목적: Korea experiences notable seasonal fluctuations in temperature and humidity, which may affect various aspects of sleep, including the Apnea-Hypopnea Index (AHI) and snoring frequency. This study examines monthly variations in AHI and snoring using Sleep Routine, a mobile, audio-based, contactless sleep-tracking service. 방법:Data were collected from December 1, 2023, to November 30, 2024, from users in Korea who used the app at least once per month. Sessions under 4 hours or over 15 hours, as well as those in which any one of the four sleep stages had a ratio of 100%, were excluded as invalid. The service uses an AI model to estimate AHI based on detected OSA events and sleep stages, and also detects snoring on an epoch basis. According to previous validation studies, the AI model demonstrates accuracies of 88%, 93%, and 94% for detecting AHI thresholds of 5, 15, and 30, respectively, in home environments, while snoring detection achieved an epoch-level accuracy of 91.4%. 결과:A total of 237,190 sessions from 1,067 users were analyzed, averaging about 19,766 sessions per month (range: 17,321-21,544). Each user measured an average of 222.3 sessions (SD = 74.21, range: 24-358). The overall mean AHI was 4.77. By month, the lowest AHI was 4.53 in October, and the highest was 4.96 in July (p < 0.001). Seasonally, AHI tended to be higher in the summer (June–August) and lower in the fall (September–November). Snoring frequency data were available from January to November due to the service’s launch schedule. The overall average was 17.57%, peaking at 19.86% in March and dipping to 14.17% in August (p < 0.001). 결론:While AHI patterns were less clearly associated with specific climate factors, snoring frequency appeared to have a strong correlation with temperature data from the Seoul region. Although reliance on a commercial service poses accuracy and usage limitations, this large-scale dataset provides valuable real-world insights into seasonal changes in sleep-disordered breathing.