Older adults who used sedative music as therapy for 45 minutes at bedtime for 3 weeks had better global sleep quality and also better individual components of sleep quality over time than those who did not. The exception was sleep disturbance. In addition, those who used music had better global sleep quality at each of the weekly time points and better scores on four components: better perceived sleep quality, sleep latency, sleep efficiency, and less daytime dysfunction. There was also a delayed effect for sleep duration by Weeks 2 and 3, and an increasing dose effect on overall sleep quality from the pretest until Week 3 (Figure 2).
Despite a meaningful clinical improvement in many par- ticipants, over half of those who improved were still poor sleepers after three weeks. These findings are comparable with the review in which Lacks and Morin (1992) found that insomniacs have substantial improvement in response to cognitive-behavioural interventions, but the majority do not become good sleepers. In our study, those with a milder sleep problem were more likely to respond.
Our findings were similar to those in the three previous studies that used qualitative (Mornhinweg & Voignier 1995), quasi-experimental (Levin 1998), or experimental (Zimmerman et al. 1996) approaches. In the experimental study, music was used in the afternoon rather than at bedtime when sleep was measured. All six types of music in our study were used at bedtime and were found useful for sleep; the six types were similar in their effects. All those who received music reported listening to it every night, and family members verified this information. The Chinese orchestra tape has not been tested before, but the remainder of the music used has been found effective for pain in postoperative studies (Good 1995, Good & Chin 1998, Good et al. 1999).
Although Hoch and Reynolds (1986) have suggested that three weeks are needed to observe a new intervention for sleep quality, others have found effects after 3-day (Zimmerman et al. 1996) and 2-week periods (Levin 1998). In the present study, a 45-minute tape of music every night at bedtime, along with brief instructions for relaxation, improved global sleep quality after only 1 week, and the effect continued to increase for two more weeks. The daily sleep log showed good agreement with weekly recall for the PSQI. As a plateau did not emerge (Figure 2), the possibility