General Rheumatology · Sleep
Poor sleep is one of the most common problems in inflammatory arthritis and one of the least talked about. Two things are worth knowing before you change anything: sometimes bad sleep is your disease talking, not your habits, and the treatment that actually works for long-term insomnia is not the sleep hygiene advice most people are handed.
The first of the EULAR fatigue recommendations' overarching principles is that health professionals should be aware that fatigue involves multiple, mutually interacting biological, psychological, and social factors. The task force agreed on that one almost unanimously (level of agreement 9.9 out of 10, with every member scoring it 8 or above). The same document says fatigue should be monitored and management options offered as part of routine care, and that responsibility for raising it should not sit with you alone.
It is also common. The Danish trial of sleep treatment in rheumatoid arthritis reports the prevalence of insomnia in RA as about 60 per cent, against 10 to 30 per cent in the general population. If you are not sleeping, you are closer to the rule than the exception.
This is the part worth acting on. The EULAR recommendation is explicit: the presence or worsening of fatigue should trigger evaluation of inflammatory disease activity and consideration of starting or changing immunomodulatory treatment where clinically indicated. That recommendation carries the guideline's strongest backing (level of evidence 1a, grade of recommendation A). The evaluation might go beyond a disease-activity score to imaging, for example MRI of the sacroiliac joints or spine, or ultrasound of the joints.
For back pain specifically, the ASAS expert criteria for inflammatory back pain use five features: onset before age 40, insidious onset, improvement with exercise, no improvement with rest, and pain at night that improves once you get up. Pain at night was one of the two strongest of the five in the analysis behind those criteria. If at least four of the five are present, the criteria identified inflammatory back pain with a sensitivity of 77.0 per cent and a specificity of 91.7 per cent. See the axSpA guide for what that means in practice.
The practical version: night pain that eases when you get up is a reason to talk to your rheumatology team, not a reason to buy a better mattress.
Cognitive behavioural therapy for insomnia (CBT-I) is a structured, time-limited programme. It is not a relaxation tape and it is not a list of tips. It works on the habits and thinking that keep insomnia going, using methods such as stimulus control and sleep restriction, guided by a sleep diary.
It has been tested in rheumatoid arthritis. A Danish randomised trial gave 62 people with RA and chronic insomnia either six weeks of nurse-led group CBT-I or usual care. The results are worth reading carefully, because they cut both ways. Measured objectively in a sleep laboratory, CBT-I did not significantly improve sleep efficiency (88.7 per cent versus 83.7 per cent, a difference of 5.03 percentage points, P = 0.068), and no laboratory-measured sleep outcome had improved by 26 weeks. The authors are candid about why: on objective testing the participants were already sleeping close to normally, despite reporting genuine clinical insomnia, so there was little room to improve.
What did change was everything the person actually experiences. At 26 weeks, insomnia severity, sleep quality, fatigue, the impact of RA on daily life, and depressive symptoms were all better with CBT-I, each highly statistically significant. Pain and patient global assessment improved too. Disease activity, physical function, and quality-of-life scores did not change. So CBT-I is not a treatment for your arthritis. It is a treatment for your sleep, your fatigue, and your mood, and those improved enough that the impact of the disease on daily life fell with them.
Almost everyone with insomnia has been handed a sleep hygiene list: consistent bedtime, dark room, no screens, less caffeine. Sleep hygiene is a real thing and it is a reasonable part of a bigger programme. On its own, it does not treat chronic insomnia.
The American Academy of Sleep Medicine has looked at this directly, and its position is not ambiguous. Its 2021 guideline on behavioural treatments gave sleep hygiene, used on its own, a conditional recommendation against, because of its lack of efficacy compared with other treatments. The same guideline gave multicomponent CBT-I a strong recommendation, on substantial evidence from multiple high-quality randomised trials. In the trials reviewed alongside it, CBT-I was superior to sleep hygiene alone, and sleep hygiene alone was less effective than sleep restriction therapy, stimulus control, or those combined. The AASM's 2026 guideline restates that position and goes further: it discourages clinicians from using sleep hygiene instructions alone even as a component of combination treatment.
