NHP Stability Programmes Under GUI-0158 Version 4.0: Protocol, Report and Expiry Dating

Most stability findings we see at Canadian natural health product sites are not caused by missing data. The samples were pulled, the laboratory ran the tests, and the numbers sit in a spreadsheet somewhere. The finding is raised because there is no approved protocol that says what was going to be tested, at what intervals, against what acceptance criteria, and no signed report that draws a conclusion from the results. The science happened. The evidence trail did not.

That distinction matters more now than it did two years ago. Health Canada published version 4.0 of the Good manufacturing practices guide for natural health products (GUI-0158) on 4 September 2025, and it came into force on 4 March 2026, replacing the 2015 version. We have already covered what version 4.0 changes across the whole quality system. This article goes narrow and deep on one section: stability. It walks through what section 52 actually requires, how to build a protocol and report that hold up, the expiry dating rule that catches out companies who hold product in bulk, and the retention periods that apply once a lot has expired.

Executive summary

Stability is one of the most strengthened areas in GUI-0158 version 4.0, and it is one of the easiest for an inspector to test. An inspector can ask for one thing, the stability file for a single product, and learn a great deal about how your quality system really works. If the protocol is missing, the report is unsigned, or the expiry date cannot be traced back to data, the gap is visible in minutes.

The headline points for licence holders and importers are these:

• The legal obligation lives in section 52 of the Regulations. GUI-0158 is guidance that explains how Health Canada expects that obligation to be met and evidenced.
• Section 52 applies to manufacturers and importers. It does not apply to packaging, labelling or distributing activities on their own.
• Every NHP marketed in Canada needs a completed stability study, an implemented protocol, and a maintained report.
• Responsibility stays with you even when someone else runs the testing, including a foreign manufacturer or the product licence holder.
• The expiry date is counted from the date of original batch manufacture, not from the date the product was packaged.
• Records, retention samples and even your test methods and equipment carry obligations that run for at least one year past expiry.

What section 52 requires

What it is. Section 52 of the Natural Health Products Regulations requires every manufacturer and every importer to determine the period of time during which, after being packaged for sale, the product will maintain its purity and physical characteristics, and its medicinal ingredients will maintain their quantity per dosage unit and their potency. That period is measured under the recommended storage conditions on the label, or at room temperature if the product has no recommended storage conditions.

Why it matters. The expiry date on your label is a claim. It tells a consumer that the product still meets its specifications on that date. Without a stability programme behind it, that claim is an assertion rather than a conclusion. An unsupported expiry date is an inspection finding, and if a product is later shown to fall out of specification before expiry, it becomes a recall.

What companies should do. Build a written stability programme procedure, then generate three artifacts for every product. GUI-0158 is explicit that manufacturers and importers must have written procedures and practices for stability studies so that finished products comply with approved specifications.

Your stability programme procedure should cover five things: assigning the initial expiry date, conducting real time studies, collecting additional stability data when needed, designing the protocol, and maintaining the report. If you are rewriting this procedure, our guidance on how to write SOPs that pass a Health Canada inspection applies directly here.

Who carries the obligation, including importers

What it is. Section 52 sits with manufacturers and importers. GUI-0158 makes clear that you are responsible for obtaining, maintaining and reviewing up to date records on the stability programme, and that this holds regardless of who actually performs the testing. It does not matter whether the work was done by you, the product licence holder, the foreign manufacturer, the distributor or another responsible person.

Why it matters. This is where importers get caught. A foreign supplier sends a one page stability summary, the importer files it, and everyone assumes the obligation is discharged. It is not. The Canadian importer must be able to defend the study, not merely possess a certificate referring to it.

What companies should do. Stability studies from foreign sites may be accepted. If you rely on them, GUI-0158 sets four specific importer duties:

• Ensure the stability protocols used comply with Canadian requirements.
• Review, evaluate and maintain the stability results.
• Obtain any required tests that are not included in the studies collected from foreign sites.
• Maintain a complete stability report for the NHP released in Canada.

