Choosing a vegan protein powder: pea, rice, soy, and the heavy-metal question

Two scoring systems get cited on labels and in marketing, and they are not the same thing. PDCAAS (protein digestibility corrected amino acid score) caps at 1.0 and estimates digestibility using a rat model applied to the whole diet. DIAAS (digestible indispensable amino acid score) replaced it in most nutrition science after a 2013 FAO expert consultation, because it measures digestibility of each essential amino acid at the end of the small intestine rather than assuming one blanket correction. DIAAS does not cap at 1.0, so a very high-quality protein can score above it. A 2017 analysis of dairy and plant proteins by Mathai and colleagues, published in the British Journal of Nutrition, put whey protein isolate at roughly 1.0 to 1.09 DIAAS. A separate 2020 comparison by Herreman and colleagues in Food Science & Nutrition, using a stricter reference pattern, scored whey lower, around 0.85, still the top tier in that analysis. Most single-source plant proteins land lower, generally in the 0.6 to 0.9 range, because they run short on one essential amino acid the body cannot make.

That shortfall is the limiting amino acid. Legumes, including peas, lentils and soy, tend to run low on methionine and cysteine, the sulfur-containing amino acids. Cereal grains, including rice, wheat and corn, tend to run low on lysine. Neither gap is fatal on its own if the rest of the day's diet fills it in, but a powder that blends a legume with a grain closes the gap inside a single scoop. That is the actual mechanism behind pea and rice blends, not a marketing flourish: pea protein supplies the lysine rice lacks, and rice supplies the methionine pea lacks. Manufacturer assays of commercial pea and rice blends, typically mixed around 70 percent pea to 30 percent rice, are not independently published, but the blend reliably scores higher on DIAAS than either source tested alone, closing most of the practical gap with whey for most buyers.

The honest comparison to whey does not end at the score. Whey has more leucine per gram than plant proteins, and leucine is the amino acid that triggers muscle protein synthesis most directly. A 2023 trial by West and colleagues found that a mycoprotein and pea protein blend produced comparable post-exercise muscle protein synthesis rates to expectations for animal protein in resistance-trained adults, but several rodent and cell studies, including one using a pea-rice mix matched to whey for both total protein and leucine content, still found a smaller rise in blood leucine and a dampened mTORC1 signaling response with the plant protein. Where it counts more, in humans doing actual resistance training over weeks rather than a single post-workout blood draw, a randomized 12-week trial in 161 young men found pea protein produced a significantly greater increase in muscle thickness than placebo at 50 grams per day, while whey did not reach significance against placebo and the overall difference between groups only trended toward significance. A separate trial on muscle damage after eccentric exercise found whey outperformed pea on recovery biomarkers, with pea landing between whey and no supplement at all but not statistically different from either. Read together, the score gap between whey and a good plant blend is real on paper and mostly disappears in outcomes once total protein intake is matched. Someone asking whether vegan protein is better than whey is asking the wrong question: it is not better, it is close enough when the blend is chosen well and the total protein and training are equal.

Plants absorb cadmium and lead from the soil they grow in, and they cannot filter it out the way an animal's liver and kidneys partially do before the protein reaches milk. That is the mechanism, not a manufacturing defect: a plant protein isolate concentrates whatever was in the crop, and the level tracks the specific crop and where it was grown far more than it tracks the brand. Cocoa is a known accumulator of cadmium, which is why chocolate-flavored powders of any protein type test higher than vanilla. Rice has its own separate problem: it is grown flooded, and that anaerobic soil chemistry pulls inorganic arsenic, the more dangerous form, up through the roots. Brown rice concentrates more of it than white rice because the outer hull, removed to make white rice, is where arsenic accumulates most, roughly 80 percent more than in the milled grain according to testing cited by Consumer Reports. Both mechanisms mean rice protein and cocoa-flavored plant powders deserve more scrutiny than a plain pea or soy isolate.

