Sleep and Depression: the bidirectional vicious circle and how to break it
Poor sleep is not only a symptom of depression. It is often its forerunner and a treatment lever in its own right. Whoever treats the sleep may treat the depression along with it and, under certain circumstances, even prevent it. What the evidence really shows, and what you can concretely do.
In the consultation I often hear the sentence: "Once the depression is gone, the sleep will sort itself out." That is only half the truth. Insomnia is not merely an appendage of depression that disappears on its own. It frequently precedes depression, it sustains it, and it often persists once the mood has already improved. Baglioni 2011 in Journal of Affective Disorders showed in a meta-analysis of 21 longitudinal studies that people with insomnia have about double the risk of developing depression. And Irwin 2022 in JAMA Psychiatry showed that treating insomnia may halve the new onset of depression in older people. Sleep is not a passive symptom. Sleep is a treatment lever. In this spoke I show why that is so and what follows from it.
This spoke is the interface between sleep medicine and mental health in the sleep cluster. We work through the bidirectional connection (insomnia as risk and as symptom), look at the REM sleep changes in depression, clarify why treating sleep may improve depression, and present CBT-I as therapy and as prevention. Added to this are the four KPNI lenses, a collection of typical thinking errors and three concrete levers for the coming weeks.
The bidirectional connection: chicken or egg?
The relationship between sleep and depression runs in both directions. That is the most important idea of this text. On the one hand, insomnia is a core symptom of depression and part of the diagnostic criteria. Almost every depressive episode is accompanied by a sleep disorder, whether as difficulty falling and staying asleep, as early-morning waking or as an increased need for sleep. On the other hand, insomnia is an independent risk factor that often precedes depression by months to years.
This second direction is frequently underestimated in practice. Whoever sleeps poorly does not only suffer. They carry a measurably increased risk of becoming depressed later. This is precisely what makes sleep such an interesting therapeutic entry point: it is potentially accessible earlier than the depression itself.
Insomnia doubles the risk of later depression
Meta-analysis Chiara Baglioni and colleagues from the Freiburg sleep research group around Dieter Riemann published in 2011 in Journal of Affective Disorders a meta-analysis of longitudinal studies on the question of whether insomnia predicts later depression. 21 studies met the inclusion criteria. In the random-effects model, an odds ratio of 2.60 (confidence interval 1.98 to 3.42) emerged for insomnia at baseline predicting depression over the course. Adjusted for outliers and in the fixed-effects model, the odds ratio was 2.10. The authors' core statement: non-depressed people with insomnia have about double the risk of developing depression compared with people without sleep problems. From this they conclude that an early treatment of insomnia may be a sensible preventive strategy for mental health.
Baglioni C, Battagliese G, Feige B, et al. J Affect Disord. 2011;135(1-3):10-19. doi:10.1016/j.jad.2011.01.011 · PMID: 21300408
Replication confirms: insomnia as a transdiagnostic risk factor
Meta-analysis Elisabeth Hertenstein and colleagues from Bern and Freiburg published in 2023 in Journal of Sleep Research a replication and update of the earlier systematic review on the longitudinal association between insomnia and mental disorder, with works from 2018 to 2022. The meta-analysis confirmed the earlier observation and even saw a somewhat stronger effect for the association between insomnia and depression. The authors classify insomnia as a possible transdiagnostic process in psychopathology, with corresponding clinical consequences for early detection and treatment.
Hertenstein E, Benz F, Schneider CL, Baglioni C. J Sleep Res. 2023;32(6):e13930. doi:10.1111/jsr.13930 · PMID: 37211915
Disrupted sleep increases the risk of depression in later life
Meta-analysis Lydia Hill Almeida and colleagues from Western Australia published in 2022 in Journal of Affective Disorders a prospective cohort study in older men with up to 17 years of follow-up, combined with a systematic review and meta-analysis. In their own cohort, the hazard ratio for a newly occurring depression in people with sleep problems was 1.67. Statistical adjustments for age, education, smoking and frailty did not substantially change the effect. The meta-analysis of a total of 15 studies yielded a pooled risk ratio of 1.82. The authors emphasize that this association does not appear to be explainable solely by reverse causality, and that older people with sleep problems are a legitimate target for preventive measures.
