The Part of Stress That Techniques Cannot Reach

Eight weeks of a mindfulness course, finished rather than abandoned. A breathing exercise before anything difficult. Five runs a week, kept up through a winter.

And the number has not moved.

That shape is common enough to deserve a better explanation than insufficient commitment. It has nothing to do with willpower, and it is not that the techniques fail. They land in the wrong column.

Stress is three quantities, not one

Nothing in the research treats stress as a single number. Epel and colleagues, writing in 2018 about how the field should measure it at all, describe a construct that emerges from interactions rather than one that sits still long enough to be counted once, and their typology separates timescales and dimensions that ordinary use collapses. Controllability is one. Chronicity is another.

The version that is usable without a laboratory has three parts.

What is arriving. How much is happening at once, and how much of it is somebody else's decision. Call it load.

What the body and the mood are doing underneath it. Sleep, digestion, a jaw that will not let go, a fuse that has shortened, attention that slides off the page. Call it strain. What that looks like in detail belongs to a different page and is not repeated here.

And what is left over. Whether one unplanned thing can still be absorbed, whether anything in the day is genuinely off, whether a night's sleep still resets anything. Call it reserve.

The reason to keep them apart is that the same load arriving on two different reserves is two different situations, and the advice that fits one is close to useless for the other.

The effects are real, and understating them would be its own dishonesty

It would be convenient for the argument to treat the techniques as a placebo. They are not, and the numbers are better than the sceptics tend to admit.

Vonderlin and colleagues pooled the randomised trials of mindfulness-based programmes run in workplaces in 2020. Across 43 trials reporting on perceived stress, the pooled effect was a Hedges g of minus 0.66, the strongest result in the review. Burnout, mental distress, somatic complaints, well-being and job satisfaction all moved as well, with effects running from small to large.

Physical activity has the larger evidence base. Singh and colleagues gathered 97 reviews covering 1,039 trials and 128,119 adults in 2023 and reported a median effect of minus 0.43 on depression, minus 0.42 on anxiety and minus 0.60 on psychological distress, compared with usual care.

And the honest version of the mindfulness picture is narrower than the headlines. Goyal and colleagues restricted their 2014 analysis to comparisons against active controls rather than waiting lists, which is the harder test, and across 47 trials and 3,320 participants found moderate evidence for improvement in anxiety, an effect size of 0.38 at eight weeks falling to 0.22 by three to six months, and 0.30 for depression. Evidence for improvement in stress and distress specifically was graded low. That is a real effect, aimed at the right organ, with limits that the enthusiasts skip.

What none of those trials changed

Go back through the outcome lists. Perceived stress. Burnout. Somatic complaints. Anxiety. Distress. Job satisfaction. Mindfulness itself.

Not one of them is a count of hours, a number of deadlines, a headcount, a notice period, or a measure of who decides the order of the week. The trials were never designed to move those, which is not a flaw in the trials. It is a fact about what an intervention delivered to a person can reach.

Lovejoy, Kelly, Kubzansky and Berkman put the structural version of this in the American Journal of Public Health in 2021. Work is a social determinant of health, and the standard American response has been the employer wellness programme, aimed at individual behaviour while the conditions producing the problem stay in place. Their review of the work redesign literature concludes that the strategies with the more reliable effects are the ones that raise workers' control and voice.

A higher tolerance is a useful thing to own, and it is not a correction.

That sentence is unpopular in a lot of rooms, and it has to be said anyway. Where the load bar is the tall one, more technique produces a person who can bear an amount that should not be borne. The amount is the thing that is wrong, and nothing performed inside the person subtracts from it.

The multiplier almost nobody adjusts

Here is where the piece stops being a complaint and starts being useful, because there is a second variable in the load column and it is far cheaper to move than the quantity.

Dickerson and Kemeny went through 208 laboratory studies of acute stressors in 2004 and found that cortisol does not respond to demand in general. What produced a reliable response was a specific combination: a task the person could not control, performed where it could be judged. Neither element did much alone.

