All topics › Parent mental health relationships work
Does sleep deprivation after the birth of a child harm parents' mental health, and do strategies to protect sleep actually help?
Newborns wake a lot at night, and everyone warns you about the sleep deprivation. But is the broken sleep itself a risk to your mental health — and do strategies like taking shifts, protecting sleep windows, or therapy for insomnia actually protect parents from depression?
Poor sleep and low mood after a baby are strongly linked, but the link runs both ways — bad sleep can drag mood down, and low mood makes sleep worse [1][2]. The clearest evidence we have is indirect: therapy for insomnia (CBT-I) during pregnancy or after birth modestly improves both sleep and mood [3][4], and treating infant sleep problems at 7–12 months slightly improved mothers' mood scores [5]. But no good trial has tested whether the most common advice — taking shifts or protecting sleep windows in the newborn weeks — actually prevents depression. These are sensible, low-risk strategies; they just aren't proven treatments.
A meta-analysis of 19 laboratory studies (1,932 participants) found that mood was more impaired by sleep deprivation than cognitive or motor performance — and that partial sleep deprivation, the kind parents experience, had large effects on mood [1]. This is lab evidence in general adults, not real postpartum sleep, but it establishes the mechanism: fragmented, insufficient sleep reliably makes people feel worse.
The longitudinal picture is messier. A meta-analysis of 21 general-population studies found insomnia predicted later depression — but the authors warned that intervening variables were inadequately controlled, so it can't establish how much is cause and how much is shared circumstance [2]. In perinatal populations specifically, poor sleep and depression were consistently associated in one review [6]. These are observational findings, and sleep and mood almost certainly reinforce each other in both directions.
In a study of 72 couples measured with wrist actigraphs late in pregnancy and at one month postpartum, both parents had more disrupted sleep and more fatigue after the birth. Fathers objectively slept less in total than mothers at both time points, while mothers reported more disruption [7]. Employment and feeding patterns mattered. A longitudinal study through the first year found at least subthreshold insomnia symptoms in around 70% of mothers (falling to 60% by 12 months) and about 50% of fathers [8]. These are questionnaire scores in one research cohort, not UK diagnosis rates — but they show sleep loss after a baby is a two-parent problem, even if it doesn't always look the same.
CBT for insomnia targets the inability to sleep when the opportunity exists — racing thoughts, not being able to fall back asleep after a feed — not the baby's unavoidable wakings. Keeping that distinction matters for interpreting the trials below.
In a randomised trial of 179 pregnant participants with insomnia, CBT-I produced insomnia remission in 64% versus 52% of controls (120 more per 1,000), with a median time to remission of 31 versus 48 days, plus a small improvement in depressive symptoms [3]. Self-reported night-time wakefulness improved, though objective total wake time did not — the treatment changed the experience of wakefulness more than its amount. Benefits persisted into the postpartum period for wakefulness excluding infant-care time [4].
A digital CBT-I trial in pregnancy found improvements in insomnia, depressive and anxiety symptoms — but participants' baseline mood symptoms were mild, so this does not prove it treats or prevents clinical depression [9]. A three-arm trial (n=127) comparing prenatal/postpartum CBT-I against a responsive bassinet and sleep-hygiene control found postpartum insomnia scores lower with CBT-I (effect size 0.56); the bassinet was not significantly better than the control (effect size 0.25), with mood outcomes only exploratory [10]. The bassinet result is a useful caution: marketed devices don't reliably protect parental sleep.
A cluster randomised trial across 49 maternal-child-health centres (328 mothers whose infants had sleep problems at 7 months) tested a behavioural sleep consultation against usual care. Infant sleep problems fell from 68% to 56% at 10 months and from 55% to 39% at 12 months, with mothers' depression scores modestly lower (mean EPDS difference −1.4 at 10 months, −1.7 at 12 months) and better mental-health quality-of-life scores [5]. This supports the idea that improving infant sleep can lift maternal mood — but it was tested at 7–12 months, not in the newborn weeks, and says nothing about sleep shifts.
No robust randomised trial has tested whether partner shifts, protected sleep windows, or similar newborn-care arrangements prevent postnatal depression. A 2021 hospital quality-improvement project and a small 2025 conference abstract (41 mothers, Fitbit data, with a commercial conflict of interest) are all we have — and they are not trial evidence [11]. This is the biggest gap between what parents are told and what has been tested.
For mothers: the evidence that broken sleep drags mood down is consistent, even if the exact causal share is unknown [1][2]. If you lie awake even when the baby sleeps — mind racing, unable to drop off — that's insomnia territory, and it's specifically the kind of problem CBT-I treats [3]. You don't have to earn help by being at rock bottom: perinatal mental-health support in the UK (GP, health visitor, talking therapies, and specialist perinatal mental-health teams) can be accessed well before things become severe.
For fathers and partners: you are not a bystander in this evidence. Fathers slept less in total than mothers in the actigraphy study, and around half of fathers had subthreshold insomnia symptoms across the first year [7][8]. Paternal mood is covered in our postpartum-depression topic — but know that your sleep matters too, and it's badly understudied. If you go back to work early, the sleep burden doesn't vanish; it collides with safety-critical tasks like driving and operating machinery, which is where sleep-deprivation research outside parenting is most alarming [1].
For your mental health: treat persistent low mood, loss of pleasure, or hopelessness as a health problem in its own right, not as something that will automatically fix itself once the baby sleeps through. The sleep–mood link runs both ways [2], so addressing both — sleep strategies and mood support — is more sensible than waiting on one.
For sleep itself: practical steps like alternating first/second shifts where feeding and recovery allow, one parent being genuinely off-duty (earplugs, separate room), and accepting offered help are reasonable and low-risk. What they are not: proven depression prevention [11]. Be honest with yourselves about whether a strategy is actually producing sleep, not just time in bed.
For relationships: exhaustion makes everything harder — patience, generosity, communication. Our couple-relationship topic covers the wider pattern. Sleep is plausibly part of why relationships strain, but that specific mediation hasn't been proven [12].
For work and finances: there is almost no direct evidence on how newborn-period sleep loss affects work performance, but the general sleep-deprivation literature shows mood and functioning degrade well before people notice [1]. If your job involves driving or safety-critical decisions, treat the early weeks like any other period of impairment: build in margins, don't push through.
For the care burden: none of these trials tested protected sleep in the context of exclusive breastfeeding, recovery from a difficult birth, or single parenting — the strategies that work for one household may be impossible in another. That's a gap in the science, and it's not your failure.
| Strategy | Outcome | Effect in absolute terms | Evidence rating |
|---|---|---|---|
| CBT-I in pregnancy | Insomnia remission | 640 vs 520 per 1,000 remit (120 more per 1,000); median 31 vs 48 days [3] | B |
| CBT-I in pregnancy/postpartum | Depressive symptoms | Small improvement; mild-symptom samples — not proven for clinical depression [3][9] | C |
| Infant-sleep intervention (7–12 months) | Infant sleep problems | 68% → 56% at 10 months; 55% → 39% at 12 months (120–160 fewer per 1,000) [5] | B |
| Infant-sleep intervention (7–12 months) | Maternal depressive symptoms | Mean EPDS difference −1.4 to −1.7 — modest [5] | B (for the modest effect) |
| Responsive bassinet | Parental insomnia | No significant benefit vs sleep hygiene [10] | B (for the null) |
| Partner shifts / protected sleep windows | Preventing depression | No direct trial evidence [11] | D (we don't know) |