Most problems arrive attached to an incident. Something gets missed, somebody notices, there is a bad week with a date on it, and the date becomes the reason a person finally says something out loud.
This one has no date.
The work gets delivered. Appraisals are good. Colleagues use the word calm. Someone is being paid, promoted and thanked for running a mechanism that is quietly expensive, and because the mechanism keeps producing results, nothing in the environment ever asks it to stop.
470M
people worldwide living with an anxiety disorder, WHO, 2023
27.6%
of people who need anxiety treatment receive any
9 to 23 yrs
median delay from onset to first treatment contact
30.5%
of anxiety cases a GP detects without a screening tool
It is not a diagnosis, and pretending otherwise helps nobody
There is no manual entry for high-functioning anxiety. No criteria list, no validated instrument, no clinician who can write it on a form. It is a popular description, and some people treat that as grounds for dismissing it.
The dismissal is a mistake, and it is worth being exact about why. The phrase caught on because it names something the clinical vocabulary handles badly. Diagnostic systems, and much more importantly the triage systems built on top of them, lean hard on visible impairment. Is it interfering with work, with study, with relationships. When the honest answer to all three is no, a real and expensive problem has nowhere to be registered.
So both halves of the sentence hold. The term is useful. The term is not a diagnosis. Anyone selling it as a condition is selling something, and anyone waving it away as a wellness invention has not asked what happens to a person whose symptoms are real and whose record is spotless.
Every other anxiety pattern eventually produces an incident
Panic produces an attack in a meeting. Avoidance produces a thing that did not get done. Social anxiety produces an invitation declined often enough that someone comments. These patterns force a conversation, eventually, because they leave evidence in the world that other people can see.
This one leaves a performance record.
The research on workplace anxiety stopped treating it as uniformly harmful some time ago. Bonnie Cheng and Julie McCarthy's 2018 theory of workplace anxiety in the Journal of Applied Psychology sets out the conditions under which anxiety raises performance rather than wrecking it, through motivation and narrowed attention, when a person has the regulatory resources to channel it. Their framing has a dark side and a bright side, and the bright side is the part relevant here. It is not a consolation. It is the mechanism by which the pattern gets funded.
Organisations are built to find people who deliver under pressure and give them more of it. Every good review, every promotion, every task that defaults to the safe pair of hands is a direct payment into the system. The reward is real, not a trick, and that is precisely what makes the pattern expensive to question. A mechanism that is being paid has no reason to stop, and no moment at which anybody is prompted to look at it.
Which explains something otherwise baffling about how long these things run.
- Generalised anxiety: 9years
- Panic disorder: 10years
- Agoraphobia: 12years
- PTSD: 12years
- Social phobia: 16years
- Specific phobia: 20years
- Separation anxiety: 23years
Those are the medians Philip Wang and colleagues found in the National Comorbidity Survey Replication in 2005, counting from first onset to the first time a person made any treatment contact at all. The lowest bar in that chart is nine years. The sample is American and the figures are two decades old, and no later survey has moved them anywhere near zero.
Nine to twenty-three years is not an access problem, or not only one. It is what happens when a condition has to generate its own reason to be mentioned.
Three layers, and only one of them is visible
The arrangement has a shape, and the shape is more useful than the label.
| Layer | What anyone can see | What it is actually doing |
|---|---|---|
| The engine | Early starts, met deadlines, an empty list by Friday | Starting far before it was necessary, because the thought of getting it wrong will not settle. The empty list produces relief, not satisfaction, and the relief gets shorter |
| The mask | Calm, capable, the reliable one | An automatic I am fine that arrives before any decision to say it, a hollow landing when praised for reliability, and a quiet worry about being found out |
| The bill | Nothing at all | Jaw and shoulders that will not drop, a stomach with opinions, nights spent re-running a day that is already over, guilt when resting, the first free morning of a holiday spent ill |
The middle row is the load-bearing one. What makes this expensive is not the anxiety. It is the anxiety that nobody can see, including the person's own doctor.
Consider how short that appointment is likely to be. Greg Irving and colleagues reviewed consultation length across 67 countries in 2017 and found averages ranging from 48 seconds in Bangladesh to 22.5 minutes in Sweden, with 18 countries covering around half the world's population at five minutes or less. Then consider what gets detected in that window. Elena Olariu's 2015 meta-analysis of 24 studies and 34,902 primary care patients put pooled sensitivity for anxiety disorders at 44.5 per cent, falling to 30.5 per cent when the physician worked without a diagnostic instrument.
Seven in ten missed, under good conditions, in patients who turned up. A patient who arrives presenting as competent, describes the problem in the register of mild self-criticism, and has no impairment to report is not a hard case to miss. They are the easiest one.
The warning light everybody watches is the wrong one
The standard way to express the cost of anxiety at a national scale is lost productivity. The World Health Organization puts depression and anxiety together at a trillion US dollars a year in it, which is the figure that gets quoted whenever a government is being persuaded to fund something.
This pattern contributes nothing to that column.
That is not a quibble about accounting. It is why the pattern is invisible to every system designed to catch it. Output is the last thing to go, so watching output tells a person nothing until the information has stopped being useful. The earlier signals are physical, and they cluster around rest rather than around work.
Two of them have actually been counted. Awake bruxism, the daytime jaw clenching and grinding that has nothing to do with sleep, came out at a pooled prevalence of 15.44 per cent across 17 studies in a 2023 meta-analysis in Clinical Oral Investigations, and it is the form of bruxism most consistently linked to psychosocial load rather than to anything dental. And the holiday that starts with illness has a small literature of its own: Ad Vingerhoets and colleagues described leisure sickness in 2002 in a Dutch sample of 1,128 men and 765 women, with roughly three per cent reporting that they reliably become unwell at weekends or at the start of a break. That was a pilot study and the figure should be held loosely, but the phenomenon it named is one of the most frequently described experiences in this whole pattern, and it is the opposite of a coincidence: a system that has been held at pressure lets go the moment the pressure stops, and the letting go is what registers as flu.
