What happens to sleep and dreams as we age

Sleep architecture changes more across a lifetime than almost any other physiological system, and the changes are large enough that a recording from a newborn and one from an eighty-year-old barely look like the same phenomenon.

Infancy

Newborns sleep 14 to 17 hours in short bouts distributed around the clock, because the circadian system is not yet entrained — the suprachiasmatic nucleus is developing, and consolidation into a day-night pattern typically emerges around three to four months.

The striking feature is REM. Newborns spend roughly 50 per cent of sleep in REM, or in its immature form, active sleep. Premature infants spend more still — estimates approach 80 per cent at 30 weeks gestation.

The standard interpretation is that REM serves a developmental function, providing endogenous stimulation to a nervous system that is wiring itself and has limited external input. Howard Roffwarg's ontogenetic hypothesis, proposed in 1966, remains the leading account.

Infants also enter sleep directly through REM, which adults do not — a feature that persists until around three months and which, in adults, is a diagnostic sign of narcolepsy.

Childhood

Total sleep declines steadily. REM proportion falls to adult levels — around 20 to 25 per cent — by roughly age five.

Slow-wave sleep is at its lifetime maximum in childhood, which is why children are famously difficult to wake and why non-REM parasomnias — sleepwalking, sleep terrors, confusional arousals — peak in this period and are largely outgrown.

Dreaming, as David Foulkes's laboratory work established, develops rather than arriving fully formed. Recall from REM awakenings is low in three- to five-year-olds, and the reports are brief and static. Narrative dreaming with the self as protagonist emerges around seven to nine, tracking visuospatial cognitive development rather than verbal ability.

Adulthood

The most stable period, and even here there is drift.

Slow-wave sleep begins declining from the twenties — measurably, and earlier than most people assume. By middle age it is substantially reduced from adolescent levels. REM is better preserved through most of adulthood.

Sleep becomes more fragmented, with more brief arousals, and sleep efficiency declines gradually.

Sex differences appear. Women report worse sleep quality than men on subjective measures while showing better sleep on objective polysomnography — a well-replicated and unexplained discrepancy. Sleep changes markedly around the menstrual cycle, during pregnancy, and through the menopausal transition, where vasomotor symptoms are a major cause of disrupted sleep that is frequently treated as primary insomnia.

Older age

Several changes converge, and distinguishing normal ageing from disorder is the central clinical problem.

Slow-wave sleep declines substantially, in some older adults to near-absence. This appears to be genuine ageing rather than pathology, though it correlates with cognitive and metabolic measures.

Sleep fragments. More arousals, more time awake after sleep onset, lower sleep efficiency.

The circadian phase advances. Sleepiness arrives earlier and waking comes earlier. This explains a great deal of what is reported as insomnia in older adults: someone sleepy at nine, dozing in a chair, going to bed at ten and fully awake at half past four has had a reasonable amount of sleep distributed inconveniently.

Circadian amplitude flattens — the difference between day and night in temperature, melatonin and activity narrows, partly because of reduced light exposure and partly because of changes in the suprachiasmatic nucleus itself.

Dream recall declines, and reported dream content shifts toward less aggression and less emotional intensity in some studies.

The disorders that arrive with age

Distinguishing these from normal change is what matters clinically, because they are treatable and normal ageing is not.

Sleep apnoea prevalence rises steeply, and presentations in older adults are frequently atypical — fatigue and nocturia rather than dramatic snoring.

Restless legs syndrome becomes more common, and iron status is frequently the treatable cause.

REM sleep behaviour disorder typically begins after fifty and is strongly associated with later Parkinson's disease and related conditions. This one is important to identify.

Nocturia from prostatic or bladder causes.

Medication effects, from an increasingly long list of drugs.

Circadian disorders, particularly advanced sleep phase.

The claim worth correcting

Older people are frequently told they need less sleep. The evidence does not support it well.

What changes is the ability to obtain consolidated sleep, not the requirement. Studies giving older adults extended time in bed find they sleep more than their habitual amount, and daytime sleepiness in older adults predicts adverse outcomes rather than being benign.

The distinction matters because "you need less sleep at your age" closes down a conversation that should have identified an untreated sleep disorder — and in this age group, there very often is one.