Smartwatch-measured nighttime scratching may be moderately associated with disease severity and clinical improvement among pediatric patients with atopic dermatitis, according to a prospective study.
In the study, researchers enrolled 50 patients aged 4 to 15 years with physician-diagnosed atopic dermatitis who wore an Apple Watch for 5 to 14 nights during the initiation of topical therapy. The device captured three metrics—the scratch count rate, scratch duration ratio, and scratch burden index (SBI)—using accelerometer data. These measures were evaluated against clinical severity using the Eczema Area and Severity Index (EASI), patient-reported symptoms using the Patient-Oriented Eczema Measure (POEM), and serum thymus and activation-regulated chemokine (TARC).
Across the monitoring period, all scratching metrics showed moderate correlations with baseline EASI, with correlation coefficients ranging from 0.60 to 0.64. The correlations persisted following treatment but were weaker, ranging from 0.43 to 0.47. Declines in scratching over time paralleled clinical improvement, with correlations ranging from 0.67 to 0.71.
Scratching metrics also aligned with biomarker levels. Associations with serum TARC ranged from 0.58 to 0.60, consistent with associations observed for clinical severity. Associations with POEM were modest. In contrast, numerical rating scale itch scores showed no statistically meaningful correlation with EASI or serum TARC and only weak associations with nighttime itch.
In exploratory analyses, the SBI-based model demonstrated the strongest ability to distinguish patients who achieved clinically meaningful improvement, defined as at least 50% reduction in EASI and a 4-point reduction in POEM. The model showed an area under the curve of 0.78, sensitivity of 75%, and specificity of 79%.
A temporal pattern was observed in which scratching metrics correlated more strongly with baseline EASI compared with with follow-up EASI, suggesting that underlying disease severity may influence subsequent scratching behavior over short observation periods.
The study was limited by its small sample size, single-center design, and short monitoring period of 1up to 2 weeks. The observational design precluded causal inference. Biomarker data were collected at a single time point, and the predictive models were exploratory and not externally validated. Pediatric-specific validation of the wearable algorithm remained limited, and treatment wasn't fully standardized.
Overall, the findings indicated that wearable-derived nocturnal scratching metrics may function as objective, treatment-responsive indicators of disease activity. However, the researchers emphasized caution. "[These metrics] may serve as treatment-responsive digital measures, [but] further prospective validation is required before clinical application,” wrote lead study author Fumiko Iwai, of the Allergy Center at the National Hospital Organization Mie National Hospital as well as the Department of Child Health and Development at the Graduate School of Medicine at Mie University in Japan, and colleagues.
The study was partially funded by a Health and Labour Sciences Research Grant from the Japanese Ministry of Health, Labour, and Welfare. Full disclosures of the study authors can be found in the study.
Source: Journal of Clinical Medicine
