Sleep quality

The Pittsburgh Sleep Quality Index (PSQI), published in 1989, is the most widely used questionnaire. It asks about the previous month in 19 items that produce seven component scores: subjective sleep quality, time taken to fall asleep, sleep duration, sleep efficiency, sleep disturbances, use of sleep medication and daytime dysfunction. The component scores add up to a single global score. In the original validation, a global score above 5 separated good from poor sleepers with a sensitivity of 89.6% and specificity of 86.5% [2].

Objective measures include polysomnography (an overnight recording of brain waves, eye movements, muscle activity and breathing) and actigraphy (a wrist-worn movement monitor that estimates sleep and wake). In 2017 a National Sleep Foundation panel reviewed 277 studies and agreed that four sleep continuity measures are appropriate indicators of good sleep quality across the lifespan: time to fall asleep, the number of awakenings longer than 5 minutes, time awake after first falling asleep, and sleep efficiency. The panel reached little or no agreement on whether the proportions of sleep stages or daytime naps indicate sleep quality [1].

The Vietnam Era Twin Study of Aging measured sleep quality with the PSQI in 1,218 men aged 51 to 60 (339 identical and 257 non-identical twin pairs). The mean global score was 5.6. Additive genetic effects explained 34% of the variation in the global score and 31% of the likelihood of being a poor sleeper (a score above 5). The remaining 66% of variation was attributed to environmental factors specific to each individual, and the shared family environment contributed nothing measurable. Heritability of six of the seven component scores ranged from 15 to 31%, and use of sleep medication showed no genetic influence [3].

The study included only men of one generation, so the figures may differ in women and at other ages. The large individual-specific share includes habits, health conditions, work schedules and bedroom environment, which are the factors examined in the sections below.

Caffeine blocks adenosine, a chemical that builds up in the brain during waking and promotes sleep. A 2023 pooled analysis of 24 studies found that caffeine reduced total sleep time by 45 minutes and sleep efficiency by 7%, lengthened the time to fall asleep by 9 minutes, added 12 minutes of waking after sleep began and reduced deep sleep by about 11 minutes. From the timing data, the authors estimated that a coffee containing 107 mg of caffeine should be consumed at least 8.8 hours before bedtime and a 217.5 mg pre-workout supplement at least 13.2 hours before [5]. In a home study, 400 mg of caffeine taken at bedtime, 3 hours or 6 hours before bed reduced objectively measured sleep by 1.1 to 1.2 hours at each time point, while the participants' own sleep diaries showed a smaller, non-significant reduction of 41 minutes for the 6-hour dose [9].

Alcohol shortens the time taken to fall asleep and consolidates the first half of the night, then increases sleep disruption in the second half. It delays the first period of REM (rapid eye movement) sleep at all doses and reduces total REM sleep at moderate and high doses [10]. A 2024 pooled analysis of 27 studies found that REM sleep was reduced after about two standard drinks and progressively more with higher doses, while faster sleep onset was seen only after about five drinks [11].

Light is the main signal that sets the internal body clock. After a summer week of camping with only natural light, the timing of the internal clock moved about 2 hours earlier, and the evening rise in the sleep hormone melatonin began near sunset [6]. In a randomised crossover study, reading on a light-emitting tablet for about 4 hours before bed on five consecutive evenings, compared with a printed book, suppressed melatonin, delayed the clock by more than 1.5 hours, lengthened the time to fall asleep, reduced REM sleep and lowered alertness the next morning, although total sleep time did not differ [12]. Regular sleep and wake times keep the clock aligned; in 60,977 UK adults, more regular sleepers had 20 to 48% lower mortality than the least regular [13]. The clock itself is described in sleep physiology, and melatonin supplements in melatonin.

Sleep apnea fragments sleep through repeated breathing pauses, often without full awakening. An estimated 936 million adults aged 30 to 69 have it [14], and untreated severe apnea was associated with about 3 times the all-cause mortality over 18 years [15]. Long daytime naps can reduce sleep pressure at night; naps of an hour or more have been associated with higher mortality, while naps under an hour were not [16].

