Circadian Sleep Disorders Network - Advocating for people with misaligned body clocks
Printer-friendly page
ON THIS PAGE (links): Top
Statistical evidence
Observational evidence
Mechanistic evidence

Can Chronotherapy Cause Non-24? (Q&A)

by James Fadden

Q: I have read that chronotherapy can cause people with DSPS to develop Non-24. What is the evidence that this can happen? Could it be that those people would have developed Non-24 anyway?

Note: This question refers to what is more specifically known as phase-delay chronotherapy, a treatment for DSPS which involves progressively delaying sleep until one "goes around the clock" to the desired sleep phase. The term "chronotherapy" has other uses in chronobiology, sometimes including any therapy for resetting the body clock, but what is said here applies only to the phase delay form.

A: Certainly not everyone who does chronotherapy develops Non-24, even among night owls. But the evidence strongly favors the idea that there are some individuals who would have remained in stable DSPS indefinitely, but who are pushed over the edge into Non-24 as a result of chronotherapy.

The evidence is of three types: statistical, observational, and mechanistic.

The statistical evidence is the strongest. DSPS is a very common condition. Non-24 is an extremely rare condition. Some persons with DSPS do go on to develop Non-24 in the course of their lives, but this is very rare. The vast majority of people with DSPS never progress to Non-24.

If chronotherapy only preceded Non-24 in those persons with DSPS who would have developed Non-24 eventually, we would expect only a very small percent of persons who do chronotherapy would develop Non-24. That does not appear to be the case. Among the online groups of Non-24 patients or among the patients of doctors treating Non-24 a history of Non-24 developing after chronotherapy is not unusual. This strongly suggests that chronotherapy brings about Non-24 in people who would not otherwise have developed this rare disorder.

The case is even stronger when considering that the percentage of people with DSPS who do chronotherapy is very low (most people have never heard of chronotherapy.) So we have to very rare sets: set 1 – those people who progress from DSPS to Non-24 (whether or not they did chronotherapy), and set 2 – those people with DSPS who try chronotherapy. If progression from DSPS to Non-24 would have happened with or without chronotherapy then membership in these two sets would be independent. The probability of someone being in both sets (doing chronotherapy and developing Non-24) would be obtained by multiplying two very low probabilities together, meaning an extremely low probability of finding someone in both sets (doing chronotherapy and developing Non-24). The fact that we find a significant number of such persons, particularly in proportion to the very rare Non-24 population, suggests a causative link.

There is also an observational argument. There is one publication on the development of Non-24 after chronotherapy (a letter to the Editor to the NEJM).[1]

The letter reports three such patients – not a small number since there were only 11 total reported cases of Non-24 world-wide at that time. I was one of those three cases and I think my case is typical. I developed DSPS between the ages of 16 and 20 while at college. This corresponds to the normal change in chronotype during adolescence but was more extreme in my case. By age 20 I was generally getting up around noon. But once that 4am to noon sleeping pattern was established it remained stable from ages 20 to 28. There was no sign of progression of my condition. At age 28 I did chronotherapy and immediately found I could no longer entrain. I remained free-running for 10 years until treated with light/dark therapy (and still lapse into free-running at times).

One might speculate that I might have eventually developed Non-24 anyway, but there was no sign of that prior to the chronotherapy. My DSPS appeared to be stable and not progressing until I did chronotherapy. I think my case is not unique in that respect.

The authors of the NEJM letter appropriately refrained from making a definite statement of causation. But I know from discussing it with them that having seen 3 cases of Non-24 all occurring after chronotherapy, they certainly had the strong impression of a causal link.

Finally there is some understanding of the mechanisms by which this causation could occur. There are reasons to expect that once a Non-24 hour cycle is adopted, it can become self perpetuating.