The Clean Label Project, a nonprofit testing group, published a report in January 2025 after contracting an independent lab, Ellipse Analytics, to run 35,862 tests for lead, cadmium, arsenic, mercury and bisphenols across 160 top-selling protein powders from 70 brands, using ICP-MS for the metals. It found 47 percent of products exceeded at least one federal or state safety threshold for heavy metals, with 21 percent measuring at least twice California's Proposition 65 lead limit, that organic products averaged three times the lead and twice the cadmium of non-organic ones, and that plant-based powders averaged roughly five times the cadmium of whey. Industry groups pushed back hard: the Council for Responsible Nutrition and the Natural Products Association both argued the report conflates detectable trace levels, which modern ICP-MS can find at parts-per-billion, with actual health risk, and criticized Clean Label Project for not fully disclosing product selection methodology when asked. Both things can be true at once: the underlying chemistry, that plant crops absorb more cadmium than dairy does, is well established and not in dispute, while a single advocacy report's specific percentage should be read as one testing snapshot from one lab in one year, not a permanent ranking of brands.

A Proposition 65 warning label on a product sold in California means a lab detected a chemical above California's daily exposure threshold, which for lead is 0.5 micrograms per day, far stricter than the FDA's own interim reference level of 12.5 micrograms per day for adults. It does not mean the product violates federal safety limits, and its absence does not mean a product is free of the same chemicals, since Prop 65 only applies to products sold in California and enforcement depends on someone testing and suing. Treat the label as a starting point for asking questions, not as a verdict either way.

Protein per scoop is a marketing number; protein per 100 grams is the number that lets you compare products with different scoop sizes honestly. A 40-gram scoop advertising 30 grams of protein and a 30-gram scoop advertising 24 grams of protein are close to identical by weight, but only the per-100g figure shows that at a glance.

Amino acid spiking, also called nitrogen spiking, exploits how protein content gets measured on a label in the first place. The standard Kjeldahl method measures total nitrogen and multiplies by 6.25 to estimate protein, because protein is the main nitrogen source in food. Free amino acids like glycine and taurine also contain nitrogen and cost a fraction of what whey or pea isolate costs, so a manufacturer can add a spoonful and inflate the nitrogen reading without adding real protein. This practice drew class action lawsuits against several US brands in the 2010s. A label listing individual free amino acids, especially glycine or taurine, high on the ingredient list next to a protein claim that looks too generous for the price is the tell.

Added sugar and gums are the other place plant powders differ from a simple whey isolate. Rice and hemp proteins in particular need more added sweetener and gum to be palatable, since neither mixes as smoothly on its own, so check added sugar per serving and look for xanthan gum or sunflower lecithin low on the list rather than high.

Third-party certification is the only verification a consumer actually has, since neither the FDA nor the USDA routinely tests supplement protein content or heavy metal levels before products reach shelves. NSF Certified for Sport and Informed Sport both test finished batches for banned substances and verify the label matches what is in the container; NSF Certified for Sport also enforces heavy metal limits under the NSF/ANSI 173 standard it is built on, and both programs require the kind of third-party lab access that makes a brand's other safety claims more credible. A certification seal is not proof of purity, but its absence on a plant protein sold at a premium is worth noticing.

Most people typing this question do not need a powder. A cup of cooked lentils has about 18 grams of protein for close to nothing per serving, a block of tofu runs 15 to 20 grams per half cup depending on firmness, tempeh runs higher still, and seitan, being nearly pure wheat gluten, delivers 20 grams or more per 3.5 ounces. All four are cheaper per gram of protein than any powder on the market and none require a lab test for heavy metals, because whole foods are eaten in modest daily amounts rather than concentrated into an isolate. A powder earns its place for a specific reason: convenience right after a workout when cooking is not realistic, a calorie target that is easier to hit with a low-volume shake, or an appetite that struggles to eat enough whole food in a day. If none of those apply, the honest answer is to skip the powder and spend the money on food.

For a full pantry list of whole-food protein sources and how they compare, see the protein staples guide. For whether a vegan diet needs any supplement at all beyond food, see which supplements are actually needed. For the myth that plant proteins must be combined at the same meal to count, see the protein combining myth. For what the science says about vegan athletes specifically, see vegan athletes.