Protein powder isn’t a superfood, and it isn’t junk. Whether it’s a win depends on what it replaces. Here’s what the trials say about the jobs it’s sold for (muscle, gut, immunity, “detox”), how whey, pea and beef compare, and what to check on the tub, including the three I recommend and why.
The grades in this post: Strong means good randomised trials agree. Mixed means trials disagree or are small and short. Mechanistic means the biology makes sense but it hasn’t been shown to change outcomes in people. None means there’s no usable evidence in people.
It depends what it replaces
I’m asked about protein powder more than almost any other product. My answer is always the same question back: instead of what?
- Instead of skipping lunch at your desk: a win.
- Instead of a processed meat sandwich as your “quick protein”: probably a win.
- Instead of a “mass gainer” that’s mostly maltodextrin and sweeteners: a clear win.
- Instead of a proper meal you had time to eat: a loss. Food brings fibre, vitamins and minerals, fullness, and the chewing that tells your body it’s been fed.
A powder is protein in a convenient form. It earns its place when it closes the gap between what you eat and what you need, and not otherwise.
How much do you need?
For most adults I work with: 1.2–1.6 g of protein per kg of body weight a day. In an analysis of 49 trials, extra protein stopped adding muscle (with resistance training) at about 1.6 g/kg a day (Morton 2018); the 1.2 lower end is my working range. Older adults, people training hard and anyone recovering from illness or injury are where gaps most often appear.
Food first. Roughly 30 g of protein is about 95 g of chicken breast, 120 g of tinned tuna, 205 g of quark or 285 g of reduced-fat cottage cheese. If you can get there with food, do.
The jobs it’s sold for, graded
| Claim | What the trials show | Grade |
|---|---|---|
| Builds muscle | Yes, when it adds to your total protein and you train. In a 2026 network analysis of 36 trials, whey ranked first for lean mass, and plant proteins showed smaller effects, though the authors rated the certainty for plant proteins low to very low (Li 2026). Head-to-head trials found pea or a soy-pea blend roughly matched whey when training and total protein were enough (Babault 2015; Santini 2025) | Strong for adding protein; Mixed for which kind. Industry interest: three of the seven authors of the pea trial worked for Roquette, which makes the pea protein tested |
| Detox / raises glutathione | Two meta-analyses found no effect of whey or milk protein on glutathione, oxidative stress or liver enzymes (Farahmandpour 2025; Mohammadi 2026) | Strong for “doesn’t”. The biochemistry (cysteine → glutathione) is real; the outcome didn’t follow |
| Boosts immunity | Small trials on blood markers only; I found no trial showing fewer infections. A scoop of whey isolate contains far less lactoferrin than the lactoferrin trials used | Mechanistic surrogate markers |
| Heals the gut | No trial of whey, pea or beef with bloating or IBS as the main outcome. Microbiome changes in small studies were minor and mixed | None for healing; the real gut issue is tolerance (below) |
| Steadies blood sugar | Whey taken before a meal lowered the glucose rise after it in people with mild or well-controlled type 2 diabetes (Chiang 2022: 5 trials, 134 people) | Mixed small, short-term trials. Supports “a shake beats a biscuit”; not a diabetes treatment |
One more finding from the network analysis: the benefit of whey levelled off at about 30–35 g a day of supplement, an estimate the authors call exploratory (Li 2026).
Whey, pea or beef?
Whey (from milk): the most studied, and highest in leucine, the amino acid most linked to switching on muscle building. After exercise it raised essential amino acids in the blood about 44% more than a pea, rice and canola blend; but muscle building over the next four hours was the same with both (van der Heijden 2024, 10 people). Isolate has had most of the lactose filtered out; concentrate has more.
Pea: matched whey for muscle in head-to-head trials when the total was enough. In the largest (161 men), the differences between pea, whey and placebo were only a trend overall; pea beat placebo in the weakest starters, and whey didn’t differ from either (Babault 2015). It’s lower in leucine, so you may need a bigger serving. Good for dairy-free and vegan eaters.
Beef: only two small trials, both roughly equal to whey (Naclerio 2017, 24 men; Sharp 2018, 41 people). But “beef protein” isn’t steak in a tub. The one I recommend is made from bones and cartilage, so it behaves like collagen: very high in glycine, low in leucine and tryptophan. That makes it useful for the collagen job (see my injury article on gelatin before rehab loading), but a poor main muscle-building protein unless you combine it with another source. “Complete protein” on the label only means all nine essential amino acids are present, not that they’re there in useful amounts.
Grass-fed: doesn’t make a different protein. On an isolate with almost no fat, the fatty-acid argument disappears.
Tolerance: the gut issues that are real
- Lactose: a whey isolate carries very little; a concentrate more.
- Milk-protein allergy: pea or beef instead.
- Sweeteners and additives: check the label; some people react to particular sweeteners or bulking agents.
- Salt: some plant powders add a surprising amount. One I recommend has about 0.6 g of salt a serving.
What’s actually in the tub: contamination and testing
- Lead. Consumer Reports tested 23 protein powders and shakes in the US. More than two-thirds had more lead in a single serving than its level of concern, 0.5 µg a day (the very strict limit used under California’s Proposition 65), and plant-based powders averaged about nine times more lead than dairy ones (Consumer Reports, October 2025). For comparison, the US Food and Drug Administration’s interim reference levels are 2.2 µg a day for children and 8.8 µg a day for women of childbearing age; these are targets for industry, not limits. I give both: the strict figure isn’t a safety threshold on its own, but lower is better, and the plant-higher pattern is worth knowing.
- Protein spiking. Some brands have been caught counting cheap amino acids like glycine as “protein”, because the standard test measures nitrogen. The check: a published amino acid profile and leucine per serving.
