Here is a strange fact about the human body: if you hand a person a sugar pill, tell them it is medicine, and ask how they feel afterward, roughly one in every seventy five people will report feeling sleepy. This is not a drug effect. It is the placebo machinery at work, producing drowsiness out of expectation, suggestion, or coincidence.
Now here is the useful finding: the best modern antihistamines produce sleepiness at almost exactly the same vanishingly low rate.
A systematic review and network meta-analysis published July 22 in the European Journal of Clinical Pharmacology has delivered what is arguably the most comprehensive accounting yet of how often newer generation allergy drugs actually make people drowsy. The answer, across 163 clinical trials involving tens of thousands of patients, is comforting for anyone who reaches for a nonsedating antihistamine during pollen season.
The lead author, Nattawut Leelakanok of Burapha University in Thailand, and eleven coauthors combed through five research databases from their inception through July 2025. They pulled every randomized controlled trial, observational study, and crossover trial that reported sleepiness rates for newer generation H1 antihistamines at standard therapeutic doses. From that mountain of data, they extracted the single number that matters most to patients: the proportion of people who felt sleepy after taking the drug.
What they found
The meta-analysis established a critical baseline. Among participants who received a placebo, 1.3 percent reported sleepiness, with a narrow confidence interval of 1.0 to 1.6 percent. That is the noise floor of subjective drowsiness in a clinical trial setting. Any antihistamine that hovers near this number is, for practical purposes, indistinguishable from nothing at all.
The drug that came closest was bilastine, taken at 20 milligrams. Its sleepiness prevalence was 1.6 percent, or just 0.3 percentage points above placebo. When the researchers performed a network meta-analysis, a statistical technique that allows indirect comparisons across drugs that have never been tested head to head, bilastine again came out on top with a relative risk of 1.06. That means bilastine users are only 6 percent more likely to report sleepiness than placebo users. By any reasonable standard, that is negligible.
Three other antihistamines also performed well. Fexofenadine at 120 to 180 milligrams, desloratadine at 5 milligrams, and bepotastine at 20 milligrams all showed sleepiness prevalence below 10 percent and failed to reach statistical significance when compared against placebo. In other words, the drowsiness these drugs produced could not be reliably distinguished from the background hum of placebo sleepiness.
Bepotastine earned an additional distinction. Among pediatric patients, it showed the lowest sleepiness incidence of any drug studied, making it a strong candidate for children who need allergy relief without classroom drowsiness.
Why it matters
The findings matter because the distinction between “sedating” and “nonsedating” antihistamines has never been clean. First generation antihistamines such as diphenhydramine, the active ingredient in Benadryl, cross the blood brain barrier readily and are known to cause significant drowsiness. Newer generation drugs were designed to stay mostly in the periphery, but the degree to which they actually cause sleepiness has been surprisingly hard to pin down.
Individual clinical trials are often too small to detect rare side effects reliably. A drug that causes sleepiness in 2 percent of users might look identical to a placebo in a trial of 200 people simply because both numbers are so low. That is where a meta-analysis of 163 studies adds real value. By pooling data across trials, the Thai research team gained the statistical power to see patterns that individual studies could not reveal.
The pattern they found is reassuring. Every single nonsedating antihistamine did produce a higher absolute prevalence of sleepiness than placebo. But for most of them, the difference was so small that it could easily be explained by chance. The study essentially gives clinicians and patients permission to stop worrying about drowsiness when choosing among bilastine, fexofenadine, desloratadine, and bepotastine at standard doses.
Limits and caveats
The study has important limitations. The 163 included trials varied widely in design, population, and how they measured sleepiness. Some used validated scales; others relied on spontaneous patient reports. The heterogeneity statistics, with I squared values around 40 to 48 percent, suggest moderate inconsistency across studies, meaning the pooled estimates should be interpreted with some caution.
The analysis also focused exclusively on sleepiness as an adverse event. It did not capture other central nervous system effects such as dizziness, cognitive slowing, or impaired reaction time, which some antihistamines may cause even without frank drowsiness. And the meta-analysis only included newer generation drugs at their usual doses. Patients who take higher than recommended doses, or who combine antihistamines with alcohol or other sedating medications, may experience more sleepiness than these numbers suggest.
Finally, the 1.3 percent placebo sleepiness rate deserves a closer look. The fact that more than one percent of people in the placebo arms of these trials reported feeling sleepy is itself a reminder of how much noise exists in subjective symptom reporting. Some of those people may simply have been tired. Some may have expected to feel drowsy because they thought they were taking allergy medicine. And some may have been reporting a genuine placebo effect. The meta-analysis cannot disentangle these possibilities.
The bottom line
The bad news, if it can be called that, is that no antihistamine is perfectly nonsedating. Even bilastine, the star performer in this analysis, produced sleepiness in 1.6 percent of users. That is slightly above placebo. The good news is that “slightly above placebo” is, for most patients, good enough. A 1.6 percent sleepiness rate means that 98.4 percent of people who take bilastine will not feel drowsy. For fexofenadine, desloratadine, and bepotastine, the numbers are similar.
The study also carries a broader lesson about how we evaluate drug side effects. When an adverse event is rare, the difference between a drug and a placebo can be vanishingly small even when it is real. The Thai team has shown that for modern antihistamines, the difference exists but is barely measurable. That is about as close to a clean bill of health as any class of drugs is likely to get.
For seasonal allergy sufferers who have been hesitant to take medication because they fear daytime drowsiness, the message is clear. The risk is real but tiny. And for the one in seventy five people who would feel sleepy on a sugar pill, no drug will ever change that number by much.
Source: Leelakanok N, Pongpun A, Sapapsap B, et al. The incidence of sleepiness in newer-generation antihistamine users: a systematic review and network meta-analysis. Eur J Clin Pharmacol. 2026;82(8):215. PMID: 42481869. DOI: 10.1007/s00228-026-04147-y.