Hill Almeida LM, Flicker L, Hankey GJ, et al. J Affect Disord. 2022;309:314-323. doi:10.1016/j.jad.2022.04.133 · PMID: 35490880
A double risk sounds dramatic, but it is not fate. It does not mean that every sleepless person becomes depressed. It means that insomnia is a changeable entry point. Whoever addresses sleep early and in a targeted way can turn a dial that lies upstream of the depression. That is an encouraging message, not a threatening one.
What changes in REM sleep in depression
In the sleep laboratory, depression shows a recurring pattern. Three changes are especially well described: a shortened REM latency, meaning the first REM sleep sets in earlier than normal. An increased REM density, that is more rapid eye movements per unit of time. And a reduced amount of deep sleep, the restorative slow-wave sleep. These findings have been the subject of biological depression research for decades.
REM and sleep abnormalities are replicable, but not specific enough
Meta-analysis Cynthia Arfken and colleagues from Wayne State University published in 2014 in Journal of Affective Disorders a literature review with meta-analysis on the question of whether polysomnographic abnormalities (REM density, REM latency, sleep efficiency, deep sleep, stage 1 and 2) are suitable as a diagnostic test for major depression. 31 works met the inclusion criteria. Most abnormalities occurred in the expected direction compared with healthy people, with moderate effect size, but with considerable publication bias and heterogeneity. The authors concluded: the sleep findings are replicable, but studies that cleanly define sensitivity and specificity are lacking. A clinically usable diagnostic test from the sleep laboratory therefore does not yet exist.
Arfken CL, Joseph A, Sandhu GR, Roehrs T, Douglass AB, Boutros NN. J Affect Disord. 2014;156:36-45. doi:10.1016/j.jad.2013.12.007 · PMID: 24412322
The practical lesson from this finding: REM changes are a fascinating biological marker and a hint that sleep and mood regulation are neurobiologically closely interwoven. But they are not suitable as a sole diagnostic substitute. Depression is diagnosed clinically, not via the sleep curve. Whoever wants to sell you a sleep laboratory as a depression test is oversimplifying.
Why treating sleep improves depression
Here the arc becomes practical. If insomnia can precede depression and sustain it, then the question naturally arises: does depression improve when you specifically treat the insomnia? The answer has become considerably clearer over the past years, and it is: yes, in many cases.
CBT-I improves mood even without a depression focus
RCT Colleen Carney and colleagues published in 2017 in Sleep a blinded randomized study in 107 people with major depression and insomnia. Three groups: antidepressant plus CBT-I, CBT-I plus placebo tablet, or antidepressant plus sleep hygiene control. Remarkably: only the CBT-I groups improved on the objective sleep measure, while the sleep hygiene control worsened. On the depression scale (Hamilton), all groups improved, even the group with CBT-I plus placebo, which therefore received no depression-specific treatment at all. The authors conclude that the antidepressant property of CBT-I should be investigated further.
Carney CE, Edinger JD, Kuchibhatla M, et al. Sleep. 2017;40(4):zsx019. doi:10.1093/sleep/zsx019 · PMID: 28199710
CBT-I may improve depression beyond the sleep domain
Meta-analysis Yuki Furukawa and colleagues from Tokyo, with Michael Perlis from Philadelphia, published in 2024 in Journal of Affective Disorders a systematic review with meta-analysis of cognitive behavioral therapy for insomnia in people with major depression and comorbid insomnia. 19 randomized studies with 4808 participants were included. CBT-I was superior to the control conditions: for the depression response an odds ratio of 2.28 emerged (confidence interval 1.67 to 3.12, GRADE certainty moderate), for insomnia remission an odds ratio of 3.57. With a response rate of 17 percent in the control group, CBT-I reached a response rate of 32 percent. The authors emphasize that the depression improvement showed itself beyond the pure sleep domain, and conclude that CBT-I may be an effective treatment for the depression associated with insomnia.
