Why The Immune System Needs Sleep

AMedicine spent much of the last century treating the drowsiness of the sick as an inconvenience, a side effect of fever that patients endured and physicians largely ignored. The assumption behind that neglect was reasonable enough: an organism fighting an infection is depleted, depletion produces fatigue, and so the extra hours of sleep observed in feverish patients seemed to need no further explanation. What unsettled the picture was a modest experiment reported in 1972 by a team at the Debrecen Physiological Laboratory, in which dogs given an extract drawn from sleep-deprived animals not only fell asleep but afterwards mounted a stronger response to an unrelated bacterium. The extract later proved to contain a fragment of bacterial cell wall. Sleep, on this reading, is summoned by infection rather than left behind by it. The question has since moved from whether the two systems speak to each other to how, and the answers now on offer divide specialists more sharply than the public summaries suggest.

BThe most developed account belongs to Ingrid Halloran of the Ravensworth Institute of Chronobiology, who maintains that the first half of the night is when the immune system does its filing. Counts of T-cells in circulating blood, she points out, fall by roughly a third between midnight and three in the morning, a decline that once looked like suppression but is now read as relocation: the cells leave the bloodstream for the lymph nodes, where fragments of any pathogen met during the day are held for inspection. Halloran calls the process consolidation, borrowing the term deliberately from research on memory, since in both cases a system that has gathered material while awake appears to sort and store it while asleep. The parallel is meant literally. Hormonal conditions peculiar to deep sleep, chiefly a trough in cortisol and a surge in growth hormone, make the meeting between cell and fragment far likelier to leave a durable record.

CEvidence for the account comes chiefly from vaccination, which offers a rare opportunity to introduce a known antigen at a known hour and then measure what the body has retained. Between 2011 and 2016 Halloran's group followed 214 volunteers who received a hepatitis vaccine either after an ordinary night at home or after a night cut to four hours under observation. Twelve months later the shortened-sleep group carried on average 42 per cent fewer specific antibodies, and 19 of them had slipped beneath the level conventionally regarded as protective. The figure was arresting. It was also, as Halloran conceded in a 2019 review, obtained only among adults under thirty-five, the age band in which the nightly cortisol trough is deepest; a smaller sample of participants over sixty showed no dependable difference whatever. Whether the mechanism genuinely fades with age, or merely becomes harder to detect in bodies whose hormonal rhythms have flattened, remains on her own account unsettled.

DNot everyone accepts that sleep itself is doing the work. Amadou Kessel, an immunologist at the Delft Centre for Human Physiology, contends that what Halloran attributes to sleep should be credited instead to the clock that governs it, since the hormonal low point on which she relies arrives at a fixed position in the circadian cycle whether or not the sleeper is unconscious. To prise the two apart he kept 19 volunteers awake through a night in dim light, holding their circadian phase intact, and found that their antibody response after vaccination fell short of that of sleeping controls by only 14 per cent, a gap he dismisses as trivial beside the shortfall Halloran describes. Critics reply that one night in a laboratory is a weak test of a process that may accumulate across weeks, and Kessel concedes the point, noting that only a pilot group of eight was tracked for fourteen nights.

EWork on animals has complicated both positions rather than settling them. At the Kavanagh Laboratory in Uppsala, fruit flies infected with a common bacterium were sorted by how much additional sleep they took in the twelve hours after exposure, and those that slept most survived at nearly twice the rate of the rest, but only where the infecting dose was small. Flies given a heavy dose slept just as much and died just as quickly, which suggests that the extra sleep buys time for a defence already capable of winning rather than supplying one that would otherwise be missing. Sceptics of the fly model observe that insects possess no adaptive immunity of the kind on which the human argument turns. The observation is fair, and it cuts in both directions: if sleeping longer improves survival in a creature with no antibodies at all, consolidation of immune memory cannot be the whole story.

FWhat all this means for advice to patients is less obvious than the popular version suggests. Halloran is careful in public: she claims a benefit for the night that immediately follows exposure to an antigen, not for sleep in general, and she has objected to campaigns that present eight hours in bed as a form of protection in itself. Kessel would rather see vaccination appointments shifted towards the morning than see patients instructed to sleep more. The two recommendations are less opposed than they look, since each implies that when the immune system is challenged matters more than how long the patient rests afterwards. A growing number of researchers suspect that both mechanisms operate together, in proportions that shift with age. What seems least likely to survive is the older and more comfortable notion that the sleep of the sick is simply the body running down.