UDCA Related Substances: How the EP Limits Actually Add Up
Share
The European Pharmacopoeia monograph for ursodeoxycholic acid sets three related-substances limits that matter, plus a fourth that most people never read. Impurity A is limited to 1.0%. Every unspecified impurity is limited to 0.10%. Total impurities are capped at 1.5%. The disregard limit, below which peaks are not counted at all, is 0.05%.
That is the whole framework. It fits in three lines, and yet it causes more confusion on the buyer's side than the impurity identities do. The confusion is structural rather than chemical: the three limits belong to three different categories of peak, and reading them as though they were a single list is where spec sheets and comparison tables go wrong.
Two specified impurities, not nine
The monograph lists nine impurity letters, A through I. That list is not a list of limits. It is a list of what the monograph is aware of, and only two of those letters carry a named limit of their own.
| Category | Letters | What it means |
|---|---|---|
| Specified impurities | A, C | Named and individually limited. These are the only two with their own numeric ceiling. |
| Other detectable impurities | B, D, E, F, G, H, I | Described in the monograph. If present at a sufficient level they would be detected by one of the tests, but they are controlled by the general unspecified-impurity criterion rather than by a limit of their own. |
Why A and C and not the others? Impurity A is chenodeoxycholic acid and impurity C is lithocholic acid, and both sit at the centre of how this molecule is made and judged. CDCA is the epimer at carbon 7 and the usual starting point for synthesis, so it is the impurity most likely to survive purification. Lithocholic acid is the one member of the family with a toxicity profile that regulators treat separately, which is why it gets a test of its own.
Impurity C actually sits outside the related-substances test altogether. It has a dedicated thin-layer chromatography method with its own limit of 0.1%, and the related-substances liquid chromatography test picks up A along with everything unspecified. If you are building a comparison table across suppliers, treat the C result and the related-substances result as two separate lines, because that is how the monograph defines them.
The unspecified pool, and the arithmetic trap inside it
Everything not named falls into one pool, and each member of that pool is allowed 0.10%. That single sentence creates the most common misreading of the whole panel, because 0.10% looks small until you ask how many such impurities can coexist.
Seven letters (B, D, E, F, G, H, I) plus any impurity the monograph has never described all share the same 0.10% individual ceiling. If seven unspecified impurities each sat at the ceiling, the sum would be 0.70%, which the 1.5% total cap would tolerate. If fifteen did, the total would breach the cap while every individual peak still passed its own limit. That is the trap: a certificate can be "compliant" line by line and still fail on the total.
In practice, well-controlled material does not look like that. The point is not that the arithmetic is likely to bite, it is that the total figure is the number that actually constrains the process, and it is the number a supplier's process capability shows up in. Two suppliers can both report "conforms" on every individual limit while one runs at 0.2% total and the other at 1.4%. Same certificate language, very different process control.
The disregard limit is the quiet one
Below 0.05%, peaks are disregarded outright. This matters for negotiation as much as for quality. If a supplier sends you a chromatogram and there are eleven small peaks visible, ask what the reporting floor is. A chromatogram reported down to 0.05% and one reported down to 0.01% will look dramatically different and mean exactly the same thing.
Also note the relationship between the numbers: the disregard limit is exactly half the unspecified limit, and one twentieth of the total cap. Those ratios are not accidents of drafting. They are set so that the counting floor sits comfortably below the limit anyone would act on.
The method is not the one people expect
The related-substances test in the UDCA monograph is liquid chromatography, and the detector is a refractometer, not a UV detector. That detail has real consequences.
| Parameter | Value |
|---|---|
| Stationary phase | End-capped octadecylsilyl silica gel, 5 µm, 0.25 m × 4.6 mm |
| Column temperature | 40 °C ± 1 °C |
| Mobile phase | Acetonitrile 30 : sodium dihydrogen phosphate 0.78 g/L adjusted to pH 3 with phosphoric acid 37 : methanol 40 |
| Flow rate | 0.8 mL/min |
| Detection | Refractometer at 35 °C ± 1 °C |
| Injection | 150 µL |
| Run time | Four times the retention time of ursodeoxycholic acid |
| Relative retention | Impurity H ≈ 0.9; impurity A ≈ 2.8 (UDCA retention about 14 min) |
| System suitability | Resolution at least 1.5 between impurity H and ursodeoxycholic acid |
Bile acids have no useful chromophore. They do not absorb in the ultraviolet in any way that gives a sensitive assay, which is why this method runs on refractive index detection. Refractive index detection is inherently less sensitive and less stable than UV, it is sensitive to temperature in a way UV is not, and it does not tolerate gradient elution. Hence the isocratic mobile phase, the tight column temperature, and the large 150 µL injection.
