Abstract
This study compared self-reported sleep latency (SL) and total sleep time (TST) to objective measures on polysomnography (PSG) during pregnancy. Thirty-three women in the third trimester (T3) of pregnancy, 16 women in the first trimester (T1) of pregnancy, and 15 non-pregnant women underwent overnight PSG, and shortly after awakening reported their perceived SL and TST. Results showed that, on average, the T3 group slightly overestimated their TSTs, whereas the T1 and non-pregnant groups underestimated TSTs when compared with objective measurement. All groups overestimated SL, and perceived SL was closest to the first epoch of 10 min of uninterrupted sleep or the first epoch of slow-wave sleep, rather than the first epoch of sleep (the current definition used for diagnostic sleep studies). The wide variation in discrepancies between estimation and PSG measurement for both TST and SL shows that self-reports made by both pregnant and non-pregnant women tend to be unreliable, which has important implications both clinically and for the many studies based on self-reported sleep patterns in pregnancy.