The third duty is the one that costs money. If a foreign study omits a test that Canadian specifications require, for example a microbial limit or a preservative effectiveness check, you have to obtain it. Build that expectation into the quality agreement before you sign it, not after the first shipment lands. Our guide to supplier qualification programmes that survive an inspection sets out how to tier suppliers and write agreements that carry this obligation.

Assigning the initial expiry date

What it is. Every finished NHP sold in Canada must carry a lot number and an expiry date. To set the initial date, you predict shelf life from known data. GUI-0158 allows two sources: results of accelerated or real time studies on the product itself, and information gathered from similar products in similar packaging.

Why it matters. There is a persistent myth that accelerated testing is now banned for NHPs. That is not what the guidance says, and getting this wrong in either direction is costly. Accelerated testing generates preliminary shelf life data quickly and may be used to predict an initial expiry date. What Health Canada says is that it may be unreliable for assigning an expiry date, that you must judge whether it is appropriate for your dosage form, and that real time testing must be used to confirm it.

What companies should do. Be cautious with accelerated data for products that can separate, melt, soften, crack or degrade under high temperature and humidity. Softgels, gummies, cream bases, effervescent formats and probiotics are the obvious candidates. For these, elevated temperature can drive a failure mode that never occurs at label conditions, or mask one that does.

If you lean on data from similar products, the bar is documentary. Keep evidence of the manufacturing and packaging process for those products, such as a batch record, and evaluate and document any differences that might affect stability. Similar formulations should differ only in ingredients assessed to have no or minimal effect on stability, for example colourants or flavours. Similar packaging should differ only in components with no or minimal effect, for example label content. Keep the stability results of the similar products you relied on, because those results are now part of the justification for your expiry date.

Real time studies confirm the shelf life

What it is. Real time stability studies must be conducted and used to verify any shelf life that was proposed by extrapolation. The data must show that each finished product still meets its label claims at the expiry date when stored according to its labelled conditions.

Why it matters. A real time study only proves something about your commercial product if the material on stability behaves like your commercial product. This is why the lot selection rules exist, and it is a common weak point in files we assess.

What companies should do. Follow an established protocol and make sure the lots qualify. Product lots must be of the same formulation and packaged in the same container closure system, or a scientifically demonstrated equivalent, as those sold commercially. The manufacturing process used for stability lots must match the commercial process, deliver the same quality, and meet the same specifications.

Storage areas need real controls. Stability samples may sit in dedicated chambers or dedicated rooms, but those areas must be monitored for temperature, humidity and, where applicable, light. Conditions must stay within the ranges in the protocol and consistent with the product licence and label, and the monitoring records must be kept.

The expiry date starts at manufacture, not at packaging

This single rule causes more quiet non compliance than any other part of section 52. The expiry date printed on the package must be assigned from the date of original batch manufacture, not the date of packaging.

Health Canada gives the example directly. If a product has a 24 month shelf life, the expiry date printed on the packaging is 24 months after the date of manufacture. That holds even if the product was held in bulk for 6 months before packaging. Companies that date from the packaging run effectively grant themselves an extra six months of shelf life that no study supports.

Bulk holding also affects study design. Because time in bulk can influence stability, you should consider including batches stored at the limits of extended hold times, for example more than one month, so the study reflects a worst case rather than an ideal one. The Regulations do not require expiry dates for bulk product, but data on hold times gives you the assurance, and the justification, that the material was still fit when it was packaged.

Designing the stability protocol

What it is. The protocol is the controlled document that makes a study interpretable. GUI-0158 says it should be designed to provide data in a form and quantity that permits realistic analysis, and that one should be created for each stability study.

Why it matters. If the acceptance criteria are written after the results are known, the study proves nothing. The protocol is what converts a set of numbers into evidence.

What companies should do. Cover the following, as applicable to the product.

