0239 A Latent Profile Analysis of actigraphic sleep and physical activity measures among Cambodian-Americans: Relationship with specific trauma symptoms

Abstract

Abstract Introduction Sleep and physical activity are related to psychological trauma. Less is known about how individuals with distinct sleep and activity profiles differ on specific clusters of trauma symptoms. Cambodian-Americans who survived the Pol Pot genocide experienced severe collective trauma. This analysis explored group differences between sleep/activity profiles on specific trauma symptoms among Cambodian-Americans. Methods Participants in a diabetes prevention trial for Cambodian-Americans (NCT02502929) met inclusion criteria for depression and high diabetes risk (but did not have diabetes). They wore wrist actigraphy (sleep) and hip actigraphy (physical activity) for 7 days (≥3 days to be included) and completed the 16-item trauma symptom scale of the Harvard Trauma Questionnaire (HTQ; N=166). Latent Profile Analyses identified profiles using 3 mean actigraphic sleep and activity variables: total nightly sleep time, sleep maintenance efficiency, and minutes in moderate-vigorous physical activity. ANOVAs explored differences between sleep/activity profiles on the HTQ, specifically total scores and the “Avoidance/Numbing” and “Re-experiencing/Hyperarousal” subscales. Models were adjusted for psychotropic medication use. Results Participants were predominantly women (79%), mean age 55.3, with elevated trauma symptoms (17% were higher than 2.5 cutpoint; mean±SD= 1.90±0.61). Sleep and physical activity patterns yielded a BIC best fit with 3 sleep/activity profiles: Inactive Poor Sleepers (n=30, 18%), Highly Active Short Sleepers (n=35, 21%), and Moderately Active Good Sleepers (n=101, 61%). Differences were observed between profiles on the HTQ total score (p=0.03). Tukey’s post hoc test revealed that Inactive Poor Sleepers exhibited greater HTQ scores than Highly Active Short Sleepers (p<0.05), but did not differ from Moderately Active Good Sleepers. There was also a significant difference between profiles in the Avoidance/Numbing subscale (p=0.01); Inactive Poor Sleepers had higher Avoidance/Numbing than Highly Active Short Sleepers (p<0.05, Cohen’s d: 0.47). There were no differences between profiles on the Re-experiencing/Hyperarousal subscale (p=.09). Conclusion Individuals with contrasting actigraphic sleep/activity profiles differed on trauma symptoms. Inactive Poor Sleepers may benefit from specific interventions for Avoidance/Numbing symptoms. Future analyses will evaluate how changes in sleep/activity profiles are longitudinally related to psychological health and diabetes risk following interventions. Support (If Any) 5R01DK103663