These are measurable effects, and they matter, but they're easy to overstate. Here's what the better studies show, and how big the effects are.
Sleep
| Study | What it found |
|---|---|
| Prather 2015: 164 healthy adults wore sleep trackers for a week, then got cold virus up the nose | Those sleeping under 6 hours a night had about 4 times the odds of developing a cold compared with those sleeping over 7, though the study was small and the range wide, roughly 1 to 17 times |
| Spiegel 2002: 11 young men restricted to 4 hours in bed for 6 nights, and given a flu vaccine after the fourth | About half the antibody response at 10 days, though this later caught up |
Sleep has a real biological role here. During sleep, growth hormone and prolactin levels rise and support immune cells, and memory forms better after vaccines. There's also a daily rhythm: numbers of immune cells in the blood go up and down throughout the day, partly driven by cortisol. T cell numbers are highest at night, and sleep lowers them slightly compared with staying awake, probably by moving some into lymph nodes, though that hasn't been shown in people.
Being ill makes you sleep more because of cytokines such as IL-1 and TNF, so the relationship runs both ways.
Stress
Short and long stress act quite differently.
| Short-term (minutes to hours) | Chronic (weeks to years) | |
|---|---|---|
| Main hormones | Adrenaline, then a cortisol rise | Prolonged high or disordered cortisol |
| Immune cells | NK cells and neutrophils flood into the blood | Fewer lymphocytes and weaker T cell function |
| Effect | Seems preparatory, and can improve vaccine responses | Slower healing, more infections, more inflammation |
Evidence for chronic stress:
- Catching colds. A 1991 study gave 394 volunteers cold viruses. Infection rates rose in step with the stress they had reported over the past year.
- Wound healing. Small test wounds healed about a quarter slower in people caring for a relative with dementia than in matched controls.
- Latent viruses. Antibody levels against EBV and other herpesviruses rise during stressful periods such as exams, suggesting that the viruses are reactivating. Astronauts on missions shed these viruses more too. See Viruses that stay.
Why stress lowers immunity
Cortisol is a natural anti-inflammatory, and a useful one. It stops inflammation getting out of hand during a sudden threat. Long term, cells can become less sensitive to it, so inflammation may actually go up at the same time as protective responses fall. It's therefore wrong to think of stress as simply "turning immunity down".
Exercise
Regular moderate exercise is linked to fewer respiratory infections, less severe infections, lower background inflammation and better vaccine responses. Most of this comes from observational studies, where healthier people are also more able to exercise. But trials support at least part of it.
An "open window" of lower immunity after hard endurance exercise has long been talked about, based on falling lymphocyte counts straight after a race. It's now disputed, because those cells may be moving into tissues rather than dying.
These rules apply to healthy people. If exertion has been making someone with a post-infectious illness worse, more exercise isn't the right advice. See After the infection.
Keeping it in proportion
Studies tend to report relative differences, and absolute changes are often smaller. Bad sleep or a stressful month won't by itself make you seriously ill, and being at your best won't stop you catching something. A vaccine will typically do much more than any of these for a given infection. But over years, these things add up, and they're worth looking after.
Poor sleep and chronic stress measurably weaken defences and are linked with reactivation of latent viruses, and regular moderate exercise helps. The effects are real, but they change the odds rather than deciding the outcome.