29 SLEEP RESTRICTION [Rev 10-27-2017]
This section is cited in Step D1 (Section 46), for a person who is not tired at bedtime, and in Step D3 (Section 48) for a person with interrupted sleep.
For some people, they are simply not tired at the time they choose to go to sleep or need to go to sleep. They may try to treat the problem by spending more time in bed, in hopes that they will eventually doze off. Instead, they are often made more frustrated and tense by the failure of their solution, creating a situation where it is even more difficult to get to sleep. This is the “not tired” problem from Part D1.
Others may fall asleep with little difficulty but wake during the night and have trouble returning to sleep, often remaining awake for hours at each awakening. This is the interrupted sleep case from Part D3.
The same treatment solution applies in both cases. We limit the amount of time that the person spends in bed. This may be counter-intuitive, because lack of sleep invites more bed time. Sleep restriction pushes for better quality of the time spent in bed.
Patient should be warned that daytime sleepiness will probably increase for a few weeks. However, the goals are regular sleep, less time spent falling asleep, and deeper sleep. Once a person attains those goals, sleep time can be increased somewhat, to relieve daytime fatigue.
Helpful resources for this treatment are Spielman, Yang and Glovinsky and : Glovinsky and Spielman (pp. 141 and 189-195).
29a. Not Tired at Bedtime (Part D1)
This problem can be a result either of the person choosing an inappropriate bed time or not being sleepy.
The first issue can be addressed by having a consistent sleep time (Section 25) and limiting the time lying awake (Section 34).
Sleep restriction combines these two into a single process.
- Have the person complete a sleep log for two weeks, to get a picture of his/her sleep pattern. It should include estimates of time spent sleeping, whenever it occurs.
- Determine the total amount of sleep the person is actually getting each day.
- Pick a single time to get that amount of sleep, with a consistent wake-up time each day. It should probably be later than the person usually goes to bed, to encourage falling asleep quickly.
- Continue until sleep has stabilized in the new pattern.
- Once the person has been on a regular schedule, check for adequate sleep and rapid falling asleep, and adjust the amount of time as needed.
For example, a person may go to bed at 10 PM each day, take an hour and a half to fall asleep, and wake at 8 AM.
- Begin by noting that she is actually getting 8 ½ hours of sleep at night.
- Decide which 8 ½ hours she wants to sleep. The most likely would be 11:30PM to 8AM or 10 PM to 6:30AM.
- If she chooses the first, have her go to bed at 11:30PM and get out of bed at 8AM for the next two weeks.
- Adjust as needed.
A table might be helpful here:
| DAY | Bedtime Hour | Amt time to sleep | Hour awake | Amt time to get up | Time lost in night | Net sleep time |
| Mon | 10:00 | 2 hrs | 7:00 | 20 min | 0 | 7 hrs |
| Tue | 9:00 | 2 hrs 30 min | 6:45 | 30 min | 10 min | 7 hrs 5 min |
In this case, net sleep time for Monday ignores time to get up. You can calculate other ways, depending on patient issues.
This could be filled out by the patient every night for a week. If it turned out that the average sleep time was about 7 hours, then he/she should be told to be in bed for no more than 7 hours each night for the next two weeks, regardless of whether he/she is sleeping.
The patient should then select the hours to be in bed, making them as consistent as possible (Section 25). Because it has been taking so long to get to sleep, this person might be advised to go to bed at midnight and get up at 7:00 AM. At first, it will probably be difficult to get to sleep, even at midnight, but the person’s circadian cycle will reinforce sleeping at the newly chosen time, making it easier to follow the plan than if another time is chosen.
At the end of one week and again after the two weeks, check to see how things are going, and adjust as needed.
29b. Interrupted Sleep (Part D3)
In this case, the person gets to sleep then awakens during the night, possibly several times. That in itself is not a problem.
However, if he/she has difficulty getting back to sleep, it is a problem. Sleep restriction may be the answer, especially if other issues have been resolved – limited use of the bedroom (Section 20), no naps (Section 32), and so on.
The procedure here is similar to part (a) above.
- Get an estimate of the total number of hours the person is sleeping, taking into account time awake during the night.
- Pick a consistent schedule that gives the person just that amount of time in bed.
- Once he/she has been on the schedule for a while, see whether time awake during the night has diminished and daytime sleepiness is tolerable. Adjust the schedule as needed.
Documenting sleep loss might be a little trickier in this case, because each night could require two or more lines, and different numbers of lines for different nights. Here is a possible example table, where each awakening gets a different line. Monday has four lines because the person awoke three times in the night.
| DAY | Bedtime Hour | Amt time to sleep | Hour awake | Amt time awake | Time lost in night | Net sleep time |
| Mon | 10:00 | 20 min | 12:30 | 45 min | ||
| 2:00 | 40 min | |||||
| 4:30 | 30 min | |||||
| 7:00 | up | 2 hrs 15 min | 6 hrs 45 min | |||
| Tue | 10:00 | 20 min | 12:45 | 1 hour |
This person went to bed Monday at 10:00 and woke at 12:30, 2:00, and 4:30, then finally got up for the day at 7:00. Time lost was the sum of amount of time to get to sleep, 20 minutes and the amounts of time awake in the night, 45 + 40 + 30 minutes. The total is135 minutes, or 2 hours and 15 minutes.
The fourth line gives Monday night’s total: 9 hours from bedtime to awakening for the day, minus 2 hours 15 minutes awake.
It is probably best to have the person keep the record-sheet in a convenient place to enter during the night, because it is unlikely that he/she will remember these numbers the next day. Even so, the numbers can only be approximate, and the total could be somewhat misleading.
A strict sleep restriction schedule would have this person going to bed for 6 hours and 45 minutes every night. Some choices will have to be made about time to go to bed and time to arise.
29c. Variations
The original protocol calls for a total time in bed each night that equals the estimated sleep time from the patient’s sleep log.
Allowing some more time in bed early on may make it easier for some patients to comply with the restrictions. (Spielman, Yang and Glovinsky, p.285).
Other people may like the idea of less time in bed at the start, with increases over time, so the task gets easier.