Zone IVb Stability: Why 30 °C/75% RH Changes the Answer
Share
A purchase order goes to a customer in Southeast Asia. The specification is agreed, the price is agreed, and the stability section of the supplier's file says 25 °C/60% RH, twenty-four months. Nobody in the thread notices that the destination warehouse runs at 31 °C for eight months of the year and that the relative humidity rarely drops below 75 per cent.
The file is not wrong. It is answering a question about a different market. That is the ordinary failure mode in stability documentation: not fabrication, but a study designed for the supplier's home climate and then offered as if climate were a detail.
What a zone actually defines
The ICH stability framework divides the world into climatic zones by measured temperature and humidity, and assigns a long-term storage condition to each. The condition is not a recommendation for how to store the product. It is the condition under which the shelf life must be demonstrated if the product is to be sold there.
| Zone | Long-term condition | Where it usually applies |
|---|---|---|
| I | 21 °C/45% RH | Temperate, cool |
| II | 25 °C/60% RH | Temperate, including much of Europe |
| III | 30 °C/35% RH | Hot, dry |
| IVa | 30 °C/65% RH | Hot, humid |
| IVb | 30 °C/75% RH | Hot, very humid |
Accelerated testing sits at 40 °C/75% RH for six months in the humid zones. The purpose of the accelerated arm is not to assign a shelf life. It is to see what fails first, and to give the long-term arm something to be interpreted against.
Five degrees and fifteen points of humidity
The difference between 25/60 and 30/75 looks small written down. In a warehouse it is the difference between a material that behaves like a solid and one that behaves like a solid trying to become a solution.
For a bile acid, three things move with humidity before anything moves with heat.
Water content. Loss on drying is a specification line, and it is the first to drift. Material at 0.4 per cent can reach 0.7 per cent in a fibre drum in a wet season without any chemical change at all. If the limit is 1.0 per cent, that is half the allowance spent on storage.
Flow and compaction. Moisture on the surface of a powder raises inter-particle friction and then, under pallet weight, produces caking. A drum that pours freely in Milan may need a mallet in Jakarta, and the customer will describe this as a quality defect because from their side it is one.
Colour and appearance. Slow oxidation at elevated temperature darkens a white powder at the margins. There is usually no numeric limit for colour, which is exactly why it becomes a complaint rather than a deviation.
Chemical degradation is the slowest of the four. Assay and impurity profile often hold within limits at twelve months even at 30/75, while the physical attributes have already failed the customer's expectations. This is the part that a specification-only review misses, because the specification has no line for "pours freely".
An illustrative dataset
Values below are illustrative and chosen to show the pattern, not to describe a batch.
| Attribute | Initial | 12 months at 25/60 | 12 months at 30/75 |
|---|---|---|---|
| Assay, dried basis | 99.8% | 99.7% | 99.5% |
| Total impurities | 0.20% | 0.22% | 0.29% |
| Loss on drying | 0.40% | 0.45% | 0.68% |
| Bulk density | 0.42 g/mL | 0.42 g/mL | 0.46 g/mL |
| Appearance | White | White | Off-white at surface |
The assay column is comfortable in both conditions. The loss on drying and bulk density columns are where the two climates part company, and the appearance line is where a customer will file a complaint that the numbers do not support.
What the pack does to the number
A stability condition describes the air outside the pack. What the material experiences is the air inside it, and the two are connected by a barrier whose quality varies more than most buyers assume.
A sealed aluminium laminate with a desiccant brings the internal humidity close to the fill humidity and holds it there. A fibre drum with a loose polyethylene liner equilibrates over weeks. A steel drum with a gasket sits somewhere between, and its performance depends on whether the gasket was compressed correctly at closing, which is a person-level variable rather than a design variable.
This is why a Zone IVb study run in one pack does not transfer to another, and why the pack description belongs in the stability section. A customer who receives the same material in a different container has received a different stability claim, whether or not anybody said so.
The extrapolation trap
There is a persistent belief that a 25/60 dataset can be pushed to 30/75 by allowing a shorter shelf life. The arithmetic feels reasonable and it is not supported.
Extrapolation under ICH Q1E uses the slope of the change at the condition in question and its statistical confidence. It assumes the degradation mechanism stays the same. Raising both temperature and humidity can change the mechanism, not merely accelerate it: a hydrolysis pathway that is negligible at 60 per cent humidity becomes the dominant one at 75. Extrapolating across that boundary produces a number with no evidential basis, and regulators have seen the argument often enough to reject it on sight.
The correct route is a study at the destination condition. For a supplier without one, the correct answer to the customer is that the data does not cover their market, followed by a plan and a date.