Furukawa Y, Nagaoka D, Sato S, et al. J Affect Disord. 2024;367:359-366. doi:10.1016/j.jad.2024.09.017 · PMID: 39242039
An important nuance is provided by Blom 2024 in Psychother Psychosom: in a double-blind RCT in 126 people with insomnia and depression, CBT-I showed strong specific effects on insomnia. However, the combination of CBT-I with an additional depression therapy brought no added value for the depression compared with the control. The authors recommend always offering CBT-I for the combination of insomnia and depression, ideally as first choice, without necessarily overloading it with a further therapy.
CBT-I as prevention: preventing depression
Perhaps the strongest finding of this entire field concerns not the treatment of an existing depression, but prevention. If insomnia precedes depression, then early treatment of insomnia should be able to delay or prevent the onset of depression. That is exactly what was tested.
CBT-I halves the new onset of depression in older people
RCT Michael Irwin and colleagues from the Cousins Center for Psychoneuroimmunology at UCLA published in 2022 in JAMA Psychiatry a randomized study on the question of whether treating insomnia may prevent the new onset of depression. 291 people aged 60 years and over with insomnia disorder, but without current depression, were randomized: two months of CBT-I or sleep education as an active comparison condition. Over the follow-up of up to three years, a first or recurrent major depression occurred in 12.2 percent of the CBT-I group, compared with 25.9 percent in the comparison group (hazard ratio 0.51, confidence interval 0.29 to 0.88). In participants whose insomnia went into sustained remission, the depression risk was even reduced by over 80 percent. The authors conclude that broad availability of CBT-I could substantially advance public health in older people.
Irwin MR, Carrillo C, Sadeghi N, Bjurstrom MF, Breen EC, Olmstead R. JAMA Psychiatry. 2022;79(1):33-41. doi:10.1001/jamapsychiatry.2021.3422 · PMID: 34817561
This finding is clinically significant because it lifts insomnia out of the role of a mere accompanying symptom. Sleep becomes here a preventive lever for mental health. A detailed guide to the method itself you will find in the sister spoke on cognitive behavioral therapy for insomnia.
Treating sleep is not symptom care, but mental health prevention.
Whoever addresses insomnia early turns a dial that lies upstream of the depression. This may improve mood along with it and, under certain circumstances, prevent a depressive episode.
The 4 KPNI lenses on sleep and depression
Nervous system
Depression and insomnia share the pattern of hyperarousal: a persistently activated stress and wake system that makes falling asleep harder and reduces deep sleep. The shortened REM latency after Arfken 2014 is an expression of this altered sleep-wake regulation. The lever is to wind the nervous system down in the evening rather than activating it further.
Immune system
Sleep deprivation and depression both go along with low-grade inflammatory activity. The research field of psychoneuroimmunology, from which the Irwin group also comes, examines how disrupted sleep triggers inflammatory signaling pathways that in turn may influence mood. Restorative sleep may potentially be inflammation-dampening here.
Hormonal system
The cortisol daily rhythm is often shifted in depression and insomnia: elevated evening values, a flattened daily course. The melatonin signal and the day-night rhythm may also be altered. A stable light-dark rhythm with morning daylight is a simple entry point to set the timing of both systems.
Metabolism
Poor sleep worsens insulin sensitivity and energy metabolism, and a disturbed metabolism may in turn burden sleep and mood. Hypersomnia and the lack of drive of atypical depression have points of contact here. Regular exercise and stable meal times support sleep and mood at the same time.
What does not work: typical thinking errors about sleep and depression
That is only half correct. Insomnia frequently persists, even when the mood has improved, and a remaining insomnia increases the risk of relapse. Baglioni 2011 and Irwin 2022 both show that insomnia is an independent factor. Sleep should be specifically treated as well, not passively waited out.
