Table 2.
Differences in sleep-related endpoints between PCOS and controls.
| Urinary Melatonin Variables | Controls | PCOS | |||
|---|---|---|---|---|---|
| Mean | Std Dev | Mean | Std Dev | p Value | |
| Mean overnight MEL (ng MEL/mg CRE) | 6213.07 | 4146.26 | 6972.42 | 3584.04 | NS |
| Mean daytime MEL (ng MEL/mg CRE) | 865.93 | 643.20 | 1297.28 | 1122.41 | NS |
| Night/day ratio (1) | 15.76 | 16.06 | 13.41 | 8.72 | NS |
| Night/day ratio (2) | 13.60 | 13.01 | 11.49 | 8.57 | NS |
| PSQI-defined sleep variables | |||||
| PSQI total | 5 | 2.7 | 7 | 4.0 | NS |
| Good sleep quality, N (%) | 9 (69) | 12 (55) | NS | ||
| Poor sleep quality, N (%) | 4 (31) | 10 (45) | |||
| Accelerometer-defined sleep variables | |||||
| Total sleep time (h/day) | 6.5 | 0.80 | 6.5 | 0.94 | NS |
| Total weekday sleep time (h/day) | 6.3 | 1.28 | 6.5 | 1.05 | NS |
| Total weekend sleep time (h/day) | 7.1 | 1.16 | 6.5 | 1.46 | NS |
| Total sleep efficiency (%) | 84 | 6.9 | 82 | 7.7 | NS |
| Weekday sleep efficiency (%) | 82 | 10.1 | 81 | 7.6 | NS |
| Weekend sleep efficiency (%) | 88 | 5.0 | 81 | 8.3 | 0.010 |
Variables were transformed as needed (logarithmic or square root) to meet assumptions. Means and SD reflect raw data. Implausible values (sleep efficiency > 100%) were excluded from their endpoint-specific analyses. Categorical analyses were conducted using Fisher’s Exact tests (two-sided). A PSQI score of >5 is categorized as poor sleep quality. MEL, 6-sulfatoxymelatonin; CRE, serum creatinine; PSQI, Pittsburgh Sleep Quality Index.