Full deep dive: sleep habits and sleep quality factors.

Core body temperature falls in the evening, and this fall is linked to the start of sleep. In community-dwelling older adults whose sleep was tracked at home with wearable monitors and bedroom sensors, sleep was most efficient and restful when the night-time bedroom temperature was between 20 and 25°C, and efficiency fell by 5 to 10% as the temperature rose from 25 to 30°C. The associations were nonlinear, and the best temperature varied considerably between individuals [7].

A systematic review of 17 studies of warm showers or baths before bed, 13 of which were pooled, found that water at 40 to 42.5°C was associated with better self-rated sleep quality and sleep efficiency, and that bathing 1 to 2 hours before bedtime for as little as 10 minutes shortened the time taken to fall asleep. The proposed mechanism is increased blood flow to the hands and feet, which speeds heat loss and the decline in core temperature [17].

Light in the bedroom at night has also been linked to poorer sleep. In a cross-sectional study of 857 Japanese adults with a mean age of 72, those in the brightest quarter of measured bedroom light had 61% higher odds of self-reported insomnia than those in the darkest quarter, and wrist-monitor measures showed lower sleep efficiency and longer time awake with brighter bedrooms [18]. As a single-time survey, it cannot show which came first.

A 2015 pooled analysis of 66 studies found that regular exercise had small beneficial effects on total sleep time and sleep efficiency, small-to-medium effects on the time taken to fall asleep, and moderate effects on self-rated sleep quality. A single session of exercise also had small beneficial effects on the following night [19]. In 22 randomised trials of regular exercise lasting at least two months, exercise lowered PSQI scores by 2.19 points compared with control groups, with similar effects for conventional exercise and mind-body exercise such as yoga; improvements were seen mainly in self-reported rather than objectively measured sleep [20].

Evening exercise has often been discouraged before bed. A review of 23 studies comparing a single evening exercise session with no exercise found no overall harm to sleep and a small increase in deep sleep, although vigorous exercise ending an hour or less before bedtime may lengthen the time to fall asleep and reduce total sleep and sleep efficiency [21].

Psychological stress is one of the strongest everyday influences on sleep quality. In 150 university students surveyed before, during and after an exam period, sleep quality fell during exams, and perceived stress was a stronger predictor of poor sleep than the amount of alcohol or caffeine consumed, which had no significant influence [22]. The relationship runs in both directions: in a Swedish population study of 1,812 adults, anxiety at baseline was associated with 4.27 times the odds of new insomnia a year later, and insomnia with 2.30 times the odds of new high anxiety and 3.51 times the odds of new high depression [23]. The wider topic is covered in stress and anxiety.

In a randomised trial in 49 older adults with moderate sleep disturbance, six weeks of mindfulness training reduced PSQI scores from 10.2 to 7.4, compared with 10.2 to 9.1 in a sleep hygiene education group [24]. When poor sleep becomes chronic insomnia, the American College of Physicians recommends cognitive behavioral therapy for insomnia (CBT-I) as the initial treatment [4]. In 20 randomised trials with 1,162 participants, CBT-I shortened the time to fall asleep by about 19 minutes, reduced time awake during the night by about 26 minutes and raised sleep efficiency by about 10 percentage points, with benefits that appeared to persist after treatment ended [8].

Much of the evidence on everyday influences comes from short laboratory or crossover studies in small groups of healthy young adults, which show effects on a single night rather than over years. Self-reported sleep quality and objectively measured sleep often disagree: caffeine reduced measured sleep by more than an hour in one study while diaries recorded a smaller loss [9], and exercise trials improved questionnaire scores more than measured sleep [20]. Observational findings, such as those on bedroom light, cannot separate cause from effect, and the twin estimates of heritability come from middle-aged men only [3].