Choose test parameters that can actually change. Select properties that may shift during storage or that could affect label claims, product safety, efficacy or quality. One specific instruction is worth noting: minerals are often assumed to be stable and left out of studies, but GUI-0158 states that mineral medicinal ingredients should still be included, because ingredients can change in unexpected ways once mixed, manufactured into a dosage form and packaged.

There is a practical problem the guidance addresses head on. When medicinal ingredients are quantified by input rather than by assay, they cannot be assayed at expiry. In that case, evaluate other observable changes over shelf life, such as colour, odour, viscosity, disintegration and microbial contamination. Health Canada also recommends assigning a shorter shelf life where medicinal ingredients cannot be quantified by assay, and revisiting the programme when suitable test methods become available.

Maintaining the stability report

What it is. The report summarizes the data generated from each protocol and records the evaluation and conclusions. It is the document that connects raw results to the expiry date you print.

Why it matters. An inspector reading a stability report should be able to answer one question without asking you: on what basis was this shelf life assigned. If the report stops at a results table, the answer is not there.

What companies should do. Include the study details from the protocol, storage conditions, sample orientation and stability specifications with test methods and acceptance criteria. Add product specifics: name or NPN, batch or lot code, date of manufacture, expiry date and packaging details. Reference any deviation reports that record departures from the protocol. Record the packaging evaluation, covering inner walls and cap, seal integrity, label appearance and adhesion, and readability of the lot number and expiry.

Then add the parts that make it a report rather than a data dump: test results for each parameter in the specifications, a scientific evaluation of the data, an overall assessment of the findings, an assessment of any out of specification, borderline or atypical trends, and the signature and approval of the QAP.

That last point deserves emphasis. The QAP for the party responsible for the stability of the product in Canada reviews, approves, signs and dates the stability reports, and addresses any deficiencies. For an importer, this means your Canadian QAP signs off on a study that a foreign site may have executed. They cannot do that credibly without the underlying protocol and results in hand.

Reducing the workload with bracketing and matrixing

What it is. If a product comes in several package types or sizes, you may be able to reduce the number of studies using bracketing or matrixing. GUI-0158 points to ICH guideline Q1D on bracketing and matrixing designs as a useful model.

Why it matters. For a company with one formulation in five bottle sizes, testing every size at every time point is expensive and adds little knowledge. A justified reduced design frees budget for the studies that matter.

What companies should do. Use these designs deliberately and write down the rationale. Two cautions apply. First, NHPs are not within the scope of ICH guidelines, so Q1D is a helpful reference rather than a binding standard, and your justification has to stand on its own. Second, bracketing and matrixing reduce the number of studies. They do not remove the requirement for real time data to confirm the shelf life you claim.

Handling out of specification and out of trend results

What it is. A stability programme that never produces an uncomfortable result is usually a programme that is not looking hard enough. GUI-0158 requires the protocol to include procedures for evaluating out of specification results and negative trends before they occur.

Why it matters. A stability failure is rarely confined to the sample on test. The lot on stability represents commercial lots that are already with consumers, which is why the guidance is explicit that you must consider the impact on all batches of the product that are on the market.

What companies should do. When you observe an out of specification, borderline or atypical trend, take action: investigate, run additional stability studies, change the shelf life, or recall the product, as the situation requires. Out of trend results should be investigated even when the result is still within specification, because that is the window in which you still have options. Investigations that stop at the laboratory are the reason so many of these findings repeat, which we cover in why CAPA keeps failing.

Shelf life is not fixed once assigned. You should confirm and adjust the expiry date when required, basing updates on real time studies of product stored under labelled conditions for the period indicated. Shelf life can be extended if real time data show the product is stable for longer than anticipated. It must be re evaluated when you make significant changes to the formulation, the manufacturing process or the packaging.

Records, retention samples and the one year rule

What it is. Several obligations continue after a lot expires. They are easy to miss because they are spread across sections 52, 58 and 61, and they are the ones most often broken by a warehouse clean out or a laboratory closure.

