A century ago, when industrial work killed workers at rates we would now find unthinkable, the standard institutional response was a poster. Be careful. Stay alert. Safety is everyone’s responsibility. The worker who lost a hand had, by definition, failed to be careful enough, and the enquiry ended there. What changed everything was not better posters. It was a discipline, safety engineering, and one idea at its core: the hierarchy of controls. Eliminate the hazard where you can. Engineer the environment where you cannot. Train people systematically for the risks that remain. And treat the flimsy last layer, the goggles, the gloves, the reminders, as exactly that: the last layer, never the strategy. Institutions that adopted the hierarchy watched their fatality rates collapse over the following decades. Institutions that kept printing posters kept holding funerals.
Researcher mental health is, right now, where industrial safety was in the poster era. The hazard is documented beyond argument, the harm is measured at population scale, and the standard institutional response is a wellbeing week, a mindfulness app licence, and a leaflet: the goggles and the posters, deployed as the strategy. This article is for the people inside institutions who suspect as much and want to do better. It translates the hierarchy of controls into researcher terms, shows what the evidence says actually works at the training layer, and ends with the questions to put to any provider, including us.
Table Of Contents:
- Sabine reviews the wellbeing budget
- What the numbers make undeniable
- The hierarchy of controls, translated for a graduate school
- “I noticed earlier when I was pushing too hard”: what nine weeks actually buys
- Five questions before you sign
- What Sabine’s provision looks like now
- A word for the researcher reading this in a hard week
Sabine reviews the wellbeing budget
Sabine, coordinating her graduate school of 140 doctoral researchers, pulls up last year’s wellbeing spending for the board. A wellbeing week in May: well attended, warmly reviewed. A mindfulness app licence: 9 percent activation after month one. A yoga course: seven regulars. A mental-health-awareness talk: one afternoon, one speaker, one feedback sheet averaging 4.5. And one line she added herself after a difficult autumn: a printed card in every new researcher’s welcome pack listing the university’s psychological counselling service. Reading the list with a year’s distance, she can sort it suddenly and uncomfortably: one of these lines is a genuine clinical pathway, and every other line is a poster. The budget was spent on the bottom of a hierarchy nobody in the room had ever named.
What the numbers make undeniable
The scale of the hazard is no longer contested, and a coordinator should hold the numbers rather than the mood. Levecque and colleagues [1], studying more than 3,600 PhD students, found one in two experiencing psychological distress and one in three at risk of a common psychiatric disorder, with work organisation among the strongest predictors. The most comprehensive evidence to date, a 2026 meta-analysis in Nature Human Behaviour covering 138,446 early-career researchers, found nearly one in three reporting elevated psychological distress, with anxiety symptoms at three to five times general-population rates, and concluded the problem is systemic rather than confined to subgroups [2]. The statistic I come back to most, in my Academic Athlete article, points the same direction. Numbers like these are the equivalent of the factory injury ledger: they end the era in which the institution could believe the hazard was rare, personal, or somebody else’s.
The hierarchy of controls, translated for a graduate school
Safety engineering’s insight was that responses to a hazard are not interchangeable; they form a hierarchy, and money spent at the bottom cannot substitute for the layers above. Translated for researcher mental health, the hierarchy reads like this, from top to bottom.
Eliminate and engineer: fix what the institution controls. Workload norms, contract security, supervision quality, feedback culture: these are the engineering controls, and no training absolves them, in the same way no pair of goggles absolves an unguarded blade. The supervision literature is blunt that supervision climate predicts doctoral mental health, and an institution serious about the hierarchy starts here, with policy and supervisor development, not with anything it can buy from me.
The clinical pathway: non-negotiable, and not training. Counselling services, psychological support, clear crisis routes. This lane must exist, be findable, and be funded before anything preventive makes sense, because training is not therapy and must never be sold or bought as its substitute. If a provider blurs that line, end the meeting.
Systematic training: the layer this article is about. Between the environment and the crisis sits the largest group in the building: researchers who are not ill but are measurably strained, running the load without the skills that make it survivable. This is the layer where evidence-based, skills-focused training operates, and this is where the research is genuinely encouraging: when you pool the studies, resilience training improves resilience and mental-health outcomes, with the strongest results for people under high strain and for sustained, practice-based formats [3], [4]. High strain and sustained formats: that is a description of doctoral researchers, and a format decision, respectively.
The last layer: apps, talks, awareness weeks. These are the goggles. They have a place, awareness genuinely matters, and a talk can move someone to seek help, but they are the last layer, and an institution whose entire provision lives here has bought posters and called it a strategy. Sabine’s 9 percent app activation is what the bottom of the hierarchy looks like in a spreadsheet.
