0319 Evaluating Illness Severity from Acute to Chronic Insomnia: Is the First the Worst?

Julia Boyle, Alexandria Muench, Knashawn Morales, Jason Ellis, Ivan Vargas, Michael Grandner, Donn Posner, Michael Perlis

Research output: Contribution to journalMeeting Abstractpeer-review


Introduction The Spielman 3P and Perlis 4P models represent illness severity over the course of insomnia disorder. The 3P model suggests that illness severity is worst when subjects experience acute illness (as compared to the subchronic and chronic phases of the disorder). The 4P model suggests that illness severity crescendos with chronicity. The present analysis uses an archival data set to assess illness severity with new onset illness (i.e.,from Good Sleep [GS] to Acute Insomnia [AI] and from AI to Chronic Insomnia [CI]). Illness severity is quantified in terms of Total Wake Time (TWT). Number of affected days/week is a tabulation of nights/week with clinical insomnia symptoms. Methods Subjects were recruited as GSs(N=911), followed up to 1-year with digital sleep diaries, and were classified as GS, AI, or CI based on standardized criteria. Data for CI subjects were anchored to the 1st of 14 days with insomnia so that day-to-day TWT, SL, WASO, and EMA were represented for ~100 days prior-to-and-following AI onset. A similar graphic (+/-Acute onset) was constructed for number of days per week with insomnia. GS data were matched to CI subjects dates (i.e., the data were temporally tethered). Segmented linear mixed regression models were applied to examine the change in slopes in the AI-to-CI period compared to GS and AI period. Results 883 subjects remained as GSs (Mage=53.1±10.8years;35.1%female;83.0%white), while 28 indi-viduals transitioned to AI and then CI (Mage=53.6±12.6years;39.3%female;89.3%white). Average TWT rose over the course of the first 2 weeks from the AI onset(b=1.6,SE=0.52minutes,p=0.002) and then was stable for 3 months(b=-0.004,SE=0.03,p=0.91). This worsening of sleep continuity was almost entirely related to increased WASO. Average number of affected days was found to be stable from AI to CI (b=0.002,SE=0.002,p=0.20). That is, while there was week-to-week variability in the number of days affected, no linear trend was evident. Conclusion These data suggest that, in our sample of CIs with primarily middle insomnia, the average severity and number of affected days was worst with the onset of AI (worst is first) and stable thereafter. Whether this temporal trend applies equally to new onset Initial and Late insomnia remains to be determined. Analyses ongoing. Support (if any) R01AG041783
Original languageEnglish
Pages (from-to)A142-A142
Number of pages1
Issue numberSupplement_1
Early online date29 May 2023
Publication statusPublished - 29 May 2023

Cite this