A New JID Study Asks Which Itch Scale Doctors Should Actually Trust

A paper set for the October 2026 issue of the Journal of Investigative Dermatology puts nine authors from German dermatology research groups, led by Felix Witte and senior author Sonja Ständer, behind a question that sounds narrow but is not: which itch scale should anyone trust. The team, working with the Society for Investigative Dermatology, compared how single-item measures like the Visual Analog Scale, Numeric Rating Scale, and Verbal Rating Scale track against multidimensional tools such as the Itch Severity Scale, the 5-D Itch Scale, ItchyQoL, and the PROMIS Itch Questionnaire, then checked those scores against cytokine levels in patients with atopic dermatitis.

The timing matters because itch scales have quietly become regulatory instruments. Trials for nemolizumab, the IL-31 receptor blocker Galderma sells as Nemluvio, used the Peak Pruritus Numeric Rating Scale as the primary efficacy endpoint in its OLYMPIA 1 and 2 trials, reporting that 56 to 58 percent of treated patients hit a four-point itch reduction versus 17 to 21 percent on placebo, according to data the company and Dermatology Times published around the drug's FDA filing. Dupilumab trials leaned on similar NRS-based endpoints. A paper that finds these scales behave differently from each other is not just a methods footnote. It is a comment on how several approved biologics were scored on their way to market.

That differential behavior is already documented, which is what makes the JID study's framing legible. A 2022 comparison tabulated in a PMC review found that VAS, NRS, and VRS correlate tightly with one another but diverge from the ISS, which tracks closer to quality-of-life measures like ItchyQoL and DLQI than to raw intensity scores. Separately, RTI International's analysis of the Peak Pruritus NRS reported differential item functioning by age, sex, and race and ethnicity, meaning identical numeric scores on the same instrument may not represent equivalent itch burden across demographic groups. Neither finding proves the new paper's conclusions, but both establish that the measurement gap it investigates was already visible before this publication.

The cytokine side of the comparison carries its own unresolved history. IL-31 is the most studied neuroimmune mediator tied to itch in atopic dermatitis, yet its correlation with reported itch severity has been inconsistent across the literature. A study in PMC found no significant correlation between serum IL-31 and itch intensity in a pediatric cohort, a result that sits awkwardly next to adult studies reporting strong correlations with subjective itch and IgE levels. One explanation that has circulated informally in the field, and which this JID paper appears positioned to test directly, is that the scale used to capture itch intensity, not just the patient population, may be driving some of that inconsistency.

What the summary and title make clear is scope: this is a measurement-properties study comparing scales against each other and against cytokines, not a clinical trial and not a drug-efficacy analysis. It cannot, on its own, tell a clinician which scale to use at the bedside for an individual patient, nor can it retroactively validate or invalidate the endpoints used in nemolizumab or dupilumab trials. What it can do is supply evidence for a narrower, falsifiable claim: whether a given itch scale tracks more closely with a specific cytokine signature than its competitors do, in the specific cohort the authors studied.

That narrower claim still has teeth for drug development. The field has been searching for a biomarker-correlated itch measure that could supplement, rather than simply replace, subjective patient-reported scales like the PP-NRS, which carry known placebo-response and recall-bias problems in trials. FDA guidance under its Patient-Focused Drug Development initiative has pushed sponsors toward validated, fit-for-purpose instruments rather than whichever scale a given trial happened to adopt. A scale shown to correlate consistently with IL-31 or TSLP levels would be a candidate for that role, though a single comparative study, however well constructed, is an early input to that regulatory conversation rather than a resolution of it.

The paper's publication venue and date are themselves a data point worth noting on their own terms. Volume 146, Issue 10 of JID lands in October 2026, meaning the findings have not yet had time to accumulate citations, replications, or rebuttals elsewhere in the literature. Readers encountering it through Elsevier's ScienceDirect listing are seeing a single measurement study against a backdrop of inconsistent prior findings on IL-31 and itch, not a settled consensus. The honest read of this work, before any secondary reporting builds around it, is as one data point in an argument about measurement that has been running since before this paper existed, not as its ending.