Compare heavy-metal concentrations, serving sizes, detection limits and lot averages without turning a laboratory report into a safety guarantee.
A protein powder's heavy-metal result needs a unit, a matching serving size, and a clear description of the sample tested. A smaller concentration does not necessarily mean less metal per serving, and “not detected” does not mean proven zero. A published product average also cannot certify the tub you are holding.
This guide explains how to compare reports for adult protein-supplement purchases. It supplies no new product tests, brand ranking or personal safety threshold. Sources checked October 30, 2026.
Record the exact product, flavor, lot, test date and sample description. Ask whether the laboratory tested the finished powder as sold, a dried sample, a raw ingredient or a prepared drink. Those results have different denominators.
Find the actual result column. A specification such as “lead ≤1 mg/kg” states an acceptance criterion; it does not tell you what the laboratory measured. If the report gives only “pass,” request the result and the criterion used.
Keep total arsenic separate from inorganic arsenic. Consumer Reports performed speciation on a subset of its protein-supplement samples, and distinguishes untested results from nondetections. October 2025 methods and results
For a mass-based result in finished powder as sold:
Micrograms per serving = result in micrograms per gram × grams of powder per serving.
For mass fractions, 1 mg/kg equals 1 µg/g, commonly called 1 ppm. A result in µg/kg, commonly called ppb, must first be divided by 1,000 to express it in µg/g. Explicit mass units avoid ambiguity about what “parts” means. NIST guide to expressing quantities
Here are invented report numbers, not tested products or acceptable exposure levels:
B has the lower concentration but the higher amount per label serving. Neither calculation establishes that either product is safe.
Use powder weight, not the weight of milk or water added. Do not multiply a laboratory digestion-solution result in mg/L by powder grams: the laboratory must account for preparation and dilution. FDA's analytical manual distinguishes mass-based and volume-based calculations and relevant corrections. FDA EAM section 3.4
Different label servings may supply different amounts of protein. For a purchase comparison, calculate metal per gram of listed protein while keeping the original serving figures visible.
Suppose the invented A serving lists 25 g protein, while B lists 50 g. A's calculated lead amount is 2.8 ÷ 25 = 0.112 µg per gram of listed protein. B's is 3.5 ÷ 50 = 0.070 µg. For an equal 25 g of listed protein, those figures correspond to 2.8 and 1.75 µg respectively.
That changes the comparison question; it does not cancel the amount in a serving you actually consume. It assumes the label protein figures are accurate and ignores measurement uncertainty. It also does not make different protein sources nutritionally equivalent. Our protein-label testing guide covers the separate question of what a protein assay measures, and our collagen-versus-whey guide examines source-specific muscle evidence.
A printed result of <0.02 mg/kg with a 35 g serving converts to a stated limit of <0.70 µg per serving, not an exact measured amount. For “ND,” ask what detection limit applies to the original sample. Without that limit, you cannot calculate a comparable bound. Detection and reliable quantitation are distinct in FDA's analytical terminology.
Consumer Reports tested two or three samples per product model, purchased between November 2024 and January 2025. Its published product values are averages. When a metal was detected in some samples, it substituted half the method detection limit for samples below that limit; when none had a detection, it substituted zero. These are calculation assumptions, not proof that every sample contained the average or that a nondetection contained no metal. The original report explains them.
Request individual results, reporting limits and uncertainty before comparing close averages. These historical averages do not establish the result for later lots.
FDA states that no safe level of lead exposure has been identified. Its assessment considers concentration, consumption and vulnerable populations; detecting lead does not by itself mean a food must be avoided. A reference value for a named population is not a personal allowance for one supplement. FDA lead guidance
Consumer Reports also explicitly says its exposure calculations do not establish California Proposition 65 compliance. Keep the report's assumptions and benchmark separate from a legal finding or a prediction that a particular person will be harmed.
Ask the manufacturer:
Please provide the finished-product report for this flavor and lot, including the actual metal results, mass basis, individual sample results, detection and quantitation limits, and uncertainty information. Explain any averaging or substituted nondetection values, and identify the benchmark behind a pass claim.
If you are concerned about your own exposure, take the label and report to a clinician. Powder arithmetic cannot diagnose poisoning, determine a blood-testing plan or justify a detox product. For shopping, the useful next step is a comparable, sufficiently detailed report rather than an unsupported “cleanest” label.
Powered by BTST