Roughly three times the effect, from putting two things together that are separately unremarkable. The amount of work was not the variable.

The field evidence points the same way and carries harder outcomes. In the Whitehall II cohort, civil servants who reported low control over their work at two separate assessments had an odds ratio of 1.93 for a subsequent coronary event compared with those who reported high control at both. The IPD-Work consortium later pooled individual data from 197,473 people and 2,358 coronary events and found a hazard ratio of 1.23 for job strain, the combination of high demand and low control, rising when events in the first years of follow-up were excluded.

So the practical consequence. A say over order, timing and notice is usually more available than less work, and on this evidence it is not a lesser substitute. It is a smaller ask aimed at a variable that was doing a lot of the damage.

There is one group-randomised trial worth knowing in detail here, because it tested exactly that and because what happened to it is instructive. Moen, Kelly and colleagues studied an initiative called STAR across 867 information technology workers, designed to increase control over work time and supervisor support for people's lives outside it. Over twelve months it reduced burnout, perceived stress and psychological distress and raised job satisfaction, and the effect ran partly through the increase in schedule control itself.

The part that usually gets left out is better than the headline. Those results held for the group surveyed before the company announced a merger. For the group surveyed afterwards, there were few effects at all. A unit of control granted inside an organisation whose ground has started moving does not do the same work, which is a more honest finding than a clean win would have been.

Two mismatches, opposite news, same total

Reading the total rather than the three bars is where most people lose the plot, and the two interesting cases look nothing alike once separated.

The second column gets filed as being a bit tired for months at a time, and the occupational health literature has been treating it as a signal in its own right for decades. Sluiter and colleagues, working with the need for recovery after work, found that it predicted later psychosomatic complaints, emotional exhaustion, sleep problems and the duration of future sickness absence, measured separately from the work demands themselves. Reserve is not a gentler way of saying load. It carries its own prognostic weight, and an empty one under an unremarkable load says the machinery is broken rather than that the week was hard. Which parts of rest actually restore what, as against which parts merely pass the time, is set out in the entry on the kinds of rest.

What the evidence for changing conditions does and does not support

The symmetrical error would be to oversell the other side, and the state of that literature should be described accurately rather than conveniently.

Fox and colleagues reviewed 83 studies of interventions that changed work itself rather than workers, in 2022, sorted into flexible work and scheduling changes, job and task modifications, relational and team initiatives, and participatory processes. Their conclusion is that such strategies have the potential to improve well-being across context-free, work-specific and work-family outcomes, and that they are a complement to individually focused programmes rather than a replacement for them.

Potential is the review's own word, and it is doing real work in that sentence. Set it next to the mindfulness literature, where a single outcome can be pooled across 43 randomised trials, and the asymmetry is obvious: there are very few designs on the conditions side as strong as the STAR trial.

The reason is mundane rather than scientific. Teaching one person to breathe is easy to randomise, cheap to run and quick to measure. Changing how a department assigns work is none of those. So the intervention that matches the problem better has the thinner evidence base, and that gap gets read as evidence that the techniques work better. It is evidence that they are easier to test.

Some stress is the point

A piece arguing that load is the neglected column could slide into implying that the goal is less of everything, and that is not what the evidence says.

Seery, Holman and Silver followed roughly 2,400 people and found U-shaped relationships rather than straight lines. A history of some but not zero lifetime adversity predicted lower global distress, less self-rated impairment, fewer post-traumatic stress symptoms and higher life satisfaction over time than either a history of none or a history of a great deal. The same group was least affected by recent adverse events.

Which fits the laboratory findings rather than contradicting them. The dimensions that decide whether a demand builds capacity or erodes it are not quantity alone. They are controllability and whether a gap follows. A life arranged to contain no demand at all is not the healthier arrangement, and nothing here is an argument for building one.