Why it matters. If Health Canada asks how you determined a shelf life for a product you discontinued last year, the answer needs to exist. The same applies to retention samples and, less obviously, to the equipment and reference materials needed to run the tests.

What companies should do. Map these retention periods into your record and sample management procedures.

The equipment obligation is the one companies overlook. GUI-0158 flags it as especially important when a product is discontinued or when a manufacturing facility or testing laboratory closes. If you are winding down a product line or changing contract laboratories, confirm you can still run the specification tests for a year past the last expiry, or make documented arrangements before the capability disappears.

Retention samples also have storage rules. Keep them under the conditions listed on the label and consistent with the product licence, in a temperature and humidity monitored area. Where a batch is packaged in two or more distinct packaging operations, or has been repackaged, take at least one retention sample from each operation so you can assess that specific run. Health Canada may request a sample large enough to complete full finished product testing in duplicate, and some companies hold triplicate so they retain material for their own investigation.

Stability compliance checklist

Work through this against one product file first, then scale across the portfolio.

• A written stability programme procedure exists and covers initial expiry dating, real time studies, additional data, protocol design and report maintenance.
• Every NHP marketed in Canada has a completed stability study, not only the products launched most recently.
• Each study has an approved written protocol that predates the results.
• Each study has a stability report containing a scientific evaluation and conclusions, signed and dated by the QAP.
• Stability lots match commercial formulation, container closure system and manufacturing process.
• Expiry dates are calculated from the date of manufacture, and bulk hold time is accounted for.
• Any shelf life set by extrapolation has real time data confirming it, or a dated plan to generate that data.
• Stability storage areas are monitored for temperature, humidity and light where applicable, with records retained.
• Imported products have protocols confirmed against Canadian requirements, with any missing tests obtained.
• Signed quality agreements define stability and retention sample responsibilities across every site involved.
• Out of specification and out of trend procedures exist, and past events show market impact was assessed.
• Shelf life is re evaluated after significant formulation, process or packaging changes.
• Retention periods for records, samples, and test capability are mapped and enforced.

Common mistakes

• Data without a protocol. The most frequent finding. Results exist, but no approved document defines the test plan or acceptance criteria, so the data cannot support a conclusion.
• Dating from packaging. Product held in bulk then dated from the packaging run, silently extending shelf life beyond what the study supports.
• Inherited supplier expiry dates. Adopting a foreign supplier's shelf life without confirming the protocol meets Canadian requirements or obtaining missing tests.
• Development batches standing in for commercial ones. Studies run on pilot material that does not match the commercial process or container closure system.
• Leaving minerals out. Assuming mineral ingredients are inherently stable and excluding them from the study.
• Unsigned reports. A complete report that the QAP never reviewed, approved, signed and dated.
• Treating an out of trend result as a laboratory issue. Closing an investigation without assessing the impact on lots already in the market.
• Losing the means to test. Discontinuing a product or changing laboratories without preserving the ability to run specification tests for a year past the last expiry.

Frequently asked questions

Is accelerated stability testing still allowed for natural health products?

Yes. GUI-0158 permits accelerated results to be used when predicting a shelf life and assigning an initial expiry date. The limitation is that accelerated testing may be unreliable for that purpose, you must judge whether it suits your dosage form, and real time testing must be used to confirm the data. Accelerated results on their own do not support a final shelf life claim.

Does section 52 apply to packagers, labellers and distributors?

No. Section 52 applies to manufacturing and importing activities. Packaging, labelling and distributing are covered by other GMP sections, and distributors and storage operators still carry obligations for storage conditions, records and retention samples. If you both package and import, the importing activity brings you into section 52.

Can I rely on stability data generated by my foreign manufacturer?

