The oils you pay for by the milligram
Borage, evening primrose and blackcurrant seed oil sit at the expensive end of the price list, and the reason they are there is one fatty acid: gamma-linolenic acid, GLA. This article evaluates the three oils primarily by their GLA content, their cost and the oxidation control they demand, which is how our formulators look at them. The questions are: what am I paying for, how much of it is in this drum, what does the evidence actually show, and how do I keep it from oxidising before the customer opens the jar. This article, a spoke of our series on plant oils by fatty-acid group, takes those four questions in order.
Two acids with the same name and different chemistry
Gamma-linolenic acid and alpha-linolenic acid are both C18:3: eighteen carbons, three cis double bonds. They are not the same molecule and they do not behave the same way. In alpha-linolenic acid the double bonds sit at carbons 9, 12 and 15, counting from the acid end, which places the last one three carbons from the methyl end: it is an omega-3 acid, the one found in rosehip and chia. In gamma-linolenic acid the double bonds sit at carbons 6, 9 and 12, six carbons from the methyl end: an omega-6 acid, made in the body from linoleic acid by the delta-6-desaturase enzyme, and the reason the GLA oils are marketed the way they are. What the two share is the part that matters for a stability study: three double bonds and two bis-allylic carbons, the carbons flanked by double bonds on both sides, which are particularly susceptible to losing the hydrogen atom that begins lipid oxidation. On that count a GLA chain is as vulnerable as an alpha-linolenic one, whatever the omega label says; the stability of the whole oil also depends on everything else in it.
How much GLA is actually in the drum
The three oils differ more in GLA content than their marketing suggests, and the content decides both the price per unit of the active and how much of the oil a formula needs.
- Borage seed oil is usually the richest. A 2016 review in the European Journal of Pharmacology gives about 21 percent GLA as a typical figure, and the borage oil used in a 2004 feeding study in the Journal of Agricultural and Food Chemistry contained 22 percent by weight.
- Blackcurrant seed oil is the variable one. The same review gives about 17 percent as typical. A 2002 survey of 36 blackcurrant genotypes bred in Scotland, also in the Journal of Agricultural and Food Chemistry, found GLA generally at 11–19 percent of total fatty acids, with three genotypes at 22–24 percent, levels the authors describe as similar to borage seed; so a given blackcurrant oil can match or exceed a given borage oil. The same seeds carried 10–19 percent alpha-linolenic acid and 2–4 percent stearidonic acid, so blackcurrant seed oil also brings omega-3 chains, with their own two and three bis-allylic carbons respectively.
- Evening primrose oil has the least. A 2023 review in Food and Function gives 8–10 percent GLA alongside 70–74 percent linoleic acid; a 1992 analysis of sixteen commercial evening primrose oil capsule brands in Lipids found a mean of 8.7 percent GLA (range 1.9–10.5), with fourteen brands at 7–10 percent and generally good agreement between the manufacturers' claims and the measured values. By composition, then, roughly three-quarters of evening primrose oil is linoleic acid and under a tenth is GLA.
The arithmetic follows directly, with a caveat. Using the 22 percent borage sample and the 8.7 percent evening primrose mean from the studies above, a gram of borage oil carries roughly two and a half times the GLA of a gram of evening primrose oil, so a formula that "contains evening primrose oil" at the same percentage as a competitor's borage delivers well under half the GLA. That is a cross-study comparison, not a description of any particular batch: the figure to compare prices on is the GLA percentage on the certificate of analysis of the oil you are actually offered, and for these oils it should be there. Composition tells you how much GLA is delivered at a given inclusion level; it does not tell you what that GLA does, or how much a product needs.
What the evidence says, and about whom
The GLA oils have a research history on skin, and it is worth being precise about it, because the two studies below were done in people with a medical condition, atopic dermatitis, and ingredient and vehicle studies of that kind do not substantiate a finished cosmetic automatically.
The study most useful to a formulator is Gehring and colleagues (Arzneimittelforschung, 1999). They applied evening primrose oil topically to atopic subjects in two vehicle-controlled cohorts of twenty each, one testing an amphiphilic emulsion and one a stable water-in-oil emulsion, for four weeks with a one-week treatment-free period, and measured transepidermal water loss and stratum corneum hydration, with a barrier challenge using sodium lauryl sulphate and nicotinic acid ester. Evening primrose oil had a stabilising effect on the stratum corneum barrier with the water-in-oil emulsion tested; with the amphiphilic emulsion tested the effect did not appear. The authors called the choice of vehicle extremely important. That is the formulator's lesson, within the limits of the two vehicles they tested: the base can decide whether the oil shows an effect at all.
