Comparisons of partners' polysomnographic sleep quality, simple reaction time, and feeling refreshed in the morning after monitoring on each intervention limb in the crossover trial
| Study night and morning results | CPAP limb (n=22) | Placebo limb (n=22) | p value |
| Mean (SD) total recording time (min) | 489 (45) | 489 (45) | 1 |
| Mean (SD) sleep latency (min)2-150 | 26.3 (13) | 32.6 (15) | 0.3 |
| Mean (SD) SEI (%)2-160 | 78 (11) | 77 (11) | 0.8 |
| Mean (SD) arousal index (events/h) | 24 (11) | 25 (12) | 0.4 |
| Mean (SD) time in stage 3+4 (min) | 43 (24) | 43 (26) | 0.9 |
| Mean (SD) time in stage REM (min) | 83 (28) | 90 (25) | 0.9 |
| Median (IQR) SURT response time (s) | 0.33 (0.28–0.39) | 0.36 (0.33–0.38) | 0.4 |
| Mean (SD) reciprocal SURT response time (1/s) | 2.9 (0.5) | 2.9 (0.4) | 0.12-151 |
| Median (IQR) 95th percentile SURT response time (s) | 0.47 (0.44–0.6) | 0.50 (0.44–0.58) | 0.5 |
| Median (IQR) refreshed in morning2-164 | 3 (2.5–4) | 3 (2–4) | 0.4 |
SEI = sleep efficiency index; SURT = simple unprepared reaction time.
↵2-150 Sleep latency data were compared during the first limb of the study only because of a treatment × order effect.
↵2-160 SEI is calculated as total sleep time/total recording time.
↵2-151 There was a weak trend favouring better (quicker) partner mean reciprocal reaction times on the CPAP limb.
↵2-164 Maximally refreshed = 5, maximally unrefreshed = 1.