For roughly twenty-five years of my career I worked directly and closely with people who were ill. Not in the hospital sense — but people who arrived at a training session or consultation coughing, with a streaming nose, with the early-onset shiver of something coming on, or with the kind of pale hollowness around the eyes that tells an experienced clinician that the body is fighting something. And sometimes they would sneeze directly at me. Sometimes — particularly when holding the pads for a boxer throwing combinations with full effort — saliva or nasal mucous would land on my face, on my lips, in my eyes. You are wearing thick leather pads over your hands. You cannot wipe it off. You continue.
I did not get sick.
Not from those exposures, not routinely, not in any way that suggested direct transmission from the person across from me to my respiratory tract. This was not luck, as I understand it. It was consistent enough across enough years and enough exposures to constitute a pattern rather than a coincidence. And it raised a question that I have been thinking about ever since, and that I think is one of the most important and most misunderstood questions in health: what actually determines whether a pathogen causes illness in a given person at a given time?
The simple answer — the one we are all given — is exposure. You are exposed to a virus or bacterium, you catch it, you become ill. That answer is right as far as it goes: no exposure, no infection, and more exposure means more risk. But exposure isn't the whole story. Given the same exposure, some people get ill and some don't, and the reasons are partly measurable.
Updated October 2026. Three of the numbers in this post were attributed to the wrong study or overstated. They now cite the right papers with their actual figures, the unsourced claims are graded or removed, and Part Two is linked at the end. What changed and why →
The Doctor Who Doesn't Get Sick
Consider the GP who has been in practice for thirty years. Every working day they sit within three feet of people who are acutely unwell with respiratory infections, gastrointestinal illness, and every variety of viral and bacterial pathogen that presents to primary care. They ask patients to open their mouths and say "ah" and then peer into their throats. They shake hands. They handle objects that have been touched by acutely infectious people.
It's a common observation that some long-serving GPs seem rarely to be ill. But I can't find good data showing that doctors as a group catch fewer colds than anyone else, and where infection in healthcare workers has been measured carefully, exposure wins. Early in the COVID-19 pandemic, front-line healthcare workers in the UK and US were at least three times more likely than the general community to report a positive test, after adjusting for how often they were tested (Nguyen 2020; observational). So the anecdote proves less than it seems to. What it does point to is real, though: among people with the same exposure, some get ill and some don't.
The concept that explains this is immune tolerance — the immune system's ability to encounter a pathogen and mount an appropriately calibrated response rather than either failing to respond or mounting a disordered, excessive one. Immune tolerance is not static. The things that seem to shift it, such as sleep, stress and social connection, are the subject of the studies below. Grade: the concept is established immunology; how much everyday lifestyle shifts it is less certain.
"Two people sit in the same room, breathe the same air, are exposed to the same pathogen. One becomes ill. One does not. They were in the same physical environment. They were not in the same biological one."
The Research That Changed How We Think About Infection
Sheldon Cohen's work at Carnegie Mellon University is some of the most useful research on this question. In a series of studies, Cohen and colleagues deliberately exposed healthy volunteers to rhinovirus — the common cold virus — via nasal drops, then isolated them and observed who developed a clinical cold.
Not everyone exposed developed illness. The rate of clinical infection varied substantially between individuals. And the variables that predicted who got sick were not primarily about the pathogen or the dose of exposure. They were about the person.
People with more types of social tie (across work, family, community and friendship) were less likely to develop a cold after the same exposure. Those with the fewest types (one to three) were about four times as likely to get a cold as those with six or more (Cohen 1997, JAMA; 276 adults).
People reporting more psychological stress were more likely both to be infected and to develop a cold, in a dose-response pattern (Cohen 1991). In an earlier sleep study, people sleeping under 7 hours were about three times as likely to develop a cold as those sleeping 8 hours or more (odds ratio 2.94, 95% CI 1.18–7.30). Poor sleep efficiency was linked even more strongly, at about 5.5 times (Cohen 2009, Arch Intern Med). In the later study that measured sleep with a wrist monitor rather than a diary, under 5 hours and 5 to 6 hours both carried about four times the odds. Six to seven hours carried no clear extra risk (Prather 2015).
How sure is this? These are controlled exposures to the same virus, which is a strong design. But sleep, stress and social ties were measured, not assigned, so they are associations, and the confidence intervals are wide. Grade: consistent associations from experimental exposure studies; no trial has shown that improving sleep or social ties prevents colds. The pathogen and the dose were the same. The person was different.
The Terrain — What Your Biology Looks Like From the Inside
Louis Pasteur and Antoine Béchamp had one of the great scientific disagreements of the nineteenth century. Pasteur's germ theory — the idea that specific pathogens cause specific diseases — became the dominant model and shaped the entire edifice of modern medicine. Béchamp argued that the state of the host — what he called the milieu intérieur, the internal environment — was as important as the pathogen in determining whether disease resulted from exposure.
The two positions are not mutually exclusive, and serious immunologists have never treated them as such. Germs cause infections; that isn't in doubt, and nothing in this post argues otherwise. Exposure still matters, and hand-washing, ventilation and vaccination work by reducing it or by preparing the immune system for it. The host's state changes the odds, not the fact that the pathogen is the cause.
The terrain — your biological environment at the time of exposure — consists of several interacting systems that collectively determine your immune response:
The Family That Gets Sick Together — And the One That Doesn't
One of the observations that most reliably prompts people to question the pure transmission model is the family where one person gets ill and the others in the same house do not. They breathe the same air. They share the same surfaces. The pathogen is present for all of them.