What to put in the enquiry
Four lines settle most of this before a sample ships: destination city and warehouse type, the climatic zone if the customer's regulatory team has already assigned one, the pack you intend to receive, and the retest period you need. The fourth is the one buyers forget, and it is the one that determines whether the study is adequate rather than merely present.
If the answer comes back with a condition that does not match the destination, that is not necessarily a refusal. It may be an offer of a different pack. The useful follow-up is to ask which pack was studied, because a change of container is a legitimate way to serve a hotter market and an unstudied container is not.
Where this leaves a supply-chain partner
We state the condition a dataset was generated at, and we say which pack it describes. Where a customer's destination sits in Zone IVb and the available data was generated at 25/60, we say so before the order rather than after the arrival, because the second conversation is a complaint and the first is a planning decision.
Climate is not a detail that can be averaged. It is the condition under which the customer's own shelf life has to hold, and a supplier who treats it as a line in a template has quietly moved that risk to the buyer.
Accelerated conditions are not a shelf life
Six months at 40 °C/75% RH is a stress condition, not a storage condition. Its job is to expose failure modes early: to show which attribute moves first, whether the pack holds, and whether the analytical method can still measure the material when it has begun to change. A product that survives six months of accelerated testing has demonstrated something about robustness. It has not demonstrated a thirty-six month shelf life, and a supplier who presents it as one has confused two different arms of the same study.
The shelf life comes from the long-term arm, at the condition of the destination market, with enough time points to describe a slope. Accelerated data supports the interpretation and justifies the extrapolation, which is why a file with a strong accelerated arm and a thin long-term arm is not stronger than one with the reverse. It is weaker in the place that matters.
The ICH Q1A(R2) text sets out the data package, and the WHO guidance on stability testing in member states is a useful cross-check for markets that apply the zone framework with local amendments. The European Medicines Agency publishes the questions its assessors actually ask about extrapolation, which is a faster route to the point than reading the guideline twice.
What the receiving warehouse needs to know
A stability condition is written for a file. A storage statement is written for a building. The two are related and they are not identical, and the gap between them is where most receiving problems start.
Three practical questions decide whether a warehouse can hold the material as labelled. Is the space air-conditioned, and if so to what set point and with what tolerance? Can the container be opened and resealed without destroying the barrier, or must it be consumed in one campaign? And is there a designated area for material awaiting quality release, because a drum that sits on a loading dock for two days in July has experienced a condition nobody wrote down.
For a destination in Zone IVb, the honest storage statement is often narrower than the label suggests. A label that reads 30 °C/75% RH describes the condition the data supports, not a licence to store the drum next to a roller door. Customers who understand this ask for the pack description and the storage statement together, which is also the pair that makes the packaging discussion concrete. Where the destination is a Zone II market and the material is then re-exported, the photostability and retention questions arrive on the same email.
Frequently asked questions
Which countries fall in Zone IVb?
The zone is defined by measured climatic data, not by latitude. Large parts of South and Southeast Asia, parts of Africa, Central America and northern South America sit in IVb, and so do some regions inside countries usually described as temperate. The zone follows the warehouse, not the border, which is why the question should be asked about the destination warehouse rather than the destination country.
What are the long-term conditions for each zone?
Zone I and II use 25 °C/60% RH as the long-term condition, with 30 °C/65% RH accepted as an alternative in some frameworks. Zone III uses 30 °C/35% RH. Zone IVa uses 30 °C/65% RH and Zone IVb uses 30 °C/75% RH. Accelerated testing is normally 40 °C/75% RH for six months, with the humidity adjusted for the drier zones.
Can a 25 °C/60% RH study be extrapolated to 30 °C/75% RH?
Not by arithmetic. Extrapolation under ICH Q1E works from data at the same condition, using the slope and its confidence interval. Moving to a warmer, wetter condition changes the mechanism as well as the rate, so a dataset at 25/60 cannot support a shelf life at 30/75 unless the higher condition was also run.
Does the pack change the answer?
It changes it more than the climate does. A sealed aluminium laminate keeps the headspace humidity close to the fill humidity regardless of the room, while a fibre drum with a polyethylene liner equilibrates slowly with the outside air. The stability condition should therefore be read together with the pack, and a study in one pack does not cover another.
How much data is needed before the first shipment?
Regulatory expectations are typically twelve months of long-term data at submission, with six months of accelerated data. For a commercial shipment ahead of a filing, the practical minimum is enough data to cover the retest period you intend to assign, and the honest position is to state the period the data supports rather than the period the customer would like.
Our product is stored in an air-conditioned warehouse. Does the zone still apply?
Then the zone describes the transport and the destination, not your warehouse. The material will still cross a dock, sit in a container and wait in a customs shed. If you want the label to say 25 °C, the whole chain has to be controlled to 25 °C, and one uncontrolled leg is enough to make the label untrue.