The first is that during chronotherapy the person is often sleeping very late relative to their temperature cycle. Given that light delays the clock before the temperature minimum and advances it after, the person will end up being awake during the delay portion of their PRC and sleeping during the advance portion. This accelerates the delay of the temperature cycle which in turn delays the sleep propensity rhythm. To re-entrain one has to break out of this cycle, which may be difficult for some persons.

An additional factor is what is called "after-effects" [2]. It was found in animal studies that if you entrain animals to a non-24-hour cycle and then release them into constant conditions, their intrinsic free-running period will have been affected by their prior entrainment. For example a mouse which normally free runs in DD with a less than 24 hour period, but that had been entrained to a 28 hour LD cycle and then released into DD would free run with a period greater than 24 hours. It would take 50-100 days in DD for its normal <24 hour period to once more manifest. It only takes a couple of weeks of forced entrainment to a non-24 hour period to produce such lasting effects.

A more recent study from Harvard Medical School has confirmed that after effects also occur in human subjects [3]. Normal subjects were entrained to a 24.65 hour day to mimic conditions that would be encountered by explorers on Mars. After 13 days of entrainment the subjects intrinsic period – the cycle their body wants to follow if it can – had increased by 0.1 hours when measured in a constant routine procedure.

Persons with DSPS already have a longer than average intrinsic period, which may account for their entrainment at a later phase [4,5]. When this is compounded by undergoing chronotherapy and exposing themselves to, say, a 27 hour LD cycle (by virtue of their eyes being closed while asleep) they might have lengthened their intrinsic period to the point where entrainment to 24 hours becomes impossible. In essence people with DSPS before chronotherapy are entrained but just barely. The additional lengthening of their period that happens during chronotherapy may be enough to push some over the edge and develop Non-24.

There are probably many factors that determine why some people can do chronotherapy without problems while others develop Non24. Some factors may involve the biology of the individual while others may involve the situation and environment, such as light exposure. Possibly doing chronotherapy when transitioning to a new time zone with a new light/dark schedule (especially if you are outdoors a lot after arriving) might be different than doing it in place.

On one mailing list we have speculated that faster chronotherapy may be less risky than a slow delay – possibly a counterintuitive idea. I made a video speculating on the reasons why which readers may find interesting. [6]

Chronotherapy is not the only factor that can cause someone with DSPS to develop Non-24. Anecdotal reports suggest that this can also happen in some persons who have not done chronotherapy. However this is rare. The vast majority of persons with DSPS will never develop Non-24.

Doing chronotherapy however, changes the equation and seems to put persons with DSPS at significant risk of developing Non-24, even people who otherwise would have remained in a stable DSPS indefinitely.

It is unfortunate that chronotherapy seems to carry this risk. It is a useful therapy and many people find it helpful and can do it without harm. But for a certain percent of persons with DSPS it does seem to precipitate the onset of Non-24.


1. Oren DA, Wehr TA. Hypernyctohemeral syndrome after chronotherapy for delayed sleep phase syndrome. N Engl J Med. 1992 Dec 10;327(24):1762.

2. Pittendrigh CS and Daan S. A Functional Analysis of Circadian Pacemakers in Nocturnal Rodents I. The Stability and Lability of Spontaneous Frequency. J. comp. Physiol. 106, 223-252 (1976)

3. Scheer FA, Wright KP Jr, Kronauer RE, Czeisler CA. Plasticity of the intrinsic period of the human circadian timing system. PLoS One. 2007 Aug 8;2(8):e721.

4. Micic G, de Bruyn A, Lovato N, Wright H, Gradisar M, Ferguson S, Burgess HJ, Lack L. The endogenous circadian temperature period length (tau) in delayed sleep phase disorder compared to good sleepers. J Sleep Res. 2013 Dec;22(6):617-24.

5. Campbell SS, Murphy PJ. Delayed sleep phase disorder in temporal isolation. Sleep. 2007 Sep;30(9):1225-8.

6. Fadden, J. Non-24 After Chronotherapy
© 2024 Circadian Sleep Disorders Network
Last modified Apr 20, 2022
Problems: contact