What long‑term randomized trials exist comparing high‑protein versus high‑carbohydrate diets for weight maintenance beyond 12 months?
Executive summary
Long-term randomized trials directly comparing higher‑protein to higher‑carbohydrate diets beyond 12 months are limited but do exist, and collectively show small, sometimes transient advantages for higher protein on weight and fat‑mass maintenance while being constrained by adherence and small effect sizes [1] [2] [3]. Major multicentre trials — notably DIOGENES, PREVIEW and the two‑year macronutrient trial published in the New England Journal of Medicine — provide the best long‑term randomized evidence and also illustrate the methodological challenges that leave the question partially unresolved [4] [5] [1].
1. The big multicentre experiments: DIOGENES and PREVIEW mapped the terrain
The European DIOGENES trial randomized adults after an initial low‑energy phase and compared ad libitum diets differing modestly in protein content and glycemic index, reporting 12‑month results that supported a slight advantage of higher protein and lower glycemic index for weight‑loss maintenance, and which were influential enough to spawn consumer books and commercial interpretations [4]. PREVIEW was a large randomized program that tested a high‑protein, low‑GI approach against alternatives and reported three‑year outcomes showing suppression of hunger with the higher‑protein, low‑GI strategy but did not demonstrate clear superiority for preventing weight regain across all endpoints, underscoring mixed long‑term effects [5].
2. The two‑year, four‑arm NEJM trial: breadth but diluted signal
A rigorously conducted randomized trial in the New England Journal of Medicine followed 811 overweight adults for two years and compared diets with targeted percentages of fat, protein and carbohydrates, explicitly designed to detect whether emphasizing protein or carbohydrates mattered for long‑term weight change; the study found that differences among diet groups were small and that adherence waned after the first months, a feature that limited persistent nutritional contrasts and blurred any macronutrient‑specific effect [1].
3. Disease‑specific long follow‑ups and 12‑month RCTs with modest gains
Type‑2 diabetes and other subgroups have been the subject of longer randomized work: the DEWL trial and other diabetes trials extended follow‑up to two years and examined high‑protein versus high‑carbohydrate advice specifically in people with diabetes, yielding variable metabolic and weight outcomes but no definitive, large long‑term advantage of one macronutrient target over another [2]. Several single‑center randomized trials reached 12 months (for example McAuley and Griffin et al.) and reported modestly better maintenance of weight or fat mass with ~30% energy from protein compared with ~15% protein, but these were modest differences and not universal across all measures [6] [7].
4. Systematic reviews and meta‑analyses: small effects, big caveats
Meta‑analyses that restricted inclusion to trials with at least 12 months of follow‑up report that higher‑protein, lower‑carbohydrate advice produces a small long‑term benefit in weight and fat mass, with effect sizes that translate into only a few hundred grams to a few kilos and which are amplified when compliance is better or when protein differences at 12 months exceed ~5% of energy [2] [8]. Broader systematic reviews also find that while higher protein increases thermic effect and helps preserve lean mass in short‑term isocaloric trials, evidence beyond 12 months is heterogeneous and subject to adherence, dropout and differing comparator diets [9] [3].
5. Why long‑term randomized evidence remains inconclusive
Trialists repeatedly flag the same barriers: large initial differences in macronutrient intake shrink over time as participants revert toward habitual diets, provided‑food centers show stronger effects than advice‑only arms, and weight‑loss trials typically suffer diminished adherence after the first few months — all of which reduce the sustained contrast needed to declare a robust, long‑term protein advantage [1] [2]. In short, the existing randomized trials beyond 12 months suggest a small and context‑dependent benefit for higher protein on weight maintenance but cannot claim a consistent, large effect across populations and delivery models [2] [8].
6. Implications and hidden agendas
Industry and commercial translations of trial findings (books, products) have amplified selective positive results — notably from DIOGENES — which may overstate practical, population‑level benefits; trialists themselves caution that providing food or intensive support, not protein per se, often drives the observed maintenance advantages [4] [2]. The best summary from the randomized long‑term evidence is pragmatic: small benefits exist in some trials and settings, but adherence and study design, not simple macronutrient physiology alone, explain much of the long‑term outcome variation [3] [1].