Classic sleeping pills may make falling asleep easier in the short term, but they do not treat the underlying chronic insomnia and carry risks of habituation and daytime tiredness. The guidelines see CBT-I, not long-term medication, as the first-line treatment for chronic insomnia. A detailed weighing-up you will find in the sister spoke on sleeping pills.
The opposite is often the case. Long times lying down without sleep destroy the association between bed and sleep and reinforce the insomnia. Sleep restriction, a core building block of CBT-I, goes the opposite way and initially shortens the time in bed. This is counterintuitive, but effective.
REM changes are described in depression, but Arfken 2014 shows that they are not specific enough for a standalone diagnostic test. Depression is diagnosed clinically. A sleep laboratory is useful for ruling out other sleep disorders such as sleep apnea, not for proving depression.
In atypical depression, hypersomnia often dominates, that is increased but unrefreshing sleeping, and the fleeing into sleep as withdrawal. More sleep does not automatically improve mood here. What matters is restorative sleep in a stable rhythm, not the pure amount of sleep.
Three concrete levers for the coming weeks
Keep a sleep and mood diary over 14 days
Note daily: time in bed and estimated sleep time, nightly wakefulness, mood in the morning and in the evening (scale 0 to 10), drive and daylight exposure. This diary makes the connection between sleep and mood visible and is the basis of any targeted treatment. It costs nothing and often delivers more insight than a single sleep laboratory.
Consider CBT-I as a first-line option, not the sleeping pill
If the insomnia is chronic (more than three months, more than three nights per week), cognitive behavioral therapy for insomnia is the evidence-based first-line treatment. Furukawa 2024 and Irwin 2022 show that it may also work on mood. Talk with your physician about CBT-I, digital CBT-I programs or a referral to sleep medicine. You will find details in the sister spoke on CBT-I.
Stabilize the day-night rhythm
Bright daylight in the morning, getting up at the same time as far as possible, dimmed light and screen reduction in the evening. A stable light-dark rhythm sets the timing of cortisol and melatonin and supports sleep and mood at the same time. This does not replace therapy, but creates the basis on which every further treatment takes hold better.
This text is general information and does not replace a medical examination. Sleep disorders and depressive complaints belong in medical or psychotherapeutic evaluation, especially when they last longer than two to four weeks, when sleep is not refreshing despite sufficient time in bed, or when drive and joy of life clearly diminish. Never stop prescribed medication on your own.
If you have suicidal thoughts or the feeling of no longer wanting to live, please get help immediately. A crisis helpline is available around the clock free of charge in Germany under 0800 111 0 111 and 0800 111 0 222. In an acute emergency situation, please dial the local emergency number, in Germany 112, without delay.
Frequently asked questions about sleep and depression
Is sleep in depression a cause or a consequence?
Both. The relationship is bidirectional. On the one hand, insomnia is a core symptom of depression and anchored in the diagnostic criteria. On the other hand, insomnia is an independent risk factor that often precedes depression. Baglioni 2011 in Journal of Affective Disorders summarized 21 longitudinal studies and found that people with insomnia and without current depression have about double the risk of later developing depression (odds ratio 2.60, confidence interval 1.98 to 3.42). Hertenstein 2023 in Journal of Sleep Research replicated this finding. In practical terms this means: sleep is not just a symptom that disappears on its own once the depression is treated. Sleep is a treatment lever in its own right.
Does insomnia really increase the risk of depression?
Yes, this is one of the most robust findings in sleep medicine. Baglioni 2011 in Journal of Affective Disorders showed in a meta-analysis of 21 longitudinal studies an odds ratio of 2.60 for the new onset of depression in people with insomnia. Hill Almeida 2022 in Journal of Affective Disorders found in its own cohort plus a meta-analysis of 14 further studies a pooled risk ratio of 1.82. Hertenstein 2023 confirmed the association in a replication and even saw a somewhat stronger effect. Important: an increased risk is not an inevitability. It means that treating insomnia may be a sensible preventive strategy, not that every sleepless person becomes depressed.
