In a cohort of more than 10,000 civil servants, low control over your own work predicted coronary heart disease with an odds ratio of 1.93, while how demanding the job was showed no significant association at all

The premise of the question is that stress management is a skill, that successful people have more of it, and that the skill is what can be copied. It is a reasonable thing to assume and it is not what the largest long-running study of work and health found.

That study found the successful people were not handling the same exposure better. They were on a different exposure.

We are writers, not clinicians. What follows is a reading of occupational health research, not medical or psychological advice.

What the Whitehall studies measured

Michael Marmot’s team at University College London has followed British civil servants since the 1960s. The first Whitehall study produced a result that has shaped the field: men in the lowest employment grade, the messengers and doorkeepers, had roughly three times the mortality rate of men in the highest grade.

Everyone in the cohort had a job, an income and access to the same health service. The gradient was not about destitution.

The obvious explanation was behaviour. Smoking, blood pressure, cholesterol: the classic risk factors, distributed unevenly by grade. Those factors accounted for about 40 per cent of the difference. After adjusting for them, men in the lowest grade still carried roughly twice the risk of coronary heart disease compared with the top.

Something else was doing most of the work, and the second study was built partly to find it.

The variable that carried the gradient

Whitehall II recruited 10,308 London civil servants aged 35 to 55, about two thirds men, across twenty departments. In 1997, Bosma, Marmot, Hemingway, Nicholson, Brunner and Stansfeld published the analysis in the BMJ after following the cohort for around five years.

The variable was job control: how much say a person had over what they did, when, and how. It was measured twice, three years apart, and assessed both by the employees themselves and independently, which matters, because self-reported working conditions are exactly the kind of measure a gloomy mood can colour.

Employees with persistently low job control across both assessments had an odds ratio of 1.93 for new coronary heart disease compared with those who consistently had high control. The association held after adjusting for employment grade, for negative affectivity, and for the classic coronary risk factors.

Then there is the finding that should be quoted more than it is.

Job demands showed no significant association with coronary risk in this cohort. Neither did workplace social support.

Not how much work. Not how hard. How little say.

And when job control was included in the model, the odds ratio for men in the lowest grade fell from 1.5 to 0.95. The gradient did not merely shrink. On this measure it disappeared into the control variable.

What that does to the premise

If control over your work is doing this much of the explanatory work, then the people at the top of an organisation are not primarily demonstrating superior coping. They are occupying the position with the most latitude, and latitude is the thing that tracked the outcome.

The advice genre has this the wrong way round. It observes that senior people appear less damaged by pressure, infers a technique, and recommends the technique to people whose actual difference from the senior person is not psychological.

A junior employee who adopts every stress practice a chief executive endorses still cannot decide when the work happens, which parts of it to drop, or whether the deadline is real. Those are the variables that moved the health outcomes here.

This does not mean individual practices are worthless, and it would be a poor reading to conclude that. It means the largest single lever identified in this literature is one that sits in the design of a job rather than in the habits of the person doing it.

The authors’ own recommendation reflects that. They did not suggest teaching civil servants to manage pressure better. They suggested giving employees more variety in their tasks and a stronger say in decisions about their work, which is a statement about organisations rather than individuals.

What this does not show

A fair amount, and it is worth being precise.

Whitehall II is an observational cohort, not an experiment. Nobody was assigned to high or low control. The association is strong, prospective and survives a lot of adjustment, which is about as much as observational work can offer, but it is not the same as a demonstration that changing job control changes heart disease.

The cohort is also specific: British civil servants, recruited in the mid-1980s, in a single organisational hierarchy with unusually clear grades. That clarity is what makes the study so informative and also what limits how far it travels. A modern workplace with flatter structures and less stable employment is not the same setting.

There is a boundary worth naming too. This research measured coronary heart disease, not anxiety. Stress and anxiety are related but they are not interchangeable, and nothing here describes what helps a person who is currently anxious. It describes what predicted who became ill.

The separate literature on reappraising your own arousal, which is where most of the practical advice in this area now comes from, is a different body of work with a different design and its own strengths. It is not what this study tested.

What would strengthen the case is straightforward to describe and hard to run: change job control deliberately in some workplaces and not others, then follow health outcomes for years. Trials of that shape exist in occupational health but they are small and short next to a cohort of ten thousand people followed for decades, which is why the observational evidence still carries most of the weight here.

If stress or anxiety has stopped being something you get through and started interfering with sleep, work or how you feel most days, a GP is better placed to help than any article about what executives do in the morning.

What stays with us is the shape of the numbers. When researchers put demands, support and control into the same model, the one that moved was the one that no amount of personal technique can supply, and the gradient that had survived every conventional explanation went with it.