marcel_kinsbourne's picture
Neurologist and Cognitive Neuroscientist, The New School; Co-author, Children's Learning and Attention Problems
Psychologist, The New School; Coauthor, Children’s Learning and Attention Problems

Shortening Sleep Will Prolong Conscious Life

Our life span is extending, but the extended life is a dwindling asset. Who would not prefer to live longer while at their peak? The time we spend asleep contributes little to our lifetime of experience, thought and action. Dreaming doesn't seem to add much. Some brain lesions and monoamine oxidase (MAO) inhibitor medications even completely abolish dreams without making any apparent difference. Could we reduce the duration of sleep (both REM and nonREM) while maintaining its benefits for the brain, whatever they might be?  I propose that we do need to sleep, but not as long as we do. The duration of sleep may be an outdated adaptation to prehistoric ecological constraints that no longer exist.

Virtually all vertebrates sleep (and invertebrates at least have quiet time). However, the duration of sleep varies wildly across species, from less than 1 hour to 18+ hours a day. For instance, rodents sleep between 8 and 17 hours, primates between 7 and 18 hours. Elephants and giraffes sleep 3-5 hours, squirrels 16-17 and bats 20 hours. The newborn of most species sleep more of each day than the adults, except that newborn whales and dolphins don't sleep at all. Within a species, the inter-individual variation of adaptively valuable traits is thought to be quite limited. Yet some people, in some families, habitually sleep only 2-4 hours a night, and function well for longer each day. Perhaps constraining the duration of sleep is not an adaptive priority in humans.

Three categorically distinct roles for sleep are: (1) maintaining the neuronal circuitry, (2) fostering learning, (3) keeping the organism out of trouble.

(1) Given its ubiquity among vertebrates and other phyla, any neurometabolic benefit of sleep must be very general and basic. If the needs of the brain determine the duration of sleep, its duration should vary systematically with some fundamental neurological variable, such as absolute or relative size of the brain, its energy utilization, the sophistication of behavioral control, or the need to replenish some key neurotransmitter. No such co-variation appears to exist.

(2) The presumed role of sleep in learning is based on continuing rehearsal. Rather than being an adaptation, the learning benefit may be a fortuitous result of the brain's continuing activity during sleep, while it is receiving no fresh information. Since the neuronal show must go on, recently acquired patterns of firing gain priority and are "rehearsed". Whether the memories are useful or useless, they are automatically rehearsed. In any case, the suggested benefit of sleep for human learning cannot be generalized to species that make a living without learning anything much, and yet require sleep.

(3) The substantial differences between people and the enormous difference between species in how long they typically sleep suggest that sleep also serves a species-specific ecological function. This is sleep's other role; sleep conserves energy and keeps animals out of trouble. It takes the members of each species a minimum time per day to make a living, that is, secure their personal survival and take advantage of any reproductive opportunity. This challenge is met anew every day. On this view, how much of the day is needed to meet adaptive goals determines the duration of the default option of sleep.

Continued activity when the day's housekeeping is done would prolong the animal's exposure to the hazards that lurk in the environment, without contributing further to basic survival and reproductive needs. Many species cannot do anything useful in the dark (and some not in the light). They gain nothing from expending more than basal metabolic energy at that time. The genetic imperative to sleep during a predetermined time of day and for a predetermined duration (or even hibernate), takes care of all that. Thus extended sleep time would be a function of the interaction between the individual and its ecology.

Predators need time for hunting; how much depends on attributes of the predator and the prey, such as speed, strength and population density. Herbivore prey needs a minimum time to graze, depending on the animal's bulk and the accessibility of food. How the remains of the day are spent would depend on how readily a secure haven can be found.

Nature is notoriously conservative, and it conserves the genetically driven imperative to sleep. The imperative to sleep is subjectively experienced as antecedent sleepiness, and the fatigue and dysphoric feeling after too little sleep. My thesis is that these feelings do not arise unavoidably from the economy of the brain, but are genetically imposed adaptations. Should a species' ecology undergo radical change, and making a living become sharply easier or more difficult, natural selection will in time reshape sleep duration accordingly. 

However, human culture evolves too quickly. Since artificial lighting was introduced, the dark no longer constrains what people can do. Since human activities are oriented as much to future as to immediate goals, all hours of the day have become potentially useful. Further, we have more effective means to secure ourselves than curling up in a quiet place and sleeping. So if a sizeable portion of the adaptation to sleep has the role of a security saving placeholder, then it would be safe to relax that portion of the sleep constraint.

The dictatorial "sleep genes", when identified, need to be modified to require a shorter sleep duration, and the circadian clock genes need to be reset. Will the state of genetic engineering become sufficiently advanced to make this prospect, though less than a sure thing, more than a pipe dream? The good news comes with the fruit fly's sleep, which is uncannily like ours; a mutation in a gene called Shaker reduces the fly's natural sleep duration by two-thirds, from about 12 to about 4 hours within 24, without detriment to the fly's well-being. The bad news is that these mutated flies don't live long. Nonetheless, I am optimistic.