Shelf-Life Dating for Botanical Raw Materials: The Stability Study Gaps Putting DSHEA Compliance at Risk

Every analytical testing laboratory we work with has a version of the same story. A supplement brand arrives with a beautiful label, a 24-month expiration date printed in clean sans-serif, and — when you ask for the stability data that justifies that date — a long pause. Sometimes it’s a supplier COA from the original batch. Sometimes it’s a literature reference that doesn’t match the brand’s actual formulation, particle size, or packaging. Occasionally, it’s nothing at all.

For a simple ascorbic acid powder or a single-compound isolate, that gap is risky. For a multi-constituent botanical extract — turmeric, ashwagandha, elderberry, lion’s mane — it’s a liability with regulatory teeth.

What 21 CFR Part 111 Actually Requires

Let’s be precise here, because this regulation is frequently misread. DSHEA’s implementing GMP rule, 21 CFR Part 111, does not require every dietary supplement to carry an expiration date. But §111.165(d) is unambiguous: if you include an expiration date on your label, you must have established data to support the shelf life that date implies.

That data requirement feeds directly into §111.87, which mandates that you establish specifications for the identity, purity, strength, and composition of each dietary ingredient. For a finished product, §111.455 requires testing against those specifications before release. If your label says the product is good for 24 months, you need test results — at or near that 24-month mark — confirming the specifications are still met.

The practical effect: an expiration date without supporting stability data isn’t just a documentation gap. It’s evidence that your release specifications may not be meaningful, which is precisely the kind of finding that appears in FDA Form 483 inspection observations. Stability-related deficiencies have been among the more frequently cited cGMP violations in dietary supplement facility inspections for more than a decade, and botanical supplement manufacturers have historically received a disproportionate share of those observations. FDA’s 2023 dietary supplement cGMP inspection summary data bears this out — specification and testing-related observations consistently rank in the top three finding categories.

Why Botanical Matrices Are So Much Harder to Stabilize

A synthetic ingredient like vitamin D3 degrades through a relatively well-understood set of pathways — primarily oxidation, accelerated by heat and UV. You can model its decline mathematically, design an accelerated study using ICH Q1A(R2) conditions (40°C/75% relative humidity for 6 months), and project a 24-month real-time profile with reasonable confidence.

Botanical extracts don’t cooperate with that model. Consider what you’re actually working with: a dried, concentrated mixture of dozens — sometimes hundreds — of phytochemicals: polyphenols, glycosides, terpenes, alkaloids, flavonoids, each with its own degradation kinetics, each capable of interacting with the others as conditions change. Turmeric extract contains curcuminoids that are simultaneously susceptible to photodegradation, alkaline hydrolysis, and oxidation. Ashwagandha withanolides are comparatively stable, but the lipophilic fraction in a whole-root extract can undergo rancidity under elevated humidity. Elderberry anthocyanins are among the most labile botanical constituents we test — half-lives at 40°C can be measured in weeks, not months.

This chemical complexity creates three problems that don’t apply to synthetic actives:

Non-linear degradation kinetics. Many botanical constituents degrade rapidly in the first 3 months and then plateau, or exhibit the reverse — a slow decline that suddenly accelerates after a threshold temperature or moisture content is crossed. Applying the Arrhenius-based projection in ICH Q1A(R2) to a multi-constituent botanical often generates numbers that simply don’t hold up against real-time data.

Matrix interference. The compounds in a botanical extract don’t degrade independently of each other. Polyphenol oxidation products can act as pro-oxidants, accelerating the degradation of neighboring constituents. Residual enzymatic activity — if the extract wasn’t properly deactivated during manufacturing — can continue in the dry state at elevated temperature and humidity.

Specification target ambiguity. Unlike a pharmaceutical active with a single defined assay value, botanical extracts are often specified against a marker compound that may not be the primary driver of the product’s quality — or its instability. Testing only curcumin content in a turmeric extract while ignoring bisdemethoxycurcumin gives you an incomplete picture of how the full extract is changing over time.

Three Stability Study Mistakes We See Most Often

Mistake 1: Relying on accelerated data alone.

ICH Q1A(R2) — designed for pharmaceutical drug substances — permits 6 months of accelerated data at 40°C/75% RH to support a 24-month real-time claim, provided the accelerated and real-time profiles track each other predictably. For many botanical extracts, they don’t. We’ve received stability samples where the accelerated set showed acceptable marker compound retention through 6 months, while the corresponding real-time samples at 25°C/60% RH developed unexpected physical changes — clumping, color shift, off-odor — that the accelerated protocol completely missed.

