Most safety analyses this desk covers find a signal. This one finds the reverse, and reporting it is the same discipline as reporting the others [1]. The hair and sexual-function questions from the same literature are in four questions about contraception, one answered.
Against DPP-4 inhibitors, skin reactions were reported proportionally less often for GLP-1 drugs, at a proportional reporting ratio of 0.27 (95% CI 0.257–0.284). Within the class the pattern varied: semaglutide had the highest raw rate, dulaglutide the lowest, and exenatide showed five times the odds of dulaglutide in the regression.
The 8.16% figure needs its unit stated or it misleads badly. That is a share of adverse event reports, not of patients: eight in a hundred reports mentioning a GLP-1 drug included a skin reaction. Since only a small fraction of people taking any medicine ever generate a report, the proportion of users experiencing a skin reaction is not calculable from this and is certainly far lower — the same denominator problem as in the erectile dysfunction signal.
The exenatide result also needs its reference point. An odds ratio of 5.01 sounds alarming and is measured against dulaglutide, not against taking nothing — a within-class ranking rather than a risk estimate. Exenatide is also an older drug with far fewer current users and an entirely different reporting history, which makes cross-drug comparisons in a spontaneous database unreliable in ways the confidence interval does not capture.
For someone choosing a product the practical content is modest and real: skin reactions happen, they are not the dominant adverse effect of this class, and if one occurs it is worth raising because it may affect whether treatment continues. The authors make that last point themselves, noting these reactions might influence adherence — which is the thread running through two thirds stop within a year and what patients said when asked.