Stability studies from foreign sites may be accepted. As the importer you must ensure the protocols comply with Canadian requirements, review, evaluate and maintain the results, obtain any required tests missing from the foreign study, and maintain a complete stability report for the product released in Canada. Possessing a supplier summary is not the same as meeting these duties.

How is the expiry date calculated if product sits in bulk before packaging?

From the date of original batch manufacture. Health Canada's example is a 24 month shelf life on product held in bulk for 6 months: the printed expiry is still 24 months from manufacture, not 24 months from packaging. You should also consider including batches held at extended bulk hold times in the study so the design reflects a worst case.

What if a medicinal ingredient cannot be assayed at expiry?

Where ingredients are quantified by input rather than assay, evaluate other changes over shelf life such as colour, odour, viscosity, disintegration and microbial contamination. Health Canada recommends assigning a shorter shelf life in these cases, and updating the stability programme when a suitable test method becomes available for your product.

How long must stability records and retention samples be kept?

Keep the record of how each shelf life was determined for one year after the last lot of that product has expired. Batch related records are kept for one year following the expiry date under section 58. Retention samples are kept in the original packaging for at least one year after expiry, and you must retain the ability to run specification tests until one year after the last batch manufactured has expired.

Do I need a separate study for every package size?

Not necessarily. Bracketing and matrixing may reduce the number of studies for products with multiple package types, and GUI-0158 points to ICH Q1D as a useful model. Because NHPs are outside the scope of ICH guidelines, document a scientific rationale for the reduced design. Real time confirmation of the claimed shelf life is still expected.

How MFLRC can help

MF License and Regulatory Consultants supports natural health product licence holders and importers across the full stability lifecycle. That work usually begins with a gap assessment of the existing programme, product by product, to separate the files that are defensible from the ones that only look complete. From there we draft the stability programme procedure, design protocols per product family, define specifications and acceptance criteria, and structure reports so a QAP can approve them with confidence. Our quality assurance and quality control services cover the surrounding systems as well, including deviation and CAPA handling, out of specification investigations, retention sample programmes and record retention schedules.

For importers, we rebuild the stability portion of the importer file: confirming foreign protocols against Canadian requirements, identifying the tests that must be obtained locally, and writing the quality agreement clauses that make those obligations enforceable. Our audit services include supplier and mock inspection audits, and our regulatory affairs, licensing and import and export team handles site licence and product licence consequences when a shelf life changes. If you would like to test your wider readiness first, start with our 25 point Health Canada GMP self assessment.

If your stability files were built under the 2015 guidance and have not been revisited since March 2026, a focused review is worth the time. It is a contained piece of work, and it addresses one of the areas an inspector is most likely to open first.

Conclusion

Stability under GUI-0158 version 4.0 is not a testing problem. It is an evidence problem. The underlying legal requirement in section 52 has not changed, and neither has the science. What has changed is that Health Canada now sets out, in detail, what a defensible stability programme looks like: a written procedure, an approved protocol per study, real time data that confirms whatever the initial expiry date was based on, and a report that a QAP has signed because it actually supports a conclusion.

For most companies the remediation is smaller than it first appears, because the testing is usually already happening. The work is to wrap that testing in documents that make it provable, and to correct the two rules that quietly undermine otherwise good files: dating from packaging rather than manufacture, and adopting a supplier's shelf life without confirming it against Canadian requirements. If you want to place this alongside the rest of the guide, our 15 point GUI-0158 version 4.0 compliance checklist covers the remaining sections.

Sources and references

• Health Canada, Good manufacturing practices guide for natural health products (GUI-0158), version 4.0, published 4 September 2025, in force 4 March 2026
• Health Canada, GUI-0158 NHP GMP guidance, sections 52 to 62, covering stability period, records, lot or batch samples and recall reporting
• Government of Canada, Natural Health Products Regulations (SOR/2003-196)
• Health Canada, Quality of natural health products guide, for stability specifications by dosage form
• International Council for Harmonisation, ICH quality guidelines, including Q1D on bracketing and matrixing designs