- “Third-party tested” for what? Informed Sport and Informed Choice test for banned substances in sport. They don’t test heavy metals or whether the protein figure is right. Ask what was tested, by whom, and against what limit.
- Country of origin. Contamination scandals have happened in China (melamine in 2007–08) and fraud cases in the US. The variable that matters is testing and traceability, not the country.
The three I recommend, and why
| Protein per serving | Leucine per serving | Best for | Watch | Testing published | |
|---|---|---|---|---|---|
| Pink Sun whey isolate On Amazon (paid link) | 23.2 g per 25 g | About 2.5 g (10.6% of protein, their certificate) | General use, muscle, older adults | Contains milk | Eurofins, one batch (6 Oct 2025): lead, cadmium and mercury below detection; arsenic 0.012 mg/kg |
| Nuzest Clean Lean (pea) On Amazon (paid link) | 19 g per 25 g (varies by flavour) | 1.7 g (stated) | Dairy-free, vegan | Salt about 0.6 g a serving | Testing claimed; results not published |
| Hunter & Gather beef On Amazon (paid link) | 20 g | About 0.8 g (from their amino acid profile) | The collagen job, alongside another protein | Low leucine and tryptophan; behaves like collagen | Eurofins heavy metals, one batch (unflavoured): lead, cadmium and mercury below detection; arsenic 0.006 mg/kg |
As an Amazon Associate I earn from qualifying purchases.
Credit where due: Hunter & Gather answered every question I sent, disclosed that they add glycine as a sweetener, and label their protein from the raw-material specification rather than from a nitrogen test that the added glycine would inflate. That’s the honest choice.
Other brands are available. A supermarket whey isolate with a published amino acid profile does the same basic job. What matters is the checks below, not the brand.
Checks on any tub
- Do you need it? Work out your daily target first.
- Amino acid profile published, with leucine per serving.
- Heavy-metal results published, with the lab and the limit used.
- “Tested” for what?
- Ingredients: sweeteners, bulking agents, salt.
- Kidneys: if you’ll be eating above about 1.6 g/kg a day, or you have kidney disease, check your kidney function first.
How strong is the evidence?
| Claim | Grade | Main source |
|---|---|---|
| Extra protein adds muscle with training, up to about 1.6 g/kg a day | Strong | Morton 2018 (49 trials) |
| Whey beats plant protein for muscle | Mixed | Li 2026 vs Babault 2015; Santini 2025 |
| Whey or milk protein raises glutathione or “detoxes” | Strong doesn’t | Farahmandpour 2025; Mohammadi 2026 |
| Protein powder boosts immunity | Mechanistic | Surrogate markers only |
| Protein powder heals the gut | None | No trial with gut symptoms as the outcome |
| Whey before a meal lowers the glucose rise in type 2 diabetes | Mixed | Chiang 2022 (5 small trials) |
| Beef protein works like whey for muscle | Mixed | Naclerio 2017; Sharp 2018 (two small trials) |
Sources
- Babault N et al. Pea proteins oral supplementation promotes muscle thickness gains during resistance training: a double-blind, randomized, placebo-controlled clinical trial vs. whey protein. J Int Soc Sports Nutr 2015;12:3. PMID 25628520. doi:10.1186/s12970-014-0064-5
- Chiang SW et al. Whey protein supplementation improves postprandial glycemia in persons with type 2 diabetes mellitus: a systematic review and meta-analysis of randomized controlled trials. Nutr Res 2022;104:44–54. PMID 35605541. doi:10.1016/j.nutres.2022.04.002
- Consumer Reports. Protein powders and shakes contain high levels of lead. October 2025. consumerreports.org
- Farahmandpour F et al. Whey protein intervention and inflammatory factors and oxidative stress: systematic review and meta-analysis of randomized controlled trials. Nutr Rev 2025;83:609–21. PMID 39196774. doi:10.1093/nutrit/nuae100
- Li Y et al. Dose-structure-population interactions of protein supplementation combined with resistance training on body composition: a Bayesian network meta-analysis with dose-response modeling. Front Nutr 2026;13:1860234. PMID 42625626. doi:10.3389/fnut.2026.1860234
- Mohammadi S et al. Impacts of milk protein supplementation on lipid profile, blood pressure, oxidative stress, and liver enzymes: a systematic review and meta-analysis. Nutr Rev 2026;84:706–33. PMID 40471664. doi:10.1093/nutrit/nuaf068
- Morton RW et al. A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength in healthy adults. Br J Sports Med 2018;52:376–84. PMID 28698222. doi:10.1136/bjsports-2017-097608
- Naclerio F et al. Carbohydrates alone or mixing with beef or whey protein promote similar training outcomes in resistance training males: a double-blind, randomized controlled clinical trial. Int J Sport Nutr Exerc Metab 2017;27:408–20. PMID 28530448. doi:10.1123/ijsnem.2017-0003
- Santini MH et al. Similar effects between animal-based and plant-based protein blend as complementary dietary protein on muscle adaptations to resistance training: findings from a randomized clinical trial. J Int Soc Sports Nutr 2025;22:2568047. PMID 41059835. doi:10.1080/15502783.2025.2568047
- Sharp MH et al. The effects of beef, chicken, or whey protein after workout on body composition and muscle performance. J Strength Cond Res 2018;32:2233–42. PMID 28399016. doi:10.1519/JSC.0000000000001936
- van der Heijden I et al. Plant protein blend ingestion stimulates postexercise myofibrillar protein synthesis rates equivalently to whey in resistance-trained adults. Med Sci Sports Exerc 2024;56:1467–79. PMID 38537270. doi:10.1249/MSS.0000000000003432
- NPR (via OPB). A study found lead in popular protein powders. Here’s why you shouldn’t panic. 16 October 2025. Source for the FDA interim reference levels quoted. opb.org