Cosmetics testing basics: safety, stability and claims

A cosmetic is any product applied to the outer parts of the body, teeth or mouth to clean, perfume, protect, change appearance or keep them in good condition. Unlike medicines, most cosmetics do not need a licence before sale, but the company that places them on the market is fully responsible for their safety and for every claim on the pack.

Laboratory testing supports three questions. Is the product safe (low microbial load, effective preservation, no harmful impurities)? Will it stay as intended until its expiry or period after opening (stability and packaging compatibility)? And can every claim be backed up (SPF, UVA protection, 'paraben-free', 'long-lasting')?

In the EU every product needs a product information file with a cosmetic product safety report. Much of the evidence in that report comes from the QC and development lab.

Cosmetic rules vary by country, but most labs work with a common set of documents. Regulations say what is allowed; ISO standards say how to test and how to manufacture well.

Every batch is checked against a specification before release. The tests are simple but must be done in exactly the same way each time, because many results depend on conditions such as temperature, spindle or probe.

• Appearance, colour and odour against an approved standard sample.
• pH — neat for fluids or on a 10% dispersion for thick creams; always state which.
• Viscosity — rotational viscometer at fixed spindle, speed, temperature and reading time.
• Texture — firmness, spreadability and stickiness on a texture analyser.
• Density, fill weight and, for some products, specific tests such as foam height (shampoos) or break strength (lipsticks).
• Assay of key ingredients: preservatives (for example parabens or phenoxyethanol by HPLC), active ingredients and UV filters.

A pH or viscosity result without its conditions (dispersion ratio, spindle, speed, temperature) cannot be compared with anything. Record them every time.

Cosmetics are not sterile, but they must have a low microbial load and stay protected when consumers dip their fingers in. Two different questions are asked. The microbial count (ISO 21149 for bacteria, ISO 16212 for yeasts and moulds) asks: how clean is this batch now? The preservative efficacy test (ISO 11930) asks: can this formula cope with contamination during use?

Both rely on neutralising the preservative during counting; otherwise the preservative keeps working on the agar plate and the count is falsely low. Method suitability is therefore shown for each product type before routine testing.

These are summaries of commonly used values; always work from the current standards and your market's rules. Microbiology must be done by trained staff in a suitable laboratory.

Lead, arsenic, mercury and cadmium can enter products as impurities in mineral pigments, talc and clays. In the EU they are prohibited substances, and traces are accepted only when technically unavoidable and safe. Other regulators publish guidance levels; for example Health Canada's and the US FDA's levels for lead are commonly cited at about 10 mg/kg.

Labs usually digest samples in closed microwave vessels and measure the four elements together by ICP-MS. Mineral-based powders may need hydrofluoric acid to dissolve silicates fully, so reports should say whether results are total or acid-extractable. Mercury in imported skin-lightening creams remains a known problem that market-surveillance labs test for.

There is no single compulsory stability protocol for cosmetics. Each company designs and justifies its own, usually following industry guidance. Samples in the final pack are stored under several conditions and checked at set intervals.

• Checks at each pull: appearance, colour, odour, pH, viscosity, texture, microscopy of emulsions, weight loss and, where relevant, assay and microbiology.
• A common rule of thumb is that 3 months at 40 °C without change suggests about 2 years at room temperature — a guide only, to be confirmed by real-time data.

SPF tells consumers how well a sunscreen protects against sunburn, mainly caused by UVB. It has traditionally been measured in vivo (ISO 24444). ISO 23675 now describes an in vitro SPF method, and in vitro tests are widely used for screening and batch consistency.

UVA protection is measured in vitro to ISO 24443: the product is spread on roughened PMMA plates, UV transmittance is measured from 290 to 400 nm, the plate is irradiated to account for photostability, and UVA-PF and critical wavelength are calculated. In the EU, UVA-PF should be at least one-third of the labelled SPF, and a critical wavelength of at least 370 nm is commonly required for broad-spectrum or UVA claims.

1. 1Verify the SPF analyser with wavelength and transmittance filters.
2. 2Apply the product to plates at the exact dose and spread it consistently.
3. 3Scan several spots on at least three plates.
4. 4Run a reference sunscreen and check it is in range.
5. 5Calculate SPF, UVA-PF, ratio and critical wavelength and review plate-to-plate variability.

ISO 22716 expects the same basics as any quality control lab: approved specifications and methods, calibrated and maintained instruments, trained staff, retained samples and complete records. Many cosmetics labs are also accredited to ISO/IEC 17025.

• Keep a retained sample of every batch for at least the shelf life.
• Record the conditions of empirical tests (pH dispersion, viscosity spindle and speed, texture probe and speed).
• Investigate out-of-specification results before retesting, and document the outcome.
• Control raw materials as well as finished products: certificates of analysis, identity tests and impurity checks for high-risk materials such as mineral pigments.