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.
Pooling 23 randomized trials with 1,478 adults aged 45 to 77 who had sleep disorders, exercise improved Pittsburgh Sleep Quality Index scores (SMD 0.62, 95% CI 0.40–0.83), and aerobic exercise showed the largest effect in the network analysis (SMD 1.34, 95% CrI 0.86–1.82). Certainty was low to moderate, the outcomes were questionnaires, and the best weekly dose of about 539 MET-min/week is a model estimate, not a prescription.
Among 60,977 UK Biobank adults who wore a wrist accelerometer for a week, the most regular fifth of sleepers had a fully adjusted all-cause mortality HR of 0.70 (95% CI 0.59–0.83) against the least regular fifth, over about 6.3 years and 1859 deaths. Regularity fitted the data better than sleep duration, but the study is observational and rests on one week of sleep data.
In 80 adults with overweight who usually slept under 6.5 hours, two weeks of sleep-extension counselling added about 1.2 hours of sleep and lowered measured energy intake by 270.4 kcal/d versus control (95% CI −393.4 to −147.4). The trial was short, single-centre and in a narrow group, so it does not show lasting weight loss or any benefit for people who already sleep enough.
In 16,083 people tested for suspected sleep apnea, higher respiratory effort during sleep, measured as negative esophageal pressure, was associated with higher all-cause mortality over a median 15 years, independent of AHI and oxygen levels (third quartile HR 1.21, 95% CI 1.03-1.43). Esophageal pressure needs a catheter, and the abstract reports hazard ratios only for the third quartile.
American journal of respiratory and critical care medicine
Among 250 young people aged 12-30 at early-intervention services in Sydney, longer baseline sleep predicted transition to bipolar or psychotic disorder (OR 2.23, 95% CI 1.38-3.74; HR 2.05, 95% CI 1.23-3.40). The sample is small and help-seeking, and the later-midpoint finding only held before correction for multiple testing.
Across 7 trials with 709 people, high-frequency rTMS aimed at the left prefrontal cortex improved sleep quality (SMD 0.97) and depressive symptoms (SMD 0.61) compared with controls. Evidence was very low to low certainty, with large heterogeneity for sleep quality (I2 = 91%), so the authors call it exploratory.
Among 41 people having overnight sleep studies, each 1 percentage-point higher sleep efficiency was associated with a lower arrhythmia rate (IRR 0.946, 95% CI 0.921-0.972, p < 0.001), while apnea-hypopnea index (AHI) was not (IRR 1.000, p = 0.954). The model was chosen after looking at the data, so the authors call it exploratory and the p-values optimistic.
Among 537 pregnant women, smartphone addiction was associated with higher odds of moderate-to-severe depression (OR 3.461, 95% CI 1.771-7.100) and anxiety (OR 7.738, 95% CI 3.064-23.172), and poor sleep quality statistically accounted for part of these links. Data were collected at one time point, so the direction of effect is unknown, and the anxiety estimate is imprecise.
In 56 male adolescents followed for 9 days, better sleep quality was associated with lower irritability the next day (standardized β -0.92, 95% CI -1.31 to -0.53), and lower irritability was associated with better sleep the next night. The sample is small, all male and exploratory, so the findings are preliminary and show association only.
In 50 NCAA Division I athletes wearing WHOOP bands, 2 weeks of instructions to sleep 10 hours per night did not raise total sleep time (7.63 to 7.51 hours, P = 0.27) but did raise sleep consistency (60.09% to 64.42%, P = 0.02). There was no control group and the athletes were measured in the offseason, so the results may not carry over to the competitive season.
Across 9 studies with 2262 ultramarathon runners, more pre-race sleep, including sleep extension, was associated with faster finish times, and better pre-race sleep quality with higher cognitive performance. The evidence is observational, so it cannot show that extra sleep causes faster races.
Among 403 university students in Ghana, dry eye disease affected 77.9%, and poor sleep quality (adjusted RR 1.11) and high perceived stress (adjusted RR 1.06) were each independently associated with it. Because exposures and eye findings were measured at one point in time, the study cannot show which came first.
In interviews with 14 first-time parents in Ireland, four themes emerged: gendered sleep roles, difficulty sleeping, wellbeing struggles from sleep loss, and adapting. Because this is a small qualitative study, it describes experiences rather than measuring how common or how harmful sleep loss is.
Pooling 16 meta-analyses covering approximately 13.8 million adults, short sleep was associated with 40% higher odds of obesity (OR 1.40, 95% CI 1.27-1.55), graded only as suggestive evidence. These are observational associations with high heterogeneity (I2 = 83.7%), and most of the included reviews were low quality.
Among 257,509 UK Biobank participants, a higher seven-component sleep health score was associated with lower risks of coronary heart disease, stroke, atrial fibrillation, heart failure and death, strongest for heart failure and coronary heart disease. Sleep was self-reported once at baseline, and the protein findings, from 26,984 people, are exploratory rather than proof of mechanism.
This review of 46 studies found that accelerometer and gyroscope methods detected apnea and hypopnea events with sensitivities from 30% to 99%, with the best results mostly in simulation-based studies under controlled conditions. Only a minority used polysomnography as the reference or clinically representative patients, and risk of bias was high.
In 40 cognitively healthy adults aged 55 to 75 years, higher plasma amyloid-β (Aβ42 and Aβ40) was associated with shorter, less dense sleep spindles, and higher evening t-tau and p-tau181 with less N3 (deep) sleep. These are exploratory correlations in a small group and do not show that sleep changes cause, or predict, Alzheimer's disease.
Among 34,166 men in the Southern Community Cohort Study, sleeping 9 hours or more on weekends, versus 7-8 hours, was suggestively linked to higher prostate cancer-specific mortality (HR 1.44, 95% CI 1.02-2.03). With only 225 prostate cancer deaths and a borderline interval, the finding is imprecise and may not be causal.
In Canadian adults surveyed in 2017-2018, dry mouth and oral pain were linked to higher adjusted odds of insomnia symptoms, unrefreshing sleep and difficulty staying awake (AOR 1.701 to 1.937). Because this is a one-time survey with self-reported measures, it cannot show whether oral problems disturb sleep or the reverse.
In 21 NCAA Division I female athletes, three nights of a 40 g pre-sleep casein, α-lactalbumin or carbohydrate drink did not change WHOOP-measured sleep, resting heart rate or HRV compared with placebo (all p > 0.05). The trial tested whether these drinks do harm, over only three nights per condition, so it does not show any benefit.
Journal of the International Society of Sports Nutrition
Of 1067 records screened, 31 studies on salivary cortisol and sleep quality met the criteria, and the review found limited consistency in using cortisol as a sleep marker. Results depended heavily on methods and context, and the authors suggest cortisol reflects broader stress physiology more than sleep itself.