Formulation
THE SHORT ANSWER
Supplement stability testing assesses whether a finished product stays within defined quality limits over time in its intended packaging and storage conditions. A useful program combines appropriate analytical tests with physical, microbiological and packaging checks. Accelerated studies can identify risks and support an initial assessment, while real-time data are needed to confirm the proposed shelf life. The formula, manufacturing process and package must be considered together.
Reviewed 28 October 2026 · Canadian NHP context with technical references
What makes a formula stable—and effective?
A launch specification is a snapshot. A stability program asks whether the product will continue to meet its quality requirements during storage. A capsule can retain its declared amount yet become slow to disintegrate; a powder can meet its assay limit while becoming hard to scoop; an oil can retain its fatty-acid content while its oxidative quality deteriorates.
Effectiveness is a separate question. Select an ingredient form, dose and intended use supported by relevant evidence, then check that the chosen format can deliver them. Stability testing does not prove clinical benefit. It helps establish whether the product retains the attributes on which its quality and claims depend.
For Canadian NHPs, section 52 assigns manufacturers and importers the obligation to determine how long the packaged product maintains specified quality attributes under its recommended storage conditions, or at room temperature if none are recommended. [1]
Real-time versus accelerated stability testing
Real-time studies: performance under intended storage
Real-time studies follow the product under its recommended conditions over the proposed shelf life. They provide direct evidence of changes as the product ages. The practical question is: will this formula, made by this process and sold in this pack, remain acceptable for the period on the label?
Accelerated studies: earlier information, with limits
Accelerated studies use more stressful conditions to increase the rate of change. They can help compare development options and identify weaknesses sooner. However, heat or humidity can produce changes that do not track normal storage, including melting, phase separation or a different degradation mechanism. There is no universal rule that a fixed number of accelerated months equals two years on the shelf. Health Canada calls for real-time confirmation of extrapolated shelf life. [2]
Which temperature and humidity should you use?
ICH Q1A(R2), a pharmaceutical reference, includes general-case examples such as 25°C ± 2°C / 60% RH ± 5% RH for long-term storage and 40°C ± 2°C / 75% RH ± 5% RH for accelerated storage. These are not blanket requirements for every supplement. Refrigerated products, moisture-sensitive systems, semi-permeable packs and different markets may need other conditions. NHPs are outside ICH’s direct scope; use its principles only where justified for the product. [2, 3]
Development stress tests, transport simulations and after-opening studies answer additional questions. Write down which question each study addresses so a useful screening result is not mistaken for a complete expiry-date justification.
An illustrated stability workflow
Use this sequence to connect formulation decisions with the evidence needed for a shelf-life decision. The steps are a planning framework, not a prescribed regulatory protocol.
1
Define the product
Record the formula, dose, process, market and proposed shelf life.
2
Assess the risks
Identify likely changes in potency, moisture, oxidation and physical quality.
3
Lock the study plan
Agree on packs, batches, methods, limits, conditions and pull dates.
4
Store and test
Collect baseline data, monitor conditions and test scheduled samples.
5
Investigate and trend
Review every result; investigate failures and unexpected changes.
6
Decide and maintain
Document shelf-life conclusions and assess changes to the product.
Choose test parameters that can reveal meaningful change
Start with the failure you are trying to detect, then choose the measurement. Ask the laboratory to explain what each method can demonstrate in your finished-product matrix, its sample requirement, and how a result will be interpreted. The following are starting points for a product-specific discussion—not a mandatory test panel for every format. [4]
Capsules and tablets
Medicinal-ingredient quantity and relevant potency, using suitable assays where available.
Moisture or water activity where relevant, appearance and capsule-shell condition.
Disintegration or other performance tests appropriate to the dosage form; microbiological quality where relevant.
Powders and stick packs
Relevant assays, moisture behaviour, caking and changes in colour or odour.
Reconstitution, dispersibility, sedimentation and sensory performance at the labelled preparation conditions.
Sachet integrity and whether the product remains usable after normal handling.
Gummies and chewables
Relevant assays plus texture, sticking, deformation and water activity or moisture behaviour.
Microbiological quality and suitability of the preservation approach.
The effect of the production process and package on flavour, colour and physical integrity.
Liquids, emulsions, oils and probiotics
For aqueous liquids: pH, appearance, precipitation or separation, relevant assays and microbiological control.
For oil products: appropriate oxidation measurements alongside composition and sensory checks.
For probiotics: a suitable viable-count method and stability assessment in the intended pack.
For emulsions and specialised delivery systems: relevant physical attributes, such as droplet-size behaviour, where they affect performance.
Release testing and stability testing serve different purposes. Do not automatically repeat every raw-material test at every stability pull, or omit a changing attribute simply because it was acceptable at release. Document the rationale for the selected panel.