When the load bar is the tall one

The unwelcome sentence first, since softening it would waste the piece: some stress is reduced by cutting the source rather than by raising the tolerance, and where that is the situation, no practice closes the gap.

The available version is smaller than it sounds. Write the pile out and mark which items have timing that is nobody else's. People are routinely wrong about that ratio in both directions, and a marked list makes the next conversation concrete where a general account of being busy never does.

Then ask for the variable that moves. Notice before something lands. The right to finish one item before the next arrives. A say in which week a thing falls rather than in whether it happens at all. None of those reduce the quantity, which is exactly why they are grantable.

And keep the medical door open, because this profile is not specific. Months of it is produced by thyroid disease, anaemia, a sleep disorder, some medications and depression as readily as by circumstance, and those get found with ordinary tests rather than by thinking harder.

Two neighbouring questions are worth separating from this one. What a person turns into under pressure is a matter of style, and it barely moves from year to year. Whether the fire has already gone out, leaving exhaustion, distance and the sense that nothing lands, is an end state with its own assessment and it does not respond to a lighter week. This one is about how much is on someone right now and how much room is left, which is the version where the answer still changes what to do.

Sources

• Epel, E. S., Crosswell, A. D., Mayer, S. E., Prather, A. A., Slavich, G. M., Puterman, E., Mendes, W. B. (2018). More than a feeling: a unified view of stress measurement for population science. Frontiers in Neuroendocrinology, 49.
• Vonderlin, R., Biermann, M., Bohus, M., Lyssenko, L. (2020). Mindfulness-based programs in the workplace: a meta-analysis of randomized controlled trials. Mindfulness, 11.
• Singh, B., Olds, T., Curtis, R., et al. (2023). Effectiveness of physical activity interventions for improving depression, anxiety and distress: an overview of systematic reviews. British Journal of Sports Medicine, 57(18).
• Goyal, M., Singh, S., Sibinga, E. M. S., et al. (2014). Meditation programs for psychological stress and well-being: a systematic review and meta-analysis. JAMA Internal Medicine, 174(3).
• Dickerson, S. S., Kemeny, M. E. (2004). Acute stressors and cortisol responses: a theoretical integration and synthesis of laboratory research. Psychological Bulletin, 130(3).
• Bosma, H., Marmot, M. G., Hemingway, H., Nicholson, A. C., Brunner, E., Stansfeld, S. A. (1997). Low job control and risk of coronary heart disease in Whitehall II study. BMJ, 314(7080).
• Kivimaki, M., Nyberg, S. T., Batty, G. D., et al. (2012). Job strain as a risk factor for coronary heart disease: a collaborative meta-analysis of individual participant data. The Lancet, 380.
• Sluiter, J. K., de Croon, E. M., Meijman, T. F., Frings-Dresen, M. H. W. (2003). Need for recovery from work related fatigue and its role in the development and prediction of subjective health complaints. Occupational and Environmental Medicine, 60.
• Moen, P., Kelly, E. L., Fan, W., Lee, S.-R., Almeida, D., Kossek, E. E., Buxton, O. M. (2016). Does a flexibility/support organizational initiative improve high-tech employees' well-being? American Sociological Review, 81(1).
• Fox, K. E., Johnson, S. T., Berkman, L. F., Sianoja, M., Soh, Y., Kubzansky, L. D., Kelly, E. L. (2022). Organisational- and group-level workplace interventions and their effect on multiple domains of worker well-being: a systematic review. Work & Stress, 36(1).
• Lovejoy, M., Kelly, E. L., Kubzansky, L. D., Berkman, L. F. (2021). Work redesign for the 21st century: promising strategies for enhancing worker well-being. American Journal of Public Health, 111(10).
• Seery, M. D., Holman, E. A., Silver, R. C. (2010). Whatever does not kill us: cumulative lifetime adversity, vulnerability, and resilience. Journal of Personality and Social Psychology, 99(6).