Guilt at rest belongs on the same list and is the one people are least willing to call a symptom, because from inside it feels like laziness being correctly identified. It is not. When stopping produces unease instead of recovery, rest is not doing what rest does, and the restoration a person believes they are getting at weekends is not arriving.
Caring about the work, and being frightened of it
The most common objection to all of this is that it describes a conscientious person, and conscientiousness is not a problem. Fair, and the distinction is sharp enough to be worth drawing properly.
Both columns produce the same calendar. That is the entire reason this goes unnoticed for a decade, and it is also why hours are such a poor way to look for it. The entry on workaholism and engagement makes the same point from the other direction: two people working the same Saturday night differ on compulsion, not on quantity, and compulsion is the variable that predicts the damage.
Neighbours, not twins
Perfectionism, people-pleasing and shame about failure get collapsed into this pattern constantly, including by people who have all four. They overlap heavily. They are not the same thing and they do not respond to the same handling, so the collapse is not harmless.
| Pattern | What the fear attaches to | The tell |
|---|---|---|
| Perfectionism | The standard, and what falling short would say about the work | Revision that will not stop. The piece is never finishable |
| People-pleasing | Another person's reaction, and the cost of disappointing them | The pressure disappears on a task nobody will ever see |
| Shame about failure | Being seen to have failed, after the event | It runs loudest afterwards, in the replay |
| Anxiety as the engine | The thing going wrong, continuously, before it has | It runs loudest beforehand, and finishing brings relief rather than pride |
Anyone taking this to a clinician should take the specifics from the middle and right-hand columns rather than the label. Four patterns that look identical in a calendar look quite different in a description of one week.
The sentence that gets past the first objection
People who function well get talked out of asking for help. Sometimes by a clinician with five minutes, far more often by themselves on the way there, and the argument that does the talking is always the same one. Nothing has gone wrong. Other people need it more.
Whether this is a common experience is not really in doubt. The European Agency for Safety and Health at Work surveyed more than 27,000 workers across the EU in 2022 and found 46 per cent reporting exposure to severe time pressure or work overload, the risk factor with the strongest link to poor work-related mental health in their data, and 27 per cent reporting stress, anxiety or depression. Those are not people who have all collapsed. Most of them are still turning up.
How it ends, which is not how people expect
The received picture of running out is a gradual decline, a slow fade in capability that gives everyone around a person time to notice.
The clinical literature says something stranger and more specific. Sweden has had a diagnosis for this since 2005, exhaustion disorder, which requires at least six months of stress exposure without adequate recovery. Elin Lindsäter and colleagues reviewed the research on it in BJPsych Open in 2022, and the qualitative studies they summarise describe a long starting phase in which the body repeatedly calls for attention through physical symptoms and disturbed sleep, while the person goes on working and managing ordinary life despite feeling ill. Then the breakdown.
So the deterioration is gradual. The deterioration is also, for its entire duration, happening in exactly the places nobody is looking. What is not gradual is the visible part, which holds and holds, because holding is the skill being practised, and then one morning the ordinary first task of the day turns out to be unliftable.
The other finding in that review is the one worth sitting with. At follow-up seven to ten years after diagnosis, 87 per cent of the sample were no longer on sick leave, and around a third were still clinically exhausted, with almost half still reporting fatigue and 73 per cent reporting reduced tolerance for stress. Returning to work and recovering are not the same event. The bill, once it arrives in full, is settled over a much longer period than anyone budgets for.
None of which is an argument for alarm. It is an argument for the unglamorous thing, which is that this is worth raising with a doctor or a therapist at the point when there is still nothing to report. That is the exact circumstance in which nobody goes, and the exact circumstance in which it is cheapest to act.
A problem that has never produced a bad day is not a problem that lacks a cost. It is a problem that has been very efficiently kept off the record.
Sources
- World Health Organization. Anxiety disorders: fact sheet.
- Wang, P. S., Berglund, P., Olfson, M., Pincus, H. A., Wells, K. B., Kessler, R. C. (2005). Failure and delay in initial treatment contact after first onset of mental disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 62(6).
- Olariu, E., Forero, C. G., Castro-Rodriguez, J. I., et al. (2015). Detection of anxiety disorders in primary care: a meta-analysis of assisted and unassisted diagnoses. Depression and Anxiety, 32(7).
- Irving, G., Neves, A. L., Dambha-Miller, H., et al. (2017). International variations in primary care physician consultation time: a systematic review of 67 countries. BMJ Open, 7(10).
- Cheng, B. H., McCarthy, J. M. (2018). Understanding the dark and bright sides of anxiety: a theory of workplace anxiety. Journal of Applied Psychology, 103(5).
- Archer, A. B., Da-Cas, C. D., Valesan, L. F., et al. (2023). Prevalence of awake bruxism in the adult population: a systematic review and meta-analysis. Clinical Oral Investigations, 27(10).
- Vingerhoets, A. J. J. M., Van Huijgevoort, M., Van Heck, G. L. (2002). Leisure sickness: a pilot study on its prevalence, phenomenology and background. Psychotherapy and Psychosomatics, 71(6).
- Lindsäter, E., Svärdman, F., Wallert, J., et al. (2022). Exhaustion disorder: scoping review of research on a recently introduced stress-related diagnosis. BJPsych Open, 8(5), e159.
- European Agency for Safety and Health at Work (2023). Mental health at work after the COVID-19 pandemic: what European figures reveal.
- World Health Organization (2016). Investing in treatment for depression and anxiety leads to fourfold return.