For a buyer, this explains something practical. A contract laboratory that quotes you a low price for "HPLC purity" on a UV system has not run this method. The number it returns is not comparable to a monograph related-substances result, no matter how convincing the chromatogram looks.
Where the numbers on a spec sheet come from
Older specifications still in circulation quote a different set: lithocholic acid 0.1%, chenodeoxycholic acid 1.0%, cholic acid 0.5%, any other impurity 0.25%, tested by thin-layer chromatography. Those figures are not wrong in the sense of being fabricated. They come from an earlier thin-layer version of the related-substances test, before the liquid chromatography method replaced it.
If you see cholic acid quoted with a 0.5% limit, or "any other impurity 0.25%", you are looking at a panel built on the superseded test. That is worth flagging for two reasons. First, the numbers describe a different analytical method and are not interchangeable with the current limits. Second, and more usefully, it tells you something about the age of the template a supplier is working from — a spec sheet is often the oldest document in a company's quality system.
Reading the related-substances block on a certificate
- Check that all three lines are present. Impurity A, unspecified (sometimes written "any unspecified impurity" or "each unspecified impurity"), and total. A block missing the total line is incomplete against the monograph.
- Check the total against the individual results. The total should be consistent with the sum of the peaks reported above the disregard limit. If it is not, ask why.
- Check the unspecified line is per-impurity, not a sum. "Unspecified 0.08%" against a 0.10% limit means the largest single unspecified peak is 0.08%. It does not mean everything unspecified adds up to 0.08%.
- Check the method reference. The certificate should point at the liquid chromatography test, not a thin-layer test, for the related-substances figures.
The impurity letters themselves — which molecule each letter refers to, and how the letters shift depending on which monograph you are reading — are set out in our A–I impurity reference. If you are cross-checking a certificate, that piece and this one are meant to be read together: the letters tell you what is being measured, and this framework tells you what the numbers are allowed to be.
Two other rows usually sit near the related-substances block on the same certificate. Rotation is covered in our note on the specific optical rotation identity test, and where the European and US monographs diverge on limits is covered in the EP vs USP comparison. Our ursodeoxycholic acid page lists what we ship against, and specifications for a specific quantity can be discussed through a quotation request.
Frequently asked questions
How many specified impurities does the UDCA monograph have?
Two: impurity A (chenodeoxycholic acid, limit 1.0%) and impurity C (lithocholic acid, limit 0.1%, tested separately by thin-layer chromatography). The other letters B, D, E, F, G, H and I are described but not individually limited.
What is the limit for unspecified impurities in UDCA?
0.10% for each unspecified impurity. That is a per-impurity ceiling, not a sum.
What is the total impurity limit?
1.5%. The total is the figure that constrains overall process control, because a large number of small individual peaks can each pass their own limit while the total exceeds the cap.
What is the disregard limit?
0.05%. Peaks below that area are not counted. A chromatogram reported to 0.05% shows fewer peaks than one reported to a lower floor, without any difference in the material.
Why does the method use a refractive index detector?
Bile acids have no useful ultraviolet chromophore, so the monograph specifies refractometric detection. That rules out gradient elution and makes temperature control critical, and it means a UV-based "purity" result is not an equivalent measurement.
Do older 0.5% and 0.25% limits still apply?
Those numbers belong to an earlier thin-layer chromatography version of the test, in which cholic acid carried a 0.5% limit and other impurities 0.25%. The current monograph controls related substances by liquid chromatography with the limits above. Confirm which version a specification is quoting.
The question that sorts suppliers faster than the certificate
Ask for the total, not the individual results. Anyone can report conformance on twelve separate lines. The total tells you what the process can actually hold, and it is the number that predicts whether the third shipment will look like the first.
The primary texts behind this piece are the European Pharmacopoeia, the United States Pharmacopeia and the ICH quality guidelines. Check the current edition before quoting any figure from them.
This article describes the limits and method in the pharmacopoeial monograph for ursodeoxycholic acid, for quality, regulatory and procurement professionals. It states no therapeutic use, dose or indication for any substance, and no order quantity or lead time. Confirm all limits and method references against the current edition of the pharmacopoeia you are working to.