0333 Sleep Patterns Throughout Mountain Warfare Training in U.S. Marines Measured Using a Consumer Wearable

Alice LaGoy,Joseph Cuellar,Max Smith,Rebecca Weller,Rebecca McClintock,Tony Duong,Timothy Dunn,Douglas Jones,Rachel Markwald
DOI: https://doi.org/10.1093/sleep/zsae067.0333
IF: 6.313
2024-04-20
SLEEP
Abstract:Abstract Introduction Poor acclimatization, high training demands, and sleeping conditions in the field can compromise the sleep of military personnel completing mountain warfare training. As chronic poor sleep can impact performance and recovery, individual differences in sleep may relate to differences in resilience during prolonged training. We characterized sleep patterns of Marines completing a mountain warfare training exercise (MTX) using a consumer wearable and explored relationships between sleep patterns and stress during training. Methods Marines (n = 59, aged 22±4 years [mean ± SD]) wore a consumer wearable (Oura Ring, Gen 2) during a 4-week MTX conducted in the fall (n = 17) or winter (n = 42). Participants also completed the Patient-Reported Outcomes Measurement Information System Sleep Disturbance scale (PROMIS; short form version 8b) and Stress Impact Scale (SIS) at three time points during each MTX (start, mid, end). Wearable-derived sleep outcome measures were averaged for each training phase and included total sleep time (TST), standard deviation of sleep midpoint, and wake after sleep onset (WASO). Longitudinal k-means clustering was used to identify different sleep patterns. Cluster by training phase (3×3) ANOVA were performed to explore relationships between sleep patterns and stress. Results Three clusters were identified. Cluster A (n = 31; 8 fall, 23 winter) had longer TST (6.8±0.7h) and low WASO (1.2±0.4h) but high PROMIS (high sleep disturbance; 56.2±5.2). Cluster B (n = 16; 6 fall, 10 winter) had short sleep (6.3±1.0h) but low WASO (1.4±0.5h) and low PROMIS (43.4±6.2). Cluster C (n = 12; 3 fall, 9 winter) had short sleep (5.9±0.8h), high WASO (1.9±0.5h), and high PROMIS (59.7±6.6). There was a significant effect of cluster on SIS (p < .001); cluster B (low sleep disruption group) reported higher scores (better stress management) than clusters A and C. Conclusion Three distinct sleep patterns were identified that varied in sleep duration with self-reported (PROMIS) and wearable-derived sleep disturbance metrics (WASO). Importantly, the low sleep disturbance cluster (Cluster B) reported managing stress better throughout MTX. Personnel with less disturbed sleep may cope better with prolonged training. Support (if any) Office of Naval Research (work unit no. 62417).
neurosciences,clinical neurology
What problem does this paper attempt to address?