Kanehara and colleagues (Journal of Dermatology, 2007) took a different route: thirty-two children with mild atopic dermatitis wore either undershirts coated with borage oil or uncoated undershirts daily for two weeks in a double-blind, placebo-controlled design. The borage group showed statistically significant improvements in erythema and itch and a decrease in transepidermal water loss from the back; the placebo group showed no significant change. It is a small, short study of a textile rather than a cream, and it is quoted here as research context. A 2017 review in the International Journal of Molecular Sciences covering the topical use of nineteen plant oils, borage among them, organises this literature under headings that include anti-inflammatory activity and barrier repair; a review is a map of the evidence, not additional evidence.
What can a UK brand do with this? Less than it would like to. Eczema and atopic dermatitis are medical conditions; a product presented as preventing or treating them may be regulated as a medicine rather than a cosmetic, depending on how it is presented and what mode of action is claimed, and the advertising rules require any efficacy claim to be substantiated on the product as sold. Our formulators' judgement, which is all the research supports, is this: GLA-rich oils are a reasonable choice for a leave-on product positioned for dry, sensitive-feeling skin; the vehicle deserves as much attention as the oil; and whatever the label says has to be shown on the finished product.
Why the percentage matters more than the name
Two things follow from the composition data and the Gehring result together. First, because the oils can differ two-fold or more in GLA, "contains borage oil" and "contains evening primrose oil" at the same percentage are different products, and a specification that names the oil without naming the GLA content has not specified the ingredient that matters. Second, because in the one vehicle-controlled study the effect appeared with one base and not the other, the percentage of GLA oil is not the only variable a formulator controls; the base is one too. In our formulators' experience these oils are used at low percentages of the oil phase, partly for cost and partly for the stability reasons below; the right percentage for a given product is a stability and cost decision that the finished-product data have to justify, and there is no sourced figure we can give you for it.
Stability: treat them like omega-3 oils
Three double bonds and two bis-allylic carbons put GLA among the more oxidation-prone fatty acids in common use, alongside alpha-linolenic acid; stearidonic acid, with four double bonds and three bis-allylic carbons, is more vulnerable still. Chapman, Kim and Min (Journal of Food Science, 2009) followed the loss of oleic, linoleic and alpha-linolenic acids in a mixture to which 10,000 ppm of oxidised alpha-tocopherol had been added, stored in the dark at 55 °C, and reported an oxidation ratio of 1 to 12 to 26 after thirty days. That experiment concerned alpha-linolenic acid, not GLA, under one set of conditions; GLA shares the bis-allylic count, so the same direction is expected, but it was not measured there. Whole oils are another matter again: their stability depends on the whole fatty-acid profile and on the minor components, which is why the batch certificate and the finished-product study, not the omega label, are what count.
Our formulators' processing practice for these oils follows from that chemistry; it is our practice, to be validated on each product against batch, process, finished-product and final-pack data:
- Add them last and cool. The GLA oil goes into the cool-down phase of an emulsion, after the heat is off, not into the oil phase that is held at emulsification temperature.
- Treat them as an addition to a base, not as the base. In our formulas the oleic and saturated groups carry the oil phase; the GLA oil rides in it at the percentage the product needs.
- Plan the antioxidant and the pack with the most oxidation-prone oil in the phase in mind. The stability study, run on the finished product in the final pack, is what supports the period-after-opening. The oxidative stability guide in this series covers the measurements.
Reading a certificate of analysis for a GLA oil
These oils are not among the named oils of Codex standard CXS 210, so there is no public composition range to check a certificate against; the supplier's specification is the reference for composition. For quality, the Codex standard for edible fats and oils not covered by individual standards (CXS 19-1981) still applies to them as foods, and its appendix gives the same quality maximums as the named-oil standard: peroxide value up to 15 milliequivalents of active oxygen per kilogram for virgin and cold-pressed oils and 10 for other oils, acid value 0.6 mg KOH per gram for refined and 4.0 for virgin and cold-pressed. Those are food-grade limits, not cosmetic acceptance specifications, but they are the only public yardstick there is.
- GLA percentage, on the certificate, per batch. The values observed in the studies cited above, not acceptance specifications, were about 21–22 percent for borage, 7–10 percent for most evening primrose oils (mean 8.7) and 11–19 percent for most blackcurrant genotypes, with some at 22–24. A certificate far outside those observations is a question for the supplier before it is a formulation input.
- Alpha-linolenic and stearidonic acid for blackcurrant. They tell you how much omega-3 chemistry you are also buying, and therefore how fragile the oil is.