“I noticed earlier when I was pushing too hard”: what nine weeks actually buys
Because “training” is an abstraction until a participant describes it, here is what the purchase actually purchases, itemised by a researcher at the end of a recent nine-week Resilience Training cohort, in their own words:
“Over the past 9 weeks, I improved my sleep, added a stretching routine before bed, used short protected focus blocks with my phone away, and started using the ABCDE model when I get emotionally triggered. Just as importantly, I became more honest with myself, understood my core beliefs better, noticed earlier when I was pushing too hard or saying yes when I meant no, and started questioning whether my decisions come from real interest or from proving myself.”
Read that as a procurement document, because that is what it is. The first half lists installed behaviours: sleep, recovery routine, protected attention, a named cognitive technique for emotional triggers. The second half lists something subtler and more valuable: earlier detection. “Noticed earlier when I was pushing too hard” is, in safety terms, a near-miss reporting system installed inside one researcher, the capacity to catch the drift before the incident. No poster produces that sentence. No app licence has ever produced that sentence. It is the output of weeks of practice with feedback, which is why the format is most of the purchase, and why participants themselves say the right moment for it is before the strain peaks, not at it, a timing question I take up in my article on when to do resilience training.
Five questions before you sign
The hierarchy plus the evidence reduces to five questions for any provider, and a serious one will enjoy them.
Which layer are you selling me? Ask the provider to place their offering on the hierarchy themselves. Honest answers sound like “systematic training, and it assumes your clinical pathway exists”. Evasive answers sound like “holistic wellbeing solution”.
Weeks or days, and what happens between sessions? The research favours sustained, practice-based formats; our own Resilience Training runs nine weeks for exactly that reason, and I wrote up the memory research on why one-day formats fade in my post on the Ebbinghaus forgetting curve.
Where is the clinical boundary, and what is your referral practice? Training serves the strained, not the ill; ask how trainers recognise the difference and what they do when they meet it. The answer tells you whether the provider has thought about the humans or only the curriculum.
How do you measure outcomes? Before-and-after measurement with a validated instrument, completion rates, and participant accounts specific enough to check. Our cohorts begin and end with the Personal Resilience Indicator so the change is shown, not asserted; ask every provider for their equivalent, and treat its absence as an answer.
What does this NOT fix? The best question in the set. A provider who answers “your workload norms, your contracts, your supervision culture, and anyone who needs clinical care” understands the hierarchy and their place in it. A provider who answers “nothing” is selling posters with better production values.
What Sabine’s provision looks like now
Sabine did not cancel the wellbeing week; awareness has a place, at its layer. What changed is the shape of the budget. The clinical pathway moved to the top of the page and got a named liaison instead of a card in a welcome pack. A nine-week, measured training became the development line, with before-and-after results she can put in front of the board. And the apps and talks moved to the bottom of the page, labelled as what they are. At the next board meeting the question was not “was it well attended” but “what moved”, and for the first time the answer had numbers in it. That is the entire difference between spending on wellbeing and engineering for it.
A word for the researcher reading this in a hard week
And if you are not buying anything, if you are a doctoral researcher or postdoc who found this page because the load has become heavy, then before any of the institutional logic above, one thing matters more. If what you are carrying feels bigger than strain, if low mood has spread beyond work, if sleep fails even when you protect it, or if hopelessness has arrived, please speak to a doctor, a therapist, or your institution’s counselling service; that is the clinical pathway, it exists for exactly this, and using it is an accurate reading of the situation, not a failure. This is a sensitive area, and if any of it feels personally close, finding the right support matters more than anything else on this page. If, instead, you recognise the strained-but-functioning middle, the load running ahead of your recovery, then the training layer was built for you: the Resilience Training for Researchers is nine weeks of implementation with over 1,500 individuals trained across our trainings at more than 30 institutions, and forwarding that page to your coordinator, with one line about why, is how institutional provision actually changes. Coordinators: book the intro call and lead with the fifth question. Industrial safety stopped printing posters a century ago and started engineering, and the fatality curves did the rest of the arguing. Researcher mental health is waiting for the same decision, one institution at a time.
FAQ
What is mental health training for PhD students?
Systematic, skills-based training for the strained-but-functioning majority: stress regulation, recovery behaviour, boundary-setting, cognitive techniques, practised over weeks with feedback. It sits between what the institution must engineer (workload, supervision) and the clinical pathway (counselling, therapy), and it is not a substitute for either. The evidence favours sustained, practice-based formats over one-day events.
Does mental health training for researchers actually work?
At its layer, yes, under conditions the research specifies. Two major meta-analyses find skills-based programmes improve resilience and mental-health outcomes, with the strongest effects for people under high strain and for sustained formats, which describes doctoral researchers and rules out one-day formats respectively. What it does not do is fix workload norms, replace therapy, or absolve supervision culture, and honest providers say so.
What should a university buy for PhD student mental health?
In hierarchy order: engineering fixes it controls (workload, contