Which is Safer – Phenoxyethanol or Water?

I would like to start this discussion by introducing the concept of “margin of safety” (MoS). This is related to the concept of “the dose makes the poison”, but it allows a numerical value to be assigned, which places the risk in a very clear context and permits a very clear picture of the true dangers in use of an ingredient.

The European Commission’s independent scientific advisory panel, the Scientific Committee for Consumer Safety (SCCS), calculate the MoS by taking the dosage figure that produces the lowest no observed adverse effect (NOAEL); that is the lowest out of all the studies available for assessment. In other words, the figures from the study that resulted in the lowest dosage required to produce an effect – a worst-case scenario. They then calculate the likely exposure to the ingredient in question (this calculation will vary, depending upon the use pattern of the ingredient). Both the NOAEL and the calculated exposure are expressed in terms of weight per unit bodyweight (“bw”)per day, usually mg/kg bodyweight/day. The exposure figure is then divided by the NOAEL figure to produce the Margin of Safety.  This makes perfect sense, because if the exposure was, for example, 100mg/kg bw/day and the NOAEL was also 100mg/kg bw/day, 100 divided by 100 = zero, and there is zero margin of safety. (Having said that, because the NO observed effect dose is used, there IS actually some margin of safety, but this will depend on the dose interval between the NOAEL and the lowest observed adverse effect level (LOAEL), which is sometimes quite large. However, for the purposes of this discussion, I will keep it simple and use the NOAEL – still with me on this?) The SCCS have a  minimum MoS of 100 as a target when determining any maximum permitted concentration, should the ingredient require any form of restriction for use in cosmetics. In other words, if the NOAEL was 1000mg/kg bw/day, the SCCS would assign a maximum permitted concentration that would result in an actual exposure of no more than 10mg/kg bw/day. The SCCS (in a previous incarnation) assigned a maximum permitted concentration of 1% to phenoxyethanol on this basis.

I realise that this all sounds rather convoluted, but it DOES make sense, and it is necessary to appreciate the significance of this in order to understand why this sort of information is both relevant and extremely important.

I was prompted to write this article by a posting on a new Facebook page (Chemical Of The Day – an offshoot of the blog of the same name, both run by Stephanie Greenwood, owner of Bubble & Bee Organic  –. One of the posts was a chart showing 7 ingredients, each with a serious warning message under the main heading “Top 7 Cosmetic Estrogen Mimickers”. Phenoxyethanol was proclaimed to be responsible for serious adverse effects (“produced significant  reproductive and developmental toxicity in studies in mammals”). On the positive side, Stephanie did include a reference to the actual study  She did, however, spectacularly fail to place the results in a proper context. Which is where I come in . . . . . . .

The study in question used different levels of phenoxyethanol in the food employed to introduce the material to the test subjects:

0         ( blank control)

0.25% (equivalent to approx 375mg/kg bw/day)

1.25% (equivalent to approx 1875mg/kg bw/day)

2.5% (equivalent to approx 3700mg/kg bw/day)

The results showed that the lowest test concentration produced no observed adverse effects and, therefore the NOAEL figure for phenoxyethanol in this study was 375mg/kg bw/day. (The LOAEL was 1875mg/kg bw/day.)

In order to calculate a Margin of Safety from these results, we need to determine the exposure to phenoxyethanol (in cosmetics). The figure used by the SCCS for daily cosmetic use is approx 17g (the last time I checked), so I will take a higher figure of 20g to make life easier in terms of calculation, but more difficult in terms of exaggerating the true exposure. Moreover, I will assume that the 20g of cosmetic products ALL contain phenoxyethanol at the  maximum permitted concentration of 1% (this is by no means the case, and is further exaggerating the actual exposure).

20g of cosmetic product containing 1% phenoxyethanol = 200mg phenoxyethanol

For a 60kg body, this equates to an exposure of 3.3mg/kg bw/day (200/60 = 3.3)

(This also assumes that all the phenoxyethanol is actually absorbed through the skin – yet another exaggeration of the true exposure scenario)

So, to calculate the MoS based on this study, we divide the NOAEL figure (375) by the exposure figure (3.3), which gives a Margin of Safety of 125 – well above the SCCS target of 100.

Furthermore, we should consider the fact that there is a significant dose interval between the one that produced the NOAEL (375) and that which produced the LOAEL (1875), and the TRUE NOAEL will be somewhere between these two figures. Calculating the MoS based on the LOAEL gives a figure of 625, so the true MoS is somewhere between 125 and 625. So, even ignoring the exposure exaggerations introduced into the calculation, there is a significantly higher MoS for phenoxyethanol than is required by the SCCS.

Comparison with water:

Water intoxication can occur when the intake is significantly greater than the ability of the kidneys to excrete. Kidneys can excrete approximately 0.8 – 1.0 kg/hour. (1 litre = 1kg) Assuming that a toxic dose would be achieved by drinking 2kg/hour, and that at least a 12-hour exposure would be required, this equates to 24kg of water/day. (The actual dose that would lead to death is almost certainly lower than this!) The referenced daily intake per day is 3.7 kg per day for adult human males, and 2.7 kg/day for adult human females, including water contained in food, beverages, and drinking water.

Using the same calculation as for phenoxyethanol, the MoS for water (in adult males) is 24/3.7 = 6.5; for adult females, it is 24/2.7 = 8.9. Please note that I have not used a NOAEL figure here but, in effect, the LOAEL, which is higher so, again, I have skewed the figures away from the argument I am trying to make but even so, by this measure, phenoxyethanol is well over 10 times safer than water! Water intoxication can, and has, lead to death – there are no known examples of death due to phenoxyethanol exposure. It’s all down to EXPOSURE!

So, despite going beyond the worst-case scenario at every stage of the calculations, the actual MoS for exposure to water is significantly less than that for phenoxyethanol. I can, therefore, conclude that phenoxyethanol is safer than water when used as intended (and NOT when used as a drink!)

Let’s remember the Golden Rule:


(Neither hazard nor exposure alone are sufficient to inform anyone of the risk involved in using any substance; you need both to do a proper risk assessment – the ONLY way to measure safety.)

I will now remind you that the inspiration for this exercise came from the public posting of claims that phenoxyethanol causes birth defects and reproductive issues. This is clearly extremely misleading information and hugely distorted out of any true context. I have chosen only phenoxyethanol out of the 7 ingredients targeted in that post, but there are similar issues with several of the other ingredients included, as well as other posts on that Facebook page. I will not even dwell too long on the fact that the post proclaimed “7 Cosmetic Estrogen Mimickers” and phenoxyethanol has NEVER been identified as an oestrogen mimic (the only study I can find on this subject failed to identify any oestrogenic activity for the substance – Morohoshi et al, 2005

I will leave you to ponder the morality of making such claims in the light of the information I have shared here.



NOAEL – No Observed Adverse Effect Level – the lowest concentration of test substance at which no adverse effects can be detected

LOAEL – Lowest Observed Adverse Effect Level – the lowest concentration of test substance at which any adverse effects were detected

MoS – Margin of Safety – the degree of separation between the lowest exposure level that causes no adverse effects and the actual exposure to the substance


I should emphasise that I have carried out the above assessment on the basis of a comparison using results on phenoxyethanol in the study that was cited in the post on the Facebook page in question. This was done in order to prove a specific point. In fact, the lowest NOAEL figure that is available for phenoxyethanol is 164mg/kg bw/day, a figure which, if substituted into my earlier calculation, STILL gives a greater margin of safety compared with water within the context described.

Related posts

Leave a Comment