Counts describe this library, not the strength of evidence.
Cómo interpretar la evidencia
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.
In a randomized trial of 72 people with chronic insomnia in Beijing, ear-based vagus nerve stimulation (taVNS) lowered the PSQI sleep score 4.2 points more than sham (95% CI -5.9 to -2.6; P < .001). The trial was small, single-center, and the sham group also improved, so wider benefit is unproven.
Across 19 studies, adolescents with sleep disturbances had higher odds of attempting suicide (OR 3.10, 95% CI 2.43-3.95) and of suicidal ideation (OR 2.28, 95% CI 1.76-2.94) than controls. The pooled studies are associational, so the data cannot show that poor sleep causes suicidal behavior or that fixing sleep would prevent it.
In 319 adults without dementia, a more fragmented 24-hour activity rhythm at baseline was associated with more brain amyloid-β on PET 7.8 years later (β 0.15, 95% CI 0.04 to 0.26), more strongly in APOE4 carriers (β 0.38). No other sleep measure was associated, and calling fragmentation a modifiable risk factor is speculative.
In 14,664 US adults who were short sleepers, more physical exercise was associated with a lower systemic immune-inflammation index (SII), with no further benefit beyond about 2400 MET-minutes per week. This is a one-time snapshot using self-reported sleep and exercise, so it cannot show that exercise lowers inflammation.
In 36 people with active ulcerative colitis, 12 weeks of sodium butyrate lowered faecal calprotectin and hs-CRP, raised several circadian-clock genes and improved the Pittsburgh Sleep Quality Index (PSQI) score versus placebo (all P < 0.001 for those outcomes). With only 36 participants and a short trial, the results need confirmation in larger studies.
In 78 people with Parkinson disease and poor sleep, real acupuncture raised Parkinson Disease Sleep Scale scores by 19.75 points more than sham acupuncture after 4 weeks (95% CI 11.02-28.49). It was a single-center trial, and the sham group also improved by 10.47 points, so part of the gain is not specific to needling.
In 30,475 college students, more physical exercise was associated with lower stress, anxiety and depression (r = -0.03, -0.058 and -0.055, all p < 0.01), partly through sleep quality and self-rated health. The correlations are small and cross-sectional, so they cannot show that exercise improves emotions.
In 3932 US adults from NHANES, higher lipid accumulation product (LAP) and visceral adiposity index (VAI) were associated with obstructive sleep apnea (OSA), and the link was stronger under age 60. The data are a single cross-sectional snapshot, so they cannot show that high LAP or VAI raises OSA risk.
Among 247 867 UK Biobank adults followed a median 12.5 years, sleeping 5 hours a day (adjusted HR 1.16) or 3 to 4 hours (HR 1.41) was associated with more type 2 diabetes than 7 to 8 hours. Sleep and diet were reported once at baseline, so the study shows association and cannot rule out other explanations.
Among 58 young adults who wore an accelerometer for 7 days, longer sleep was linked to higher accuracy on the incongruent Stroop test, and sleeping under 6 hours was linked to lower accuracy than 7 hours or more (-0.012, CI -0.023 to -0.002, p = 0.024). The sample is small and the results are associations, with sleep measured by a wrist-type activity monitor rather than a sleep lab.
In 502,383 UK Biobank participants, insomnia (HR 1.05), daytime napping (HR 1.09) and dozing (HR 1.19) were each associated with higher stroke risk after adjusting for cardiovascular risk factors, and genetic analysis supported a causal link only for insomnia and stroke (OR 1.31, 95% CI 1.13-1.51). The sleep measures were self-reported, and most associations with dementia weakened or vanished in the genetic analyses.
Pooled across studies, 23% (95% CI 20%-26%) of people with post COVID-19 syndrome had depression and 23% (95% CI 20%-26%) had anxiety, while sleep disorder reached 45% (95% CI 37%-53%). Heterogeneity was extreme (I2 from 98.7% to 99.9%) and publication bias was possible, so these figures are a rough range, not a rate for any one group.
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.
Across 10 studies with 16,726 people, the triglyceride-glucose (TyG) index was higher in obstructive sleep apnea than in healthy controls (SMD 0.856, 95% CI 0.579 to 1.132), but its pooled diagnostic AUC was only 0.681. The authors endorse routine use, which goes beyond a modest AUC and a literature with few studies per question.
Among 10 212 US adults in NHANES, poor sleep patterns were linked to higher odds of cardiovascular disease (CVD; OR 1.76, 95% CI 1.26-2.45), while a higher oxidative balance score (OBS, more antioxidant exposure) was linked to lower odds (Q4 vs Q1: OR 0.67, 0.47-0.94). Sleep and diet were self-reported in a one-time survey, so the data show association, not cause.
A meta-analysis of 42 longitudinal studies found small, two-way links between sleep and mental health in adolescents: good sleep predicted fewer internalizing symptoms later (r = -.20) and more internalizing symptoms predicted worse sleep (r = -.21). Correlations are small and observational, so they do not show that improving sleep treats mental health problems.
International journal of clinical and health psychology : IJCHP
Among 826 older adults in the Seattle Longitudinal Study, being a short sleeper (HR 3.67, 95% CI 1.59-8.50) and having more variable sleep duration (HR 3.06, 95% CI 1.14-5.49) were each linked to incident cognitive impairment. Sleep was self-reported as median nightly hours, and the study is observational, so it cannot show that sleep change causes decline.
Among 88,975 UK Biobank participants, those with a sleep regularity index (SRI) of 41 (5th percentile) had a hazard ratio of 1.53 (95% CI 1.41-1.66) for all-cause death relative to the median SRI, while an SRI of 75 (95th percentile) gave 0.90 (0.81-1.00). Sleep was measured over only 7 days, and this is an association in older adults.