What changes in REM sleep in depression?
In depression, three patterns have been described in sleep laboratory studies: a shortened REM latency (REM sleep sets in earlier than normal), an increased REM density (more rapid eye movements per unit of time) and a reduced amount of deep sleep. Arfken 2014 in Journal of Affective Disorders examined in a meta-analysis of 31 works whether these polysomnographic abnormalities are suitable as a diagnostic test for major depression. Result: the findings go in the expected direction and are replicable with a moderate effect, but they are not specific enough to form a clinical diagnostic test. REM changes are therefore an interesting biological marker, not a standalone diagnostic substitute.
Does depression improve when you treat sleep?
Yes, the evidence for this has become considerably stronger in recent years. Furukawa 2024 in Journal of Affective Disorders analyzed 19 randomized controlled trials with 4808 participants and found that cognitive behavioral therapy for insomnia (CBT-I) in people with major depression and comorbid insomnia may more than double the depression response compared with control conditions (odds ratio 2.28). The depression improvement showed itself beyond the pure sleep domain. Already Carney 2017 in Sleep had shown that even a CBT-I without a depression-specific component can lower depression scores. Treating sleep is therefore not only symptom relief, but may be an antidepressant lever in its own right.
What is CBT-I and why is it useful in depression?
CBT-I is cognitive behavioral therapy for insomnia, the structured and guideline-based first-line treatment for chronic insomnia. It combines several building blocks: sleep restriction (time in bed is initially shortened and adjusted to the actual sleep time), stimulus control (the bed is re-associated with sleep), cognitive work on dysfunctional sleep thoughts, relaxation techniques and sleep hygiene. In depression, CBT-I is doubly useful: it improves insomnia and may help improve depression as well. Furukawa 2024 in Journal of Affective Disorders showed the dual effect in a meta-analysis. You will find a detailed account of the method in the sister spoke on CBT-I.
Can treating the sleep disorder prevent depression?
There is one of the most convincing pieces of evidence in prevention research for this. Irwin 2022 in JAMA Psychiatry treated 291 older people over 60 years with insomnia, but without current depression, randomized either to CBT-I or to sleep education as an active comparison condition. Over the follow-up, depression occurred in 12.2 percent of the CBT-I group, compared with 25.9 percent in the comparison group (hazard ratio 0.51). With sustained remission of insomnia, the depression risk was even reduced by over 80 percent. This is a strong argument for taking insomnia seriously and treating it early, not only once depression has become manifest.
Is too much sleep in depression also a problem?
Yes. Depression presents not only as insomnia, but in some of those affected as hypersomnia, that is an increased need for sleep, sleeping for long periods and still unrefreshing sleep. This is typical in so-called atypical depression. The phenomenon of fleeing into sleep, that is sleep as avoidance and withdrawal, also belongs in this picture. The distinction is important: unrefreshing sleep despite long time in bed may point to depression, but also to other causes such as sleep apnea or thyroid disorders. A medical evaluation of the sleep architecture and the differential diagnoses is sensible before lengthening or shortening time in bed across the board.
When should I seek medical help for sleep problems and low mood?
When the sleep disorder lasts longer than three to four weeks, when mood is persistently low for at least two weeks, when interest and joy are lost, when drive clearly diminishes, or when sleep is not refreshing despite sufficient time in bed. Urgent and immediate: if you have thoughts of no longer wanting to live, or suicidal thoughts, please seek medical or psychotherapeutic help without delay, for example via a crisis helpline, or in an emergency via the local emergency number. Sleep problems and their connection to mood can be treated well, but they do not replace a medical examination. This text is information, not a diagnosis.
More from the cluster "Treating sleep disorders holistically"
- Pillar: Treating sleep disorders holistically
- Spoke 9: CBT-I for insomnia
- Spoke 14: Sleep and Depression (you are here)
Connections to other topics
The pillar article places insomnia, sleep phases, diagnosis and treatment into the overall context. Every sleep path starts here.
