1224 Comparing Actigraphy Scoring Protocols in Adolescents with Anxiety and Related Disorders (ARD)

Le résumé fourni par la source

Abstract Introduction Estimating sleep from actigraphy is a challenge in psychiatric populations where criteria may be susceptible to altered physical activity. Expert-consensus procedures may introduce bias. Alternative scoring protocols may resolve ambiguities, avoid the need for consensus, and increase reproducibility. This study compared two scoring procedures in a treatment-seeking sample of adolescents with ARD. Methods Thirty-six adolescents (24 F; ages 13-17yrs [m±sd: 15±1.7yrs]) diagnosed with ARD wore a patch activity monitor on their dominant triceps and completed sleep diaries for 8.7±3.1 nights resulting in 180 unique nights. Sleep was estimated in 1-minute epochs (Sadeh); sleep periods were scored with sleep diaries using two scoring protocols: (1) “3-5” (onset at first of 3 consecutive minutes of sleep; offset at last of 5 consecutive minutes of sleep); and (2) “15-15” (onset at first of 15 consecutive minutes of sleep; offset at last of 15 consecutive minutes of sleep). We tabulated nights requiring group consensus (e.g., ambiguity of sleep within an hour before, or after, the major sleep episode). We examined differences in sleep period time and total sleep time (in minutes) and sleep midpoint, onset, and offset (in decimalized clock-hour) from five randomly selected weeknights. Results 15-15 scoring produced fewer nights requiring consensus (70 nights, average: 17% per participant) compared to 3-5 scoring (147 nights, average: 30% per participant). 15-15 scoring reduced estimated sleep period time by 30.2±51.1 minutes (3-5: 543.4±86.2 min; 15-15: 513.2±71.9 min; difference 95%CI: [-46.91, -13.54]; p=.001), but not total sleep time (3-5: 448.7±69.7 min; 15-15: 443.5±66.2 min; difference: -5.2±33.5 minutes [-16.2, 5.8]; p=.39). 15-15 scoring shifted estimated sleep midpoint 0.25±0.4 hours earlier versus the 3-5 rule [3-5: 4.5h±0.7h; 15-15: 4.3h±0.6h; difference 95%CI: [-0.39h,-0.11h], p=.001). The two procedures did not shift estimates (p’s>.42) of sleep onset (3-5: 19.6h±4.3h, 15-15: 19.4h±4.6h) or offset (3-5: 7.5h±1.3h, 15:15: 7.4h±1.2h). Conclusion The 15-15 scoring protocol reduced ambiguities and the need for consensus compared to the 3-5 protocol in a psychiatric adolescent sample. It affected estimates of sleep period time and sleep midpoint but not total sleep time. These data indicate that alternative scoring may be considered in populations where symptoms may contribute to sedentary activity. Support (if any) P20GM139743 (MAC).

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.