Potency loss: check the method before interpreting the trend
A raw-material certificate of analysis does not establish the stability of the finished blend. Other ingredients, extraction efficiency and degradation products can affect measurement. Ask whether the procedure is suitable for your matrix and can distinguish the target from relevant interference. Stability-indicating analytical principles are described in ICH Q2(R2); that pharmaceutical guidance is a technical reference, not a universal NHP method requirement. [5]
For a complex botanical, be precise about what is measured. A marker result can be useful without representing every constituent or proving the biological effect of the complete extract. Where an ingredient is quantified by input rather than finished-product assay, identify that limitation and establish a justified approach instead of presenting an unmeasured amount as a demonstrated stability result.
A practical review should compare actual numeric results, method changes, units and sample preparation. A sequence of “pass” certificates hides the direction and size of change. Overages also need justification: adding extra material does not resolve a poor method, uncontrolled degradation or an unsuitable package.
Moisture content and water activity are different measurements
Moisture content describes how much water a material contains. Water activity describes the water’s thermodynamic availability; it is not simply another expression of moisture percentage. Composition and temperature influence the relationship between the two. Two products with similar moisture contents can therefore behave differently. [6]
For a development review, ask what the moisture-related failure actually looks like: clumping, shell softening, texture drift, a reaction between ingredients or an increase in microbiological risk. Select measurements that can distinguish these possibilities. A single water-activity reading does not establish overall safety or replace the rest of the control strategy.
Compare the baseline, stored samples and package performance together. For example, a rising moisture result in one pack but not another is a reason to investigate the closure and moisture barrier as well as the formulation. If a desiccant is used, assess the complete system, including its compatibility and capacity over the intended use period.
Oxidation: look beyond a single number or a smell check
For applicable oil products, peroxide value examines primary oxidation products, while p-anisidine value reflects certain secondary oxidation products. The two describe different stages. A low peroxide result on its own does not establish that an oil is fresh. Codex’s fish-oil standard also notes that flavourings can interfere with oxidation measurements, so confirm method suitability for the actual matrix. [7]
Plan the investigation around oxygen exposure, light, temperature, headspace, closure integrity and the antioxidant system. Compare packaging options under a controlled study rather than assuming that an amber bottle or an antioxidant addition solves the problem.
Use a defined sensory procedure as supporting information: specify the attributes, reference samples, scoring approach and conditions of assessment. Do not claim that taste or odour will always detect a problem before an analytical limit is reached. Conversely, an acceptable assay does not make an objectionable flavour commercially acceptable. Record both.
Packaging compatibility is part of the stability question
The study pack should represent the commercial container-closure system or have a scientifically justified equivalence. A change in bottle material, seal or pack size deserves assessment; it should not inherit an expiry date solely because the ingredient list stayed the same. [2]
For each packaging candidate, our recommended comparison sheet records the container material and size, closure and liner, seal, fill level, headspace, light protection, any desiccant, and relevant supplier barrier information. Link that sheet to a packaging revision number so the test results remain traceable.
Separate three questions: can moisture or oxygen enter; can product components interact with the packaging; and can the pack survive handling and use? Depending on the product, investigate leakage, seal integrity, changes to the inner surface, sorption or other compatibility concerns. A successful container test does not automatically answer every question about the finished product.
Consider consumer use as well. Repeated opening, a scoop introduced into a powder, or a dropper used over several weeks can create a different exposure from an unopened study unit. Where relevant, define an after-opening or reconstitution study instead of extending an unopened result by assumption.
A sample stability study plan
Illustrative planning example: a room-temperature capsule product with a proposed 24-month shelf life and two candidate bottle systems. This is a discussion template, not a validated protocol or a prescribed Canadian testing schedule. Select conditions, batches, methods and criteria for the actual product.
Product definition: record the formula revision, dosage form, intended dose, market and proposed storage statement.
Representative material: justify the batches, scale and manufacturing process; document manufacture dates and any bulk holding time.
Packaging: define both candidate bottle systems completely, including closure, seal, fill count and any desiccant.
Real-time schedule example: baseline, 3, 6, 9, 12, 18 and 24 months.
Accelerated schedule example: baseline, 1, 2, 3 and 6 months, where accelerated conditions are scientifically appropriate.
Test panel: select relevant assays, physical performance, moisture-related measurements, microbiological checks and package assessments.
Acceptance and action rules: approve the specifications, methods and investigation triggers before results are generated.
Governance: identify who pulls samples, tracks conditions, receives results, investigates concerns and approves the report.
How many samples should you reserve?
Build the sample calculation from the laboratory’s actual needs. If this example uses three justified batches, two package systems and two unopened units per pull set, seven real-time sets plus four additional accelerated sets would require 132 units: 3 × 2 × 2 × 11. This counts one shared baseline per batch/package combination and excludes contingency and additional studies. It is illustrative arithmetic; two units may be insufficient for your test panel.
