidence is not the same for those three jobs.

This page is not a shopping list. For a complete-food product, see the Huel Black Edition review. For drink logging, stay with hydration apps and nutrition products.

What a meal replacement is, and what it is not

In plain language, a meal replacement is a product designed to stand in for a plate. Energy is fixed. Protein and micronutrients are listed. You still need water. You still need a plan for the other sittings.

NICE NG246 (chapter captured 13 September 2026) tells services to keep an energy deficit and to prefer a balanced long-term pattern, pointing at the NHS Eatwell Guide. It warns against nutritionally unbalanced restriction.

A flavoured shake that you add to an unchanged evening graze is not that model. It is a snack with a marketing line.

Total diet replacement: the clinic trials

DiRECT (Lean et al., Lancet 2018) was a cluster-randomised primary-care trial in Scotland and Tyneside. The intervention stopped diabetes and blood-pressure drugs, then used an 825–853 kcal formula diet for 3–5 months, then stepped food back in, then offered maintenance support. The intention-to-treat groups each had 149 people.

At 12 months, 36 intervention participants (24%) lost 15 kg or more. None in usual care did. Remission, defined as HbA1c under 48 mmol/mol (6.5%) after at least two months off diabetes medicines, occurred in 68 (46%) versus 6 (4%). Mean weight change was −10.0 kg versus −1.0 kg. Remission tracked weight loss. Two serious events in one person were judged possibly related (biliary colic and abdominal pain).

At 24 months (Lean et al., Lancet Diabetes & Endocrinology 2019), 53 of 149 (36%) in the intervention arm were still in remission, versus 5 of 149 (3%). Mean weight difference had narrowed to −5.4 kg. That is meal replacement evidence for a supported TDR pathway, not for an unsupervised tub.

DROPLET (Astbury et al., BMJ 2018) randomised 278 adults with obesity in Oxfordshire primary care. The TDR arm used 810 kcal a day as sole food for eight weeks, then food return, with weekly support for 12 weeks and monthly support for three months. Usual care was nurse support plus a modest energy cut. At 12 months, weight change was −10.7 kg versus −3.1 kg (adjusted difference −7.2 kg; 95% CI −9.4 to −4.9). Forty-five percent versus 15% lost at least 10%. Moderate-or-worse adverse events were similar (11% versus 12%). Seventy-three percent were re-measured at one year.

NICE 2025 used this literature. Low-energy diets (800–1,200 kcal) may be considered for adults living with obesity, or with overweight plus type 2 diabetes, only in specialist or long-term-condition services, with dietitian or nutritionist access, and for no more than 12 weeks. Very-low-energy diets (under 800 kcal) are narrower still. Neither is a long-term strategy on its own.

Partial replacements and satiety

Astbury, Piernas, Hartmann-Boyce, Lapworth, Aveyard and Jebb (Obesity Reviews 2019) asked a different question. They excluded total diet replacement and diets under 800 kcal. They kept 23 randomised trials and 7,884 adults. Searches ran to August 2018.

Mean year-one weight change favoured meal-replacement arms in every grouped comparison. Versus other diets, the difference was −1.44 kg (95% CI −2.48 to −0.39). With support on both sides, it was −2.22 kg. Versus unsupported comparators, −3.87 kg. With enhanced support versus regular support, −6.13 kg. Heterogeneity was high in some pools. Only six trials were low risk of bias on every Cochrane domain.

That is meal replacement evidence for structured partial swaps plus people. It is not a proof that any bar you like will hold hunger at 16:00.

Satiety research is messy. Protein and fibre often help. Liquid meals can be faster to finish than a plate. We did not find a single validated “fullness index” that ranks current consumer brands. If a shake leaves you raiding biscuits at 15:30, the product failed its job for you, whatever the year-one mean in a trial.

Protein still matters on food days. Use the everyday protein intake guide when the formula phase ends.

NHS services are not the supermarket aisle

NHS England describes a Type 2 Diabetes Path to Remission programme: 800–900 kcal total diet replacement for 12 weeks, then lifestyle support, for eligible adults with type 2 diabetes and overweight or obesity. Early service data have been published separately. Eligibility is clinical. Buying soups does not enrol you.

NICE also tells clinicians to discuss regain, medicine changes, and eating-disorder risk before a low-energy phase. Constipation, fatigue, and hair loss appear in the very-low-energy warning list. Those are reasons for supervision, not for a silent start.

What meal replacement evidence still leaves open

We did not attach a head-to-head RCT of every high-street shake versus cooked lunches in healthy office workers. Brand taste tests are not outcomes. Store ratings are omitted here because they are not trials.

Long-term weight after year two depends on eating skills. DiRECT’s second year already showed regain for many people. Apps such as WaterMinder or Waterllama can log fluid. They cannot watch for gallstones or low glucose.

Plant Nanny is a game-like reminder. It is not a dietitian.

How to use a product without borrowing a trial

If you want one sitting of convenience, read energy, protein, fibre, and free sugars. Sit down. Use a glass. Keep a cooked meal in the day so chewing does not vanish.

If you want a DiRECT-style phase, ask a clinician whether a service exists and whether your medicines need a change. Do not copy 850 kcal from a paper if you take insulin or you have a history of disordered eating.

Judge progress by function, not a single morning weight. Persistent faintness, chest pain, or mood collapse tied to restriction needs care, not another flavour.