The explanation offered is usually individual variation in immune function — which is true, but incomplete, because it stops short of asking what produces that variation. The person who gets sick and the person who doesn't are not simply biologically different in some fixed, predetermined way. They are different in their current biological state. Their sleep quality. Their stress load. Their nutritional status. Their gut health. Their cortisol pattern. Their social connection.
There is also the phenomenon you have likely observed — the family member who hears that someone is coming down with something and begins to feel symptoms before any meaningful exposure has occurred. That isn't weakness or hypochondria. Expectation changes how we notice and report body sensations. Whether it changes immune function enough to alter infection risk hasn't been shown. Grade: untested.
Robert Sapolsky's work on stress physiology — particularly his research on social hierarchy and glucocorticoid patterns in primates, and his documentation of how social threat and social status directly modulate cortisol and immune function — is a big influence on how I think about this. Most of it is in primates and in mechanism, so it's a framework rather than a finding in people.
Immune Tolerance and the 90-Year-Old Smoker
The 90-year-old who smoked for sixty years and never developed lung cancer is one of the most commonly misused anecdotes in health debates. It is used to argue that smoking is not as dangerous as claimed, or that individual genetics can protect against any exposure, or that lifestyle choices are ultimately irrelevant if you are constitutionally resilient enough.
None of these is the right lesson. Smoking is the main cause of lung cancer, and there is no level of smoking that is safe for some people. The exposure matters most, and the risk rises with how much and how long someone smokes. Some long-term smokers never develop lung cancer, and the reasons are partly chance, partly genes and partly things we can't yet measure. The idea that some people have more "biological reserve" (repair, detoxification and immune surveillance) is a reasonable one. But it isn't something anyone can measure in advance, so nobody can know they are one of the lucky ones. Grade for "reserve" as an explanation: mechanistic and untested.
My grandfather died of lung cancer at 67. He smoked. The smoking was the cause. What I take from it is the narrower point this post is about: exposure sets the risk, and the state of the person changes the odds within it. Reducing exposure and looking after the person are both worth doing. Neither replaces the other.
What This Means Practically — Building the Terrain
If the host's state changes the odds, the practical question adds to "how do I reduce exposure" rather than replacing it: what is the state of my biological environment, and what is depleting it?
The six immune-supporting nutrients — Vitamins A, C, D, E, Selenium, and Zinc — are the foundation layer. Food first, as I have always maintained. A diet of colourful vegetables, oily fish, nuts, seeds, fermented foods, and quality protein provides most of what the immune system needs without supplementation. Supplements address deficiency once it is identified — they do not substitute for the nutritional foundation.
But nutritional status alone is insufficient as a framework if gut health is compromised, if cortisol is chronically elevated, if sleep is disordered, if social connection is depleted, or if chronic low-grade infections are draining immune reserve without producing obvious symptoms.
This is where I use testing: GI-MAP for gut findings, DUTCH Plus for the daily cortisol pattern, and blood chemistry for nutrient status and inflammatory markers. They can show things a symptom history can't. But be clear about the evidence: no trial has shown that testing well people this way reduces infections. Some findings (low vitamin D, low ferritin) have clear treatments. Others (a low stool sIgA, a Blastocystis result) are leads whose meaning for infection risk hasn't been established. Grade: not trial-tested for infection susceptibility.
Exposure matters, and it isn't the whole story. In experiments where people get the same dose of the same virus, short sleep, high stress and fewer social ties are each linked to higher odds of a cold.
Parts of the terrain can be measured. Some (vitamin D, iron) have clear meaning. Others are leads, and whether testing well people prevents infections hasn't been tested.
Parts of the terrain can be changed. Sleep, movement, nutrition, stress and social connection are worth working on for many reasons. Whether changing them prevents colds is mostly untested; the links above are associations.
Sources
- Prather AA et al. Behaviorally assessed sleep and susceptibility to the common cold. Sleep 2015;38:1353–9. PMID 26118561, doi:10.5665/sleep.4968
- Cohen S et al. Sleep habits and susceptibility to the common cold. Arch Intern Med 2009;169:62–7. PMID 19139325, doi:10.1001/archinternmed.2008.505
- Cohen S, Tyrrell DA, Smith AP. Psychological stress and susceptibility to the common cold. N Engl J Med 1991;325:606–12. PMID 1713648, doi:10.1056/NEJM199108293250903
- Cohen S et al. Social ties and susceptibility to the common cold. JAMA 1997;277:1940–4. PMID 9200634
- Buijze GA et al. The effect of cold showering on health and work: a randomized controlled trial. PLOS ONE 2016;11:e0161749. PMID 27631616, doi:10.1371/journal.pone.0161749
- Nguyen LH et al. Risk of COVID-19 among front-line health-care workers and the general community: a prospective cohort study. Lancet Public Health 2020;5:e475–83. PMID 32745512, doi:10.1016/S2468-2667(20)30164-X
- Nieman DC et al. Upper respiratory tract infection is reduced in physically fit and active adults. Br J Sports Med 2011;45:987–92. PMID 21041243, doi:10.1136/bjsm.2010.077875
If you keep getting ill, start with the basics.
Frequent infections can have medical causes your GP should check first. After that, the GI-MAP, DUTCH Plus and blood chemistry can show gut findings, the cortisol pattern, nutrient status and inflammation. They are leads for a plan, not a proven way to prevent colds.
TDG Five-Test Programme → Book a free 20-minute call →