The 1 in 14 that was counted in the dermatology clinic
Showing the misleading chart
A skincare newsletter we drew says nearly 1 in 14 Europeans react to the commonest mix of perfume ingredients, citing 6.9% positive across a 12-country patch-testing network. Those 6.9% are dermatology patients, sent for patch testing because a rash looked like an allergy. When a population study patch tested a random sample of adults over overlapping years, the same test, done the same way, came back positive in 2.6%.
01The claim
Nearly 1 in 14 Europeans is allergic to the commonest perfume ingredients, so anyone choosing a skincare product should treat fragrance as a likely problem.
02The trick
Every bar is ESSCA’s published figure, and the axis, the sample sizes and the fine print are all in order. What changed is one word: the source counts patients and the headline says Europeans. ESSCA pools results from dermatology departments that patch test their patients one after another, and a person reaches that chair because a rash made a doctor suspect a contact allergy. That is a sample chosen by the very thing being measured, so its rate describes people with suspected allergy, not the population they came from, and 51,000 of them can’t fix that, because the size of a sample doesn’t change who is in it. The paper itself notes that referral patterns, along with test technique and exposure, probably explain why its rates vary between regions. (Clear Label and its advice are our invention, drawn to show the mechanism; every figure is ESSCA’s.)
03The fix
The redraw puts each fragrance allergen on one axis twice: once for the clinic patients and once for the EDEN Fragrance Study, which interviewed 12,377 adults aged 18–74 from random population samples in five countries between 2008 and 2011 and patch tested a random 3,119 of them, whether or not they had ever had a rash. Tested the way most of the clinics test it (Finn Chamber, in petrolatum), fragrance mix I falls from 6.9% to 2.6%, from nearly 1 in 14 to about 1 in 38; fragrance mix II halves, from 3.8% to 1.9%; Lyral barely moves, from 1.7% to 1.4%. The test itself matters too, which is why the comparison has to be like for like: in the same 3,119 people, fragrance mix I came out anywhere from 0.7% to 2.6% depending on the test batch and patch system, and EDEN’s age-standardised headline figure is 1.8%. Even the top of that range is well under the clinics’ 6.9%. The two studies also differ in their countries and in the age and sex of the people tested, so the ratios are approximate; but the direction is the one the selection predicts, and a random draw, non-responders and all, doesn’t fill up with people who came because of the outcome. And a positive patch test is not always an everyday problem: counting only people who also reported an itchy rash and avoiding scented products, EDEN put fragrance allergy at 1.9% of adults, or 0.8% on a stricter definition, and fragrance contact allergy was about twice as common in women as in men. The clinic number isn’t wrong; it is the one a dermatologist deciding what to test needs. It just has to travel with its label: “of patients patch tested for suspected allergy”.