Covers allergen hazards in gelatin, collagen peptides, plant peptides, and hyaluronic acid, including protein and DNA methods, regulatory thresholds, VITAL doses, LOQ, and action limits.
For functional ingredients such as edible gelatin, pharma gelatin, collagen peptides, plant protein peptides, and hyaluronic acid, an allergen is any protein-containing material recognized as a priority food allergen by the applicable regulatory framework. The Codex priority list, the US FALCPA/FASTER Act framework, and EU Annex II define the major allergen groups that testing programs typically address; the exact list applied to a given lot depends on its destination market and on the certification scheme governing the finished product.
Within a functional ingredient lot, three hazard categories must be separated before any test is designed. An intentional ingredient is an allergen formulated into the product by design. Carry-over is residual allergen transferred from a preceding production run through shared equipment or rework. Cross-contact is unintended incorporation during handling, packaging, storage, or transport, often at trace level. Testing scope must focus on the latter two when the ingredient is declared allergen-free, and on identity and quantity when an allergen is intentional.
Analytical target selection follows hazard chemistry. Protein-target methods detect allergenic proteins or peptides and align directly with immunological risk; DNA-target methods detect species-specific sequences and serve as an indicator of source material, not of intact allergenic protein. This distinction matters for hydrolyzed collagen peptides and heat-processed gelatin, where protein fragmentation and DNA degradation may diverge. Procurement and quality teams should define, per material, which allergens are plausible from raw material origin, shared-line history, and co-packed items, then map each to a protein or DNA target before accepting any analytical plan.
Allergen limits used in functional ingredient release combine declarative rules, quantitative action levels, and scheme-specific reference doses. They are not interchangeable, and a compliant result depends on which framework applies to the finished product rather than to the ingredient alone.
Declarative regulation requires labeled disclosure of listed allergens when they are intentionally present, generally regardless of concentration. Trace, unintentional presence is governed separately through advisory labeling and risk-based action levels. The Voluntary Incidental Trace Allergen Labelling (VITAL) program, currently at version 3.0, provides reference doses expressed as milligrams of total protein from the allergenic source; these doses are converted into an action concentration for a given serving size and are used to decide whether precautionary labeling is warranted. VITAL reference doses are a risk-management tool, not universal statutory limits.
Gluten is governed by a distinct quantitative threshold. The Codex standard for gluten-free foods (CXS 118-1979, revised) sets 20 mg/kg gluten in food as presented to the consumer; many jurisdictions adopt this value in their gluten-free rules. Because gelatin and collagen peptides are inherently non-cereal materials, gluten findings usually indicate cross-contact from shared lines, adjunct ingredients, or packaging environments rather than source composition.
For peanut, milk, and egg, private certification schemes and manufacturer specifications commonly set action levels in the low parts-per-million range expressed as protein equivalents, with scheme-defined values often cited between approximately 2.5 and 10 ppm protein. These numbers are valid only under the issuing scheme; they must not be generalized into legal limits or transferred across allergens.
Two analytical distinctions control interpretation. The limit of quantification (LOQ) is the lowest concentration the laboratory can report with defined accuracy; the action limit is the decision threshold tied to risk or specification. A result below the action limit but above the LOQ is still a quantified positive. Results must also be read in the analyte stated by the method: ppm of gluten, ppm of total milk protein, ppm of casein, and ppm of skim-milk powder are not equivalent values. Quality teams should require certificates and reports to state the analyte, the action limit, the LOQ, and the scheme that defines each, so that batch-release and labeling decisions use a single, traceable basis.
No single allergen method fits every functional ingredient matrix. The choice must follow the analyte that survives processing, the chemistry of the ingredient, and whether the decision requires screening or quantified release data. The comparison below summarizes target, matrix fit, output, and principal limitations.
| Method | Target analyte | Matrix fit | Output | Principal limitation |
|---|---|---|---|---|
| Sandwich ELISA | Intact soluble allergen protein | Unprocessed or mildly processed powders and extracts | Quantitative | Under-recovery after heat, hydrolysis, or denaturation; hook effect at high load |
| Competitive ELISA | Small or denatured protein epitopes | Hydrolyzed collagen peptides and processed gelatin | Quantitative | Narrower working range and sensitivity to extraction quality |
| Real-time PCR | Species-specific DNA sequences | Source verification where protein is degraded | Quantitative or semi-quantitative | DNA does not measure allergenic protein; fragments may be lost |
| Lateral flow device | Selected allergen protein | Rapid cleaning checks and field screening | Qualitative | Not a certified quantitative release method |
| LC-MS/MS | Allergen-specific peptide markers | Hydrolyzed, fermented, or ELISA-inhibited matrices | Quantitative | Higher method complexity and need for matrix-specific markers |
Polyphenols and tannins can bind protein or assay antibodies; hydrolysates and severe heat can destroy epitopes and fragment DNA. Method selection must therefore be matrix- and allergen-specific, with orthogonal confirmation when the primary signal is unreliable.
