Why Vitamin D Potency Results Vary Between Labs — and What It Means for Your Label Claims

A supplement brand sent the same vitamin D3 softgel to two ISO 17025 accredited labs last year. Lab A came back with 2,180 IU per softgel. Lab B returned 1,730 IU. The label said 2,000 IU. Both labs were right — at least by the standards of their own validated methods. And that’s the problem.

This isn’t a hypothetical. Inter-laboratory disagreement on vitamin D potency is one of the most consistent findings in supplement testing proficiency programs, and most brands have no idea it’s happening until they get two COAs that don’t agree. The disconnect isn’t about lab quality. It’s about method variation — and understanding it is the difference between a defensible label claim and a compliance problem waiting to happen.

Why Vitamin D Is Harder to Quantify Than It Looks

Vitamin D3 (cholecalciferol) is a fat-soluble vitamin. It doesn’t dissolve in water, and it doesn’t sit cleanly in a capsule shell the way ascorbic acid does. Almost every vitamin D product on the market — softgels, gummies, chewables — delivers the vitamin suspended in an oil matrix: MCT oil, sunflower oil, olive oil, or a blend.

Before any chromatographic measurement can happen, you have to destroy that matrix. The standard approach is saponification: heating the sample with potassium hydroxide in ethanol, which hydrolyzes the fat into glycerol and fatty acid salts. The vitamin D, being unsaponifiable, is then extracted into an organic solvent and cleaned up for injection.

This sounds routine. It isn’t. Saponification introduces at least three independent sources of result variation before a single drop hits the HPLC column:

• Temperature and hold time: Most methods specify 70°C for 30 minutes under reflux. A lab running at 60°C for 20 minutes — still within a plausible range of “similar” protocols — recovers less vitamin D from oil-rich softgels, systematically underreporting potency.
• KOH concentration: Some methods use 50% aqueous KOH; others specify 60%. Higher alkalinity drives more complete hydrolysis of glycerides but also accelerates vitamin D isomerization if the reaction temperature isn’t tightly controlled.
• Antioxidant protection: Vitamin D oxidizes and photodegrades readily. Adding pyrogallol or BHT (butylated hydroxytoluene) to the saponification flask protects the analyte during the reaction. Labs that omit this step — or use too low a concentration — will underreport potency by 5–12% in oil-matrix products, depending on extraction duration and ambient light exposure.

None of these variables requires negligence to vary between labs. Validated methods have allowable ranges, and those ranges are wide enough to produce real, reproducible differences in analyte recovery.

The Calibration Problem No One Talks About

After extraction, the lab builds a calibration curve and quantifies the vitamin D against it. The curve’s accuracy depends entirely on the reference standard used to construct it.

The authoritative source is USP Cholecalciferol Reference Standard — a certified material with traceable potency tied to NIST measurements. Labs that use this standard, recertify it on schedule, and control the dilution series carefully produce calibration curves with well-documented uncertainty.

Not all labs do. Some maintain in-house secondary standards prepared from a single lot of bulk cholecalciferol purchased years ago and diluted down over time. Secondary standards aren’t inherently wrong — they’re common in high-throughput environments. But if the lot was never verified against primary USP material, or if storage conditions weren’t rigorously controlled (vitamin D3 degrades above 25°C and is sensitive to light and oxygen), the calibration curve itself carries a systematic offset. Every result it generates will be off by the same factor, and it won’t show up in routine system suitability checks.

External proficiency data from programs like CAP’s Chemistry Proficiency Surveys and FAPAS (Food Analysis Performance Assessment Scheme) consistently shows inter-laboratory coefficients of variation of 10–18% for vitamin D assays across ISO 17025 accredited participants. For a product labeled at 2,000 IU, that CV translates to a plausible result spread of roughly 1,640–2,360 IU across labs using different but fully defensible methods. This isn’t a scandal. It’s the honest analytical reality of a fat-soluble vitamin in a complex food matrix, and brands that don’t know about it are flying blind.

HPLC-UV vs. LC-MS/MS: Why the Detection Method Changes Everything

Most labs run vitamin D3 by reverse-phase HPLC with UV detection at 265 nm. This is the approach described in AOAC Official Method 2002.05, and it performs reliably for straightforward softgel and hard capsule matrices where the oil extraction is clean.

Gummies are a different story entirely. Gelatin, sugars, colorants, and co-extracted lipid oxidation products all absorb near 265 nm. When they co-elute at or near the vitamin D peak, the UV detector integrates those interferences into the result — inflating or deflating reported potency depending on the co-eluting compound’s spectral profile. On a gummy with a natural carrot-derived colorant, we’ve seen apparent potency inflate by 8–11% on HPLC-UV versus the true value confirmed by LC-MS/MS. That’s not analyst error. That’s the physics of the detector.

LC-MS/MS sidesteps this entirely. Detection by multiple reaction monitoring — monitoring the precursor ion at m/z 385.3 for cholecalciferol and the characteristic product ion at m/z 159.1 — is essentially immune to co-eluting matrix components that don’t share the same fragmentation pattern. For complex matrices, LC-MS/MS routinely delivers inter-assay coefficients of variation of 3–6%, compared to 8–15% for HPLC-UV on difficult substrates.