The review also names the harm, which is the part that matters here: using sleep hygiene alone may delay effective treatment while insomnia continues or worsens, and someone whose sleep hygiene fails may reasonably conclude that behavioural treatment does not work for them and decline the treatment that would have. If you have tried the tips and they did not work, that is the expected result, not your failure.
The American Academy of Sleep Medicine published a guideline in 2026 on combining CBT-I with medication. Its two recommendations, both conditional and both on low-certainty evidence, are worth knowing because together they set an order. It suggests CBT-I plus a sleep medication over medication alone. It suggests against CBT-I plus medication over CBT-I alone. Their own summary of what that means: CBT-I above combination, and combination above medication alone.
The guideline adds a remark worth carrying: someone who places a high value on getting more total sleep early on, or a lower value on improving daytime symptoms, may reasonably choose the combination instead. That is a preference, not a error, and it is the sort of thing to decide with your clinician rather than alone.
On the medicines themselves, the 2026 guideline records that the AASM's 2017 medication guideline said sleep medicines should be considered mainly for people who cannot take part in CBT-I, who still have symptoms after a proper trial of it, or as a temporary addition to it. It also records that the same guideline gave conditional recommendations against trazodone, tiagabine, diphenhydramine, melatonin, tryptophan, and valerian, on insufficient evidence of efficacy, absence of high-quality data, or other considerations. Several of those are exactly what people buy for themselves: the antihistamine in over-the-counter sleep aids is diphenhydramine, and melatonin and valerian are on the same list. That does not make them dangerous; it means the evidence behind them is not what the marketing implies. Bring what you are taking to your appointment, including anything from a pharmacy shelf.
The 2026 guideline is endorsed by the Canadian Sleep Society, so it is reasonable guidance to raise here in Ontario.
Sleep apnoea is worth naming. In the Danish RA trial, 40 per cent of participants turned out to have obstructive sleep apnoea on their baseline sleep study, and all of them were referred on. The authors specifically suggest screening for sleep apnoea in people with RA who stay exhausted despite their arthritis being well controlled. If you snore, wake unrefreshed, or your partner has noticed you stop breathing, say so.
Steroids are worth naming too: if you take prednisone, ask your team about timing, because sleep disturbance is one of the effects people on steroids rate as most bothersome.
The 2023 EULAR fatigue recommendations set out four things, with their evidence grades:
Health professionals should build regular assessment of fatigue severity, impact, and coping strategies into consultations.
Level of evidence 5, grade D (expert opinion)
Offer access to tailored physical activity and encourage long-term activity. The effect on fatigue was small in RA and large in spondyloarthritis.
Level of evidence 1a, grade A
Offer access to structured, tailored psychoeducational programmes, which go beyond information alone.
Level of evidence 1a, grade A
New or worsening fatigue should trigger evaluation of disease activity and consideration of starting or changing immunomodulatory treatment.
Level of evidence 1a, grade A
On CBT-I, EULAR is careful, and we are repeating their care rather than overselling. They name cognitive behavioural therapy for insomnia as an example of something with insufficient evidence at guideline level across inflammatory rheumatic disease, but which might help an individual, within a stepped model of care where the least resource-intensive effective option comes first. Their research agenda still lists sleep hygiene training among the single components whose efficacy needs establishing. In other words: the RA trial above is encouraging and CBT-I is first-line for insomnia in general, but EULAR has not yet graded it for inflammatory arthritis as a whole.
Tell your rheumatology team that you are not sleeping, and describe the pattern rather than just saying you are tired: when you wake, whether you can get back to sleep, whether moving helps. Ask specifically whether CBT-I is available to you, in person or through a digital programme. Ask about sleep apnoea if you snore or wake unrefreshed. Ask about steroid timing if you take prednisone. And if fatigue is new or worse, ask whether your disease needs looking at, because the guideline says it should be.