The detailed guide to cognitive behavioral therapy for insomnia: sleep restriction, stimulus control, cognitive work. The toolkit behind the studies in this spoke.
How depression can be understood and treated beyond sleep: nervous system, immune system, lifestyle. Sleep is one building block of several.
Why classic sleeping pills do not solve chronic insomnia and what role medication realistically plays. Linked in the pillar.
Sources and further reading
- Baglioni C, Battagliese G, Feige B, Spiegelhalder K, Nissen C, Voderholzer U, Lombardo C, Riemann D. Insomnia as a predictor of depression: a meta-analytic evaluation of longitudinal epidemiological studies. J Affect Disord. 2011;135(1-3):10-19. doi:10.1016/j.jad.2011.01.011 · PMID: 21300408 [Meta-analysis]
- Hertenstein E, Benz F, Schneider CL, Baglioni C. Insomnia - A risk factor for mental disorders. J Sleep Res. 2023;32(6):e13930. doi:10.1111/jsr.13930 · PMID: 37211915 [Meta-analysis]
- Hill Almeida LM, Flicker L, Hankey GJ, Golledge J, Yeap BB, Almeida OP. Disrupted sleep and risk of depression in later life: A prospective cohort study with extended follow up and a systematic review and meta-analysis. J Affect Disord. 2022;309:314-323. doi:10.1016/j.jad.2022.04.133 · PMID: 35490880 [Meta-analysis]
- Arfken CL, Joseph A, Sandhu GR, Roehrs T, Douglass AB, Boutros NN. The status of sleep abnormalities as a diagnostic test for major depressive disorder. J Affect Disord. 2014;156:36-45. doi:10.1016/j.jad.2013.12.007 · PMID: 24412322 [Meta-analysis]
- Carney CE, Edinger JD, Kuchibhatla M, Lachowski AM, Bogouslavsky O, Krystal AD, Shapiro CM. Cognitive Behavioral Insomnia Therapy for Those With Insomnia and Depression: A Randomized Controlled Clinical Trial. Sleep. 2017;40(4):zsx019. doi:10.1093/sleep/zsx019 · PMID: 28199710 [RCT]
- Furukawa Y, Nagaoka D, Sato S, Toyomoto R, Takashina HN, Kobayashi K, et al. Cognitive behavioral therapy for insomnia to treat major depressive disorder with comorbid insomnia: A systematic review and meta-analysis. J Affect Disord. 2024;367:359-366. doi:10.1016/j.jad.2024.09.017 · PMID: 39242039 [Meta-analysis]
- Blom K, Forsell E, Hellberg M, Svanborg C, Jernelöv S, Kaldo V. Psychological Treatment of Comorbid Insomnia and Depression: A Double-Blind Randomized Placebo-Controlled Trial. Psychother Psychosom. 2024;93(2):100-113. doi:10.1159/000536063 · PMID: 38286128 [RCT]
- Irwin MR, Carrillo C, Sadeghi N, Bjurstrom MF, Breen EC, Olmstead R. Prevention of Incident and Recurrent Major Depression in Older Adults With Insomnia: A Randomized Clinical Trial. JAMA Psychiatry. 2022;79(1):33-41. doi:10.1001/jamapsychiatry.2021.3422 · PMID: 34817561 [RCT]
- Felder JN, Epel ES, Neuhaus J, Krystal AD, Prather AA. Efficacy of Digital Cognitive Behavioral Therapy for the Treatment of Insomnia Symptoms Among Pregnant Women: A Randomized Clinical Trial. JAMA Psychiatry. 2020;77(5):484-492. doi:10.1001/jamapsychiatry.2019.4491 · PMID: 31968068 [RCT]
- Chiu HY, Lee HC, Chen PY, Lai YF, Tu YK. Associations between sleep duration and suicidality in adolescents: A systematic review and dose-response meta-analysis. Sleep Med Rev. 2018;42:119-126. doi:10.1016/j.smrv.2018.07.003 · PMID: 30093362 [Meta-analysis]