The problem: labs don’t always disclose which method they used, and brands rarely ask. If your product is a gummy vitamin D supplement and the COA was generated by HPLC-UV, the number on that report carries more analytical uncertainty than the two decimal places suggest. You might be within spec. You might not be. Without LC-MS/MS confirmation, it’s genuinely hard to know.

What FDA’s cGMP Framework Actually Requires

Under 21 CFR Part 111, dietary supplement manufacturers must establish specifications for identity, purity, strength, and composition, then verify that each finished lot meets those specifications before release. FDA does not set a universally mandated potency range for vitamin D — that specification is the manufacturer’s responsibility to define and defend.

Most well-run brands set release specifications at 100–120% of label claim and establish a minimum end-of-shelf-life specification of 100%. The intentional overage compensates for vitamin D3 degradation over an 18–24 month shelf life under ambient storage conditions — a degradation rate that varies considerably based on oxygen headspace, container closure system, moisture ingress, and UV exposure during distribution.

Problems arise when brands have no specification at all, or when the contract manufacturer’s outgoing COA and the brand’s independent verification test were generated by different methods and nobody reconciled the difference. We’ve worked with brands that received a CMO COA showing 108% of label claim and an independent third-party result of 91% on the same production lot. Both results were analytically plausible given the method differences involved. But without knowing each lab’s extraction and detection parameters, there’s no way to determine whether the lot is actually compliant — or to decide whether to release it.

That kind of ambiguity is expensive. It costs in investigation time, potential rework, and delayed release dates. It’s also entirely preventable.

What ISO 17025 Accreditation Controls — and What It Doesn’t

ISO 17025 accreditation means a lab has documented its methods, validated them against defined performance criteria, demonstrated proficiency through internal and external testing, and maintains calibration traceability to national measurement standards. It eliminates a large class of errors from the risk picture: uncontrolled instrument drift, undocumented method deviations, untrained analysts running unsupervised.

But it doesn’t standardize method selection across labs. Two accredited labs can both be fully compliant with ISO 17025 while using different saponification temperatures, different reference standard sources, and different detection platforms. Accreditation guarantees that each lab is doing what it says it’s doing — not that both labs are doing the same thing.

This is why choosing a testing lab for vitamin D goes beyond verifying accreditation status. The right questions to ask before committing a product to a lab’s scope:

1. What reference standard do you use for vitamin D calibration, and when was it last certified against USP primary material?
2. What extraction method do you run — specifically, what is the saponification temperature, hold time, and KOH concentration?
3. Do you use HPLC-UV or LC-MS/MS for final detection, and do you make that choice based on matrix type?
4. Do you participate in external proficiency testing for vitamin D, and can you share your most recent z-scores?

A lab that answers all four questions with specifics — not just “our methods are validated and accredited” — has the analytical transparency your label claim requires. Vague answers to any of these questions are a signal worth taking seriously.

Protecting Your Label Claim Before It Becomes a Compliance Issue

The most practical step any brand can take is method alignment before the first production lot ships. Ask your testing lab exactly which method they’ll use. Lock that in as your reference method in your quality agreement. If you later switch labs or want to dual-qualify a second testing source, run a bridging study: test the same characterized lot at both labs and document the method-to-method offset. If Lab B consistently returns 5% lower than Lab A on your specific matrix type, that offset belongs in your specification — not in a dispute resolution file.

At Qalitex, we run vitamin D3 quantification by HPLC-UV for softgels and hard capsules with clean oil matrices, and LC-MS/MS for gummies, powder blends, and multi-ingredient formulas where UV interference is a real concern. Our calibration standard is USP Cholecalciferol Reference Standard, recertified on an annual schedule. We report measurement uncertainty alongside every potency result on our COAs — a figure that tells brands exactly how much analytical variability is built into the method.

That uncertainty estimate tends to surprise brands who’ve never seen it on a COA before. But once you understand what it means — and what it means when a competing lab’s COA doesn’t include one — you start asking for it everywhere. It’s one of the clearest signals that a testing lab actually understands the method it’s running, rather than just running it.

---

Written by Nour Abochama, Vice President of Operations, Qalitex Laboratories. Learn more about our team

Talk to our team about your testing needs. Contact us

Related from our network

• Supplier COA Verification for Vitamin D Raw Materials — Ayah Labs provides independent verification of manufacturer-issued COAs for raw ingredients including cholecalciferol and vitamin D2, helping brands catch specification mismatches before they reach finished goods.
• Vitamin D Supplement Testing Under Health Canada’s NHP Framework — Androxa covers Canadian GMP and NHP licensing requirements for supplement brands seeking to sell vitamin D products in the Canadian market.
• Building a Defensible 21 CFR Part 111 Quality System for Supplement Manufacturers — Aurora TIC helps supplement brands design compliant quality systems and prepare for FDA inspections under the dietary supplement cGMP regulations.