Accelerated conditions are a stress screen and an early warning system. For botanical raw materials, they cannot substitute for real-time data.

Mistake 2: Testing the wrong marker compounds.

Standardization markers are often chosen for commercial reasons: withanolide content in ashwagandha is marketable, so it becomes the specification. But withanolides aren’t necessarily the most stability-sensitive fraction of the extract. A defensible stability protocol starts with a characterization of the extract’s full phytochemical profile — often by HPLC-DAD or LC-MS — followed by deliberate identification of the constituents most likely to degrade. Those become your stability sentinel markers, whether or not they appear on the label.

Mistake 3: Not testing in the finished container-closure system.

A botanical extract’s stability is inseparable from its packaging. A moisture-sensitive elderberry powder that performs well in a nitrogen-flushed glass jar may fail within 90 days in a non-desiccated HDPE bottle with a standard induction seal and no desiccant canister. Running stability on bulk powder in generic lab containers — then applying those results to a specific packaging configuration — invalidates the data for label claim purposes and won’t hold up to scrutiny.

What a Defensible Stability Protocol Looks Like

A protocol that will satisfy an FDA investigator’s request for documentation — or a quality audit from a major retail buyer — needs to address each of these elements:

Time points: T=0 (baseline characterization), then 3, 6, 9, 12, 18, and 24 months for a 24-month label claim. Intermediate points aren’t optional; they’re where degradation inflection points are most likely to emerge.

Storage conditions: Real-time at 25°C/60% RH is the primary stability condition and cannot be skipped. Layer in accelerated conditions (40°C/75% RH) as a stress-screening tool, and consider intermediate conditions (30°C/65% RH) for products destined for warm-climate distribution or warm-weather storage scenarios.

Test battery: Identity confirmation at each time point — not just T=0 — using HPTLC or HPLC. Quantitative marker compound assay. Moisture content by Karl Fischer. Microbial enumeration per USP <61>/<62> at select time points. Physical appearance, color, and odor. For extracts in a finished dosage form, also include disintegration testing where applicable.

Container fidelity: Run the study in the actual packaging configuration used for market: bottle material and size, liner, closure torque, desiccant type and quantity, label. Not a proxy container.

Retained samples: Pull and reserve sufficient sample volume at each time point to allow full retesting if an out-of-specification result triggers an investigation.

This isn’t a short-term project. A 24-month real-time study takes 24 months. That’s why brands building a portfolio of botanical products need to initiate stability work at the same time they lock down their formulas — not 6 months before launch, and certainly not after the product is already on shelves.

What This Means for Midwest Brands Specifically

For supplement brands operating in the Chicago metro and broader Midwest, the stability testing gap has a logistics dimension on top of the regulatory one. Most ISO 17025-accredited laboratories capable of running a complete botanical stability protocol — with validated HPLC methods, ICP-MS for heavy metals at time-point confirmation, and USP microbiology — are concentrated on the coasts. Shipping time-sensitive stability samples across the country introduces handling variability that can compromise results and complicates chain-of-custody documentation.

Our model at Ayah Labs addresses that directly. With sample receiving at our Countryside, IL facility, Chicago-area and Midwest brands can deliver stability samples locally, within controlled temperature conditions, and have them transferred to our ISO 17025-accredited analytical testing laboratory in California for the actual assay work. Initial time-point results come back within 5–7 business days. Chain of custody is fully documented, cold-chain requirements are maintained where the extract demands it, and the accreditation behind the Certificate of Analysis is the same ISO 17025 credential your biggest retail customer will ask to verify.

That matters when a buyer’s quality team calls for documentation — or when an FDA investigator walks through your door.

Start the Clock Before You Think You Need To

The most common stability program failure we see isn’t a protocol error or a wrong testing method. It’s timing. Brands that begin stability studies after their product is already on shelves are, by definition, selling product with an expiration date they cannot yet substantiate. Every month the product sits on retail shelves without supporting data compounds the exposure.

If you’re formulating a new botanical product today, your T=0 samples should enter the stability chamber the same week the formula is locked. If you have products already on the market without stability data, start a bracketing study immediately — it’s an imperfect retrofit, but it is infinitely better than having no data when FDA or a retail quality audit asks for it.

The expiration date on your label is a regulatory commitment. Treat it like one.

---

Written by Nour Abochama, VP Operations, Qalitex | Quality Consultant, Ayah Labs. Learn more about our team

Ship your sample to our Chicago facility — get a Qalitex CoA in 5–7 days. Contact us

Related from our network

• Supplement and botanical raw material testing services — Qalitex Laboratories provides ISO 17025-accredited analytical testing for dietary supplements, botanical extracts, and raw material qualification.