Counts describe this library, not the strength of evidence.
How to read this evidence
Sleep health has several dimensions. Regularity cohorts and sleep-extension trials answer different questions; a prediction from observational data is not proof that changing a score prevents disease. Exercise effects also differ between sleep quality and daytime sleepiness.
A literature review ranked sleep detection methods by accuracy as questionnaire, sleep diary, contactless devices, contact devices, then polysomnography, with actigraphy sensitivity above 90% but low specificity. The ranking comes from comparing validation studies of differing designs, and many consumer devices and apps have not been validated at all.
In 51 people with current suicidal ideation followed for a week, shorter sleep (objective and self-reported) and poorer sleep quality predicted more severe suicidal thoughts the next day, but daytime thoughts did not predict the following night's sleep. The sample was small and week-long, so this shows a short-term pattern, not that fixing sleep prevents suicidal thoughts.
In 69 children with autism and 62 age-matched controls, touch hypersensitivity had the strongest link to sleep disturbance in the autism group, explaining 24% of the variance in total sleep disturbance scores. The data come from parent questionnaires at one point in time, so they show association only and not that touch sensitivity disrupts sleep.
In 1728 children aged 9-11 from four countries, both modelling approaches estimated higher body fat percentage when time was moved from moderate-to-vigorous physical activity (MVPA) to any other behaviour. The cross-sectional design cannot show that changing time use alters body fat, and the abstract gives no effect sizes.
In 5151 Dutch children followed to age 9, early sleep problems did not predict later autistic traits once earlier traits were accounted for, but autistic traits and an autism diagnosis (81 children) were linked to more sleep problems later. Only 81 children had a diagnosis, so the findings describe patterns, not causes.
In a meta-analysis of randomized controlled trials, Baduanjin, a mindfulness-based Chinese exercise, reduced musculoskeletal pain (SMD -0.88, 95% CI -1.02 to -0.74) and improved sleep quality (SMD -0.48, 95% CI -0.95 to -0.01) in people with chronic illness. The sleep result was very inconsistent between trials (I² = 84.42%) and the confidence interval just reaches zero.
International journal of environmental research and public health
In a crossover trial of 33 people with traumatic brain injury and sleep problems, 2 mg prolonged-release melatonin lowered Pittsburgh Sleep Quality Index scores versus placebo (7.68 vs 9.47; difference -1.79; 95% CI -2.70 to -0.88), meaning better sleep quality, but it did not change sleep onset latency. The trial was small, lasted 4 weeks per treatment, and ran at two Melbourne hospitals.
Pooling 21 studies, higher alcohol consumption was associated with a 25% higher relative risk of sleep apnoea (RR 1.25, 95% CI 1.13-1.38), but studies disagreed a lot (I2 = 82%) and there was evidence of publication bias (p = 0.001). Because the studies are observational, the link may partly reflect other factors and cannot prove that drinking causes apnoea.
In 42 healthy, normal-weight adults who usually slept 5 to under 7 hours, a 4-week sleep hygiene consultation increased sleep duration by 0:21 (h:mm) versus controls and lowered free sugar intake (-9.6 g versus 0.7 g in controls; P = 0.042). It was a small pilot with self-reported food diaries, and energy balance and cardiometabolic markers did not differ.
Pooling 16 prospective cohort studies of 24,308 people, severe obstructive sleep apnoea was linked to about double the risk of major adverse cardiac events (RR 2.04, 95% CI 1.56 to 2.66) and stroke (RR 2.15, 95% CI 1.42 to 3.24), while mild apnoea showed no significant link. These are observational studies, so the associations do not prove that apnoea itself causes the events.
In 695 people with newly diagnosed sleep apnea followed for a median of 3.0 years, adherence to CPAP was 89% in severe, 71% in moderate and 55% in mild disease. Higher apnea severity and sleepiness were independently linked to staying on CPAP and smoking to stopping, but this retrospective study of one clinic cannot show causes.
In 552 Canadian children aged 3 to 4, the overall composition of sleep, sedentary time and physical activity was significantly associated with BMI z-scores (p = 0.006) but not with waist circumference (p = 0.718). No single behavior, relative to the others, was significantly associated with either measure, and sleep was reported by parents, not measured.
Across 69 articles covering 148,524 children aged 0 to 4, shorter sleep was associated with higher adiposity (20/31 studies), poorer emotional regulation (13/25), impaired growth (2/2), more screen time (5/5) and more injuries (2/3). Most studies were cross-sectional and evidence quality ranged from very low to high, so these are associations, not proof that sleep is the cause.
Among 3755 UK university students with insomnia, digital cognitive behavioural therapy (CBT) for insomnia reduced insomnia, paranoia and hallucinations at 10 weeks versus usual practice (insomnia adjusted difference 4.78, 95% CI 4.29 to 5.26). The paranoia and hallucination effects were small (Cohen's d 0.19 and 0.24), and the authors say results in non-student groups remain untested.
In a survey of 4951 clinical nurses in Mainland China general hospitals, 63.9% (3163) scored 5 or higher on the Pittsburgh Sleep Quality Index, and frequent night shifts and high occupational stress were among the factors linked to it. A one-time self-report survey cannot show that these factors cause poor sleep.
Across 37 studies, elite athletes showed high levels of sleep complaints, including longer time to fall asleep, fragmented sleep and daytime fatigue. The quality of the evidence was generally low, so how common insomnia really is in elite sport remains uncertain.
Among 61 Harvard undergraduates tracked for about 30 days, the most irregular sleepers had melatonin onset at 00:08 versus 21:32 in the most regular sleepers (p < 0.003), and sleep regularity correlated with grade point average (r = 0.37, p < 0.004). The sample was small and the design observational, so the data cannot show that irregular sleep lowers grades or shifts the body clock.
In a British birth cohort of 2232 twins, loneliness was associated with worse overall sleep quality, especially subjective sleep quality and daytime dysfunction, and the link held within monozygotic twin pairs. The abstract gives no effect sizes, and sleep was self-reported at one time point, so the direction of the link is unclear.