Add the agreed quantity for investigations, repeats where justified, after-opening studies and other work. Keep stability samples distinct from any separate regulatory retention-sample obligation. Confirm sampling and shipping instructions before the first pull.
Read the trend, investigate the result, document the decision
An out-of-specification result falls outside the approved acceptance criteria. An out-of-trend result may still meet those criteria but depart from the expected pattern. Define how your team will flag both. A passing result deserves attention when it is moving quickly toward a limit.
Our suggested review sequence is to check the sample identity, storage history, method and calculations; assess possible laboratory and product causes; document the investigation; and evaluate the consequences for the study and marketed product. Do not discard an inconvenient result or keep retesting until a passing value appears.
When a justified statistical analysis is used, consider batch variability, relevant attributes and the uncertainty of extrapolation. ICH Q1E discusses these concepts for pharmaceuticals; it is useful background, not permission to extend any supplement’s shelf life by a standard multiplier. [8]
The conclusion should state which formula and packaging configuration the data support, the justified storage conditions, remaining uncertainties and any required follow-up. A change to the formula, manufacturing process or package should trigger an assessment of whether the supporting evidence still applies. [2]
What determines study cost and turnaround?
The main cost drivers are the number of batches, pack configurations, storage conditions, pull points and analytical methods. Method development can be a significant part of a project when a finished-product assay is not already suitable. Special storage, microbiological work, packaging investigations and additional sample handling can add work.
Ask for a quote that separates protocol development, setup, storage, each testing round, data review and the final report. Also clarify ownership of raw data and the cost of investigations or protocol changes. A cheap storage quote is not the complete cost of a stability program.
Distinguish laboratory turnaround after a pull from the calendar length of the study. Faster reporting does not make a 24-month real-time observation occur sooner. Earlier development decisions may be possible from screening or interim results, with their limits stated clearly.
Before you start: a study-planning checklist
Define the target market, product, proposed shelf life and storage statement.
Confirm the formula, ingredient forms, dose and manufacturing revision.
Identify plausible chemical, physical, microbiological and packaging failure modes.
Justify the chosen batches, scale and commercial pack configurations.
Agree on real-time conditions and whether accelerated or additional studies are appropriate.
Confirm suitable methods, units, acceptance criteria and sample requirements.
Set baseline and pull dates with a clear rule for missed or delayed events.
Calculate samples for every test, condition, batch and pack, plus justified contingency.
Assign responsibility for storage monitoring, shipping, testing and data access.
Define investigation and escalation procedures for failures and unexpected trends.
Agree on interim reviews, reporting, approval and record retention.
Connect future formula, supplier, process and packaging changes to stability review.
Supplement stability testing: common questions
Can accelerated testing prove a 24-month expiry date?
Not by itself as a universal rule. Its usefulness depends on the product, the failure mechanisms and the justification for extrapolation. An appropriate real-time program is needed to confirm a proposed shelf life. [2]
Is a supplier COA enough?
No. It describes results for the material or batch covered by that document. It does not demonstrate how your finished formula behaves throughout storage in its commercial packaging.
Should the study include the final packaging?
Use the commercial system or document a scientifically justified equivalent. Treat a different closure, barrier material, fill volume or pack size as a change to assess, not automatically as an interchangeable pack.
Do all supplements need the same testing panel?
No. Select tests for the formulation, format, risks and applicable requirements. A powder, an aqueous liquid and an oil product can need very different measurements.
Can a formula be chemically stable but still fail?
Yes. The product may retain the assayed ingredient while its texture, disintegration, microbial quality, dispensing performance or packaging becomes unacceptable. Define the relevant attributes before the study.
What should I send Kompete to discuss a stability plan?
Send the current formula and specifications, target markets, proposed label claims and shelf life, manufacturing information, packaging details and any existing data. Highlight known issues such as clumping, odour changes, separation or declining assay results.
Related practical guides
Before the first production run: Is your supplement formula ready for manufacturing?
Before approving artwork: Canadian NHP labelling requirements
For project support: explore analytical testing or request a formulation and stability review.
Sources and scope
References checked 28 October 2026. The checklist and sample schedule are planning examples. Pharmaceutical and food references are identified for their technical principles; their full requirements should not be assumed to apply automatically to every NHP.
[1] Natural Health Products Regulations, section 52
[2] Health Canada GUI-0158: Stability period, section 52
[3] ICH Q1A(R2): Stability Testing of New Drug Substances and Products
[4] Health Canada: Quality of Natural Health Products Guide
[5] ICH Q2(R2): Validation of Analytical Procedures (EMA)
[6] FDA: Water Activity (aw) in Foods