From digital phenotype to closed-loop therapy: artificial intelligence–enabled wearable and connected technology for allergic diseases

Background The prevalence of allergic diseases continues to rise, yet episodic outpatient care remains constrained by delayed symptom recognition, the "black box" of medication adherence, and fragmented management. Wearables, the Internet of Things (IoT), and artificial intelligence (AI) now converge in "digital phenotyping." From a pharmaceutical perspective, this review appraises these AI-enabled technologies for allergic diseases along the thread of "from digital phenotype to closed-loop therapy." Methods A narrative review synthesizing recent literature on the technical framework, real-time monitoring, risk early warning, long-term management, therapeutic wearables, psychological digital phenotypes, and equity and safety. Results These technologies span the "sensing–warning–management–treatment" chain. Adaptive baseline learning, time-aware modeling, and multimodal fusion enable individualized early warning from noisy physiological and environmental signals. The strongest evidence concerns long-term management: digital inhalers and smartwatches, validated in randomized controlled trials and meta-analyses, improve adherence and asthma control and reduce exacerbations, with economic value in averting biologic escalation; devices are extending from passive sensing to active interventions—nasal acoustic therapy, "feedback as therapy," and conceptual closed-loop epinephrine injection. However, most applications remain at small-sample, single-center validation; portable quantitative pollen detection is an empty niche; and drug confounding of digital phenotypes (β 2 -agonist contamination of heart rate variability) and closed-loop amplification of skin-tone–related bias challenge clinical translation. Conclusions AI-enabled wearables are shifting allergy care from intermittent, subjective consultation toward continuous, closed-loop management, with evidence most mature for digital inhalers. Future work should develop multimodal digital phenotypes and allergy-specific foundation models and strengthen medication-covariate modeling, hard-endpoint validation, and whole-lifecycle equity governance.