A method is acceptable for a functional ingredient only when its performance has been demonstrated in that ingredient, or in a matrix demonstrably equivalent in composition, processing history, and extraction behavior. Kit-manufacturer validation performed in a different commodity does not establish fitness for gelatin, collagen peptides, plant protein peptides, or hyaluronic acid.
The core performance parameters are as follows.
Spike levels must bracket the decision point rather than only demonstrate high-concentration detection. For processed ingredients, incurred reference material—allergen carried through the actual heat, hydrolysis, or fermentation step—is preferable to a post-processing spike, because denaturation and fragmentation alter extractability. Where incurred material is unavailable, recovery should be assessed at the most relevant process stage. Validation records should state the matrix, allergen, analyte (protein, peptide, or DNA), equipment, acceptance criteria, and the action limit they support. Procurement decisions should treat a certificate citing generic kit validation as incomplete until matrix-specific data are provided; this is the parameter most often responsible for apparently compliant lots failing finished-product verification.
Allergen contamination in powdered, granulated, or particulate functional ingredients is rarely uniform. A single grab sample can miss localized carry-over from a shared line, a contaminated packaging seam, or a small rework addition, producing a non-representative negative. The sampling plan must therefore be designed around lot heterogeneity before analytical sensitivity is considered.
A representative plan defines the number of incremental samples based on lot size, packaging configuration, and known heterogeneity risk; increments are drawn from multiple positions across bags, totes, or bulk containers and from different depths. Two strategies follow. A composite strategy combines and homogenizes increments to estimate average lot contamination and is appropriate where the decision is lot-level release against an action limit. An individual-sample strategy retains increments separately to detect localized hot spots and is preferred after changeovers, supplier changes, or cleaning failures, where averaging could dilute a real signal. Minimum sample mass must be large enough to capture rare particles; subsampling for analysis occurs only after adequate homogenization.
Homogenization controls the analytical particle size. Coarse granules and flakes must be ground or milled to a consistent fineness so that the test aliquot, often only a gram or less, reflects the bulk laboratory sample. Grinding itself is a contamination risk: equipment must be dedicated or validated as clean between allergen and non-allergen materials, and high-risk samples should be processed after a documented cleaning or on reserved equipment.
Extraction conditions are part of sample preparation, not a laboratory formality. Buffer composition, extraction temperature, time, and mixing must match the validated method and the matrix; collagen and gelatin solutions can gel or vary in viscosity, plant peptides can bind polyphenols, and hyaluronic acid solutions can alter phase behavior. Deviations change recovery. Quality teams should specify sampling frequency, increment count, composite or individual strategy, minimum mass, homogenization target, and cross-contamination controls in the sampling protocol, and should reject results where the chain from lot increment to test aliquot is undocumented.
Allergen testing frequency is driven by the probability that a given production event introduces or redistributes allergen, not by a fixed calendar schedule. For functional ingredient lines, the manufacturing variables below define where testing intensity must increase.
Shared equipment is the dominant structural risk. A line that alternates between allergen-containing and allergen-free products carries residual hazard from conveyors, dryers, mills, sieves, transfer piping, and filling heads. Risk rises with difficult-to-clean geometry, adhesive materials, and dry processing, where proteinaceous dust can persist in dead spaces. Production sequencing determines exposure: running an allergen-free product immediately after an allergen-containing campaign creates the highest changeover risk, while dedicated campaigns separated by validated cleaning reduce it.
Cleaning method changes the verification burden. Wet cleaning with detergent and rinse generally removes soluble protein more effectively than dry cleaning, but incomplete drying can create microbiological and residue redistribution issues. Dry cleaning (vacuuming, wiping, air purging) is common in powder handling but may leave embedded residues; it requires more frequent visual, swab, and rinse verification. Cleaning must be validated for the specific allergen, soil load, and equipment, and verification should combine visual inspection with analytical testing of swabs, rinse water, or first-off product.
Rework control is a frequent root cause. Rework containing an allergen must be segregated and routed only into compatible product; uncontrolled rework converts a localized issue into lot-wide contamination. Supplier or origin changes alter the raw-material hazard profile: a new collagen, gelatin, plant peptide, or adjunct supplier may use different equipment, adjuncts, or transport, warranting increased incoming testing until a history is established. Seasonal campaigns, toll manufacturing, and shared storage introduce similar step changes.
Testing frequency should therefore rise at allergen changeover points, after cleaning deviations, following engineering maintenance that opens equipment, upon supplier or formulation changes, and during high-risk campaigns. Routine monitoring can be reduced only after a documented period of compliant changeovers. This linkage lets quality teams allocate analytical spend to the events that actually change risk rather than testing stable periods uniformly.
Allergen results should be interpreted against a pre-defined decision flow that links the reported value to the LOQ, the action limit, measurement uncertainty, and labeling obligations. Decisions made after the result is known tend to be inconsistent and difficult to defend.
A non-detected result means the analyte was not observed above the method LOD in the tested sample; it is not proof that the lot is allergen-free, because sampling, extraction, and method scope all limit coverage. Such a result supports release when the method is validated for the matrix, the sampling plan represents the lot, and no contextual trigger (recent changeover, cleaning failure, supplier change) contradicts it.
A quantified result above the action limit indicates that the lot does not meet the allergen-free specification under the governing scheme. The default action is hold, followed by root-cause investigation and a disposition decision that may include relabeling with the allergen declared, diversion to a compatible product, or rejection. Automatic rejection is not always the only outcome, but release against an allergen-free claim is not supported.
A result between the LOQ and the action limit is a quantified positive that still meets the numerical criterion; it should be trended and investigated if repeated, because it may signal deteriorating cleaning or supplier control. A borderline result—one close enough to the action limit that measurement uncertainty could cross it—requires a defined protocol: repeat analysis on retained homogenized sample, analysis of a second independent preparation, and where appropriate confirmation by an orthogonal method such as LC-MS/MS or PCR. Repeat testing should not be used to shop for a favorable value; the number of replicates and acceptance rule must be set in advance.
A generic batch scenario illustrates the flow. A collagen peptide lot tested immediately after a milk-containing campaign returns a milk protein value above the LOQ but below the scheme action limit. The lot is placed on hold, a second homogenized aliquot and an independent extraction are analyzed, and cleaning and changeover records are reviewed. If confirmation remains below the action limit and swabs show no residue, the lot may be released against the specification, with the finding logged for trend review. If confirmation crosses the action limit, release against an allergen-free claim is stopped and disposition proceeds through relabeling, diversion, or rejection.
Hold-and-release logic should quarantine affected lots pending investigation. Root-cause triggers include changeover history, cleaning records, rework additions, supplier certificates, and environmental swabs. Precautionary labeling is a risk-management decision based on reference-dose assessment, not a substitute for control, and should not be applied simply because a detectable signal exists below the action limit. Documentation should capture the result, method, LOQ, action limit, uncertainty, retests, contextual evidence, disposition, and labeling decision, providing an auditable basis for procurement and regulatory review.
Unexpected allergen results in functional ingredients usually arise from matrix chemistry, extraction behavior, sampling structure, or contamination during preparation rather than from random laboratory error. Diagnosis should proceed from the sample and the matrix inward to the assay.
Confirmation should use an orthogonal principle—for example, LC-MS/MS after an ELISA positive, or PCR after a protein-method negative in a heat-processed matrix—rather than repeating the same kit. Blaming the laboratory before examining matrix, sampling, and preparation causes recurring failures because the underlying interference remains in place.
Is a gluten-free result the same as an allergen-free result? No. Gluten testing addresses cereal prolamins from wheat, rye, barley, and related grains against the gluten threshold; it says nothing about milk, egg, peanut, soy, fish, shellfish, tree nuts, sesame, or other listed allergens, which require separate targets.
Can a supplier certificate of analysis replace incoming allergen testing? A COA supports qualification but does not replace verification unless it states the matrix-validated method, analyte, LOQ, action limit, sampling basis, and applicable scheme. Periodic or risk-based independent testing remains appropriate, especially after supplier or process changes.
When should testing be repeated after a process change? After any change that alters residue behavior—new shared equipment, modified cleaning, changed production sequence, new rework routing, new supplier or origin, or maintenance opening enclosed equipment. Repeat at the first affected lots and continue until a compliant history is re-established.
How are hydrolyzed or fermented ingredients tested? Protein fragmentation can invalidate sandwich ELISA and degrade PCR targets. Competitive ELISA and LC-MS/MS peptide-marker methods are generally preferred, with validation performed on the processed matrix rather than the starting material.
What does an advisory statement such as "may contain" imply? It indicates that a risk assessment found plausible unintended allergen presence that cannot be eliminated; it is not a substitute for control and should be based on documented evaluation rather than used as a general precaution.
Functional ingredient allergen testing requirements are defined by hazard origin, applicable thresholds, matrix-validated methods, representative sampling, and manufacturing context rather than by a single universal assay. Intentional ingredients, carry-over, and cross-contact require different controls, and protein targets must be distinguished from DNA targets, particularly in hydrolyzed and heat-processed gelatin and collagen materials. Declarative rules, VITAL reference doses, the 20 mg/kg gluten threshold, and scheme-defined protein action levels must be applied separately, with LOQ and analyte clearly stated. Methods—ELISA, PCR, lateral flow, and LC-MS/MS—each have matrix-specific limits, and acceptance depends on in-matrix validation that brackets the action limit. Sampling design, homogenization, and extraction determine whether a result represents the lot, while shared equipment, cleaning, rework, and supplier changes drive testing frequency. Positive, negative, and borderline findings should be handled through a pre-defined hold, retest, and root-cause flow, with orthogonal confirmation used when matrix interference is plausible.
Contact our technical experts for personalized assistance.