- Peroxide value, per batch. For an oil this prone to oxidation the batch value at delivery is, in our procurement practice, the number we weigh most; the CXS 19 maximums above are the public yardstick, and the trend across batches from the same supplier tells you more than any single figure.
- Refined or cold-pressed. The 2023 review of evening primrose oil lists triterpenes, phenolic acids, tocopherols and phytosterols among its minor components; how much of them a given refining process leaves behind is a question for the supplier, and one worth asking.
- Storage and age. In our procurement practice we ask how the oil was stored and when it was pressed, because time and warmth before delivery are oxidation time the product never gets back.
When the premium is justified, and when it is not
| Situation | Our formulators' judgement | Why we think so |
|---|---|---|
| Leave-on product positioned for dry, sensitive-feeling skin, in a water-in-oil or rich occlusive base, with a stability programme | The premium can be worth paying | The one vehicle-controlled study of topical evening primrose oil found an effect with the water-in-oil emulsion it tested (Gehring, 1999); it tested atopic subjects, not ordinary cosmetic users, so the finished product still has to prove itself |
| Light lotion or gel-cream where the oil phase is small and the product is sold in a clear pack | Usually not, in our judgement | Three double bonds in a small oil phase exposed to light and air; the same study found no effect with the amphiphilic emulsion it tested (it did not test gel-creams or clear packs) |
| Rinse-off products | Not, in our judgement | Short contact time for an expensive, oxidation-prone ingredient; none of the cited studies tested rinse-off use |
| A "contains evening primrose oil" claim at a token percentage | Not, in our judgement | At 8–10 percent GLA, a token percentage of the oil delivers a token amount of GLA while the oxidation risk is real |
| Anything presented for eczema, atopic skin or a named condition | Do not make prevention or treatment claims for a cosmetic | Such presentation may make the product a medicine under UK rules; where the presentation is borderline, obtain MHRA advice |
The GLA oils are the clearest case in this series of an oil chosen for a single fatty acid and priced by it. Read the GLA percentage on the certificate, respect the three double bonds in the process and the pack, choose the vehicle with the Gehring result in mind, and test the finished product for whatever you intend to say about it. That is what the premium buys; without those steps it buys a line on the label.
Sources
- Codex Alimentarius. Standard for Edible Fats and Oils not Covered by Individual Standards, CXS 19-1981 (amended 2015): Appendix, quality characteristics. FAO PDF
- Sergeant S., Rahbar E., Chilton F. H. Gamma-linolenic acid, dihomo-gamma-linolenic acid, eicosanoids and inflammatory processes. European Journal of Pharmacology, 2016. PMC4975646
- Effect of long-term dietary supplementation of high-gamma-linolenic canola oil versus borage oil on growth, hematology, serum biochemistry, and N-6 fatty acid metabolism in rats. Journal of Agricultural and Food Chemistry, 52(12), 3960–3966, 2004. PubMed 15186123
- Ruiz del Castillo M. L., Dobson G., Brennan R., Gordon S. Genotypic variation in fatty acid content of blackcurrant seeds. Journal of Agricultural and Food Chemistry, 50(2), 332–335, 2002. PubMed 11782203
- Farag M. A., Reda A., Nabil M., Elimam D. M., Zayed A. Evening primrose oil: a comprehensive review of its bioactives, extraction, analysis, oil quality, therapeutic merits, and safety. Food and Function, 14(18), 8049–8070, 2023. PubMed 37614101
- Gibson R. A., Lines D. R., Neumann M. A. Gamma linolenic acid (GLA) content of encapsulated evening primrose oil products. Lipids, 27(1), 82–84, 1992. PubMed 1318991
- Gehring W., Bopp R., Rippke F., Gloor M. Effect of topically applied evening primrose oil on epidermal barrier function in atopic dermatitis as a function of vehicle. Arzneimittelforschung, 49(7), 635–642, 1999. PubMed 10442214
- Kanehara S., Ohtani T., Uede K., Furukawa F. Clinical effects of undershirts coated with borage oil on children with atopic dermatitis: a double-blind, placebo-controlled clinical trial. Journal of Dermatology, 34(12), 811–815, 2007. PubMed 18078406
- Lin T. K., Zhong L., Santiago J. L. Anti-Inflammatory and Skin Barrier Repair Effects of Topical Application of Some Plant Oils. International Journal of Molecular Sciences, 19(1), 70, 2017. PubMed 29280987
- Chapman T. M., Kim H. J., Min D. B. Prooxidant activity of oxidized alpha-tocopherol in vegetable oils. Journal of Food Science, 74(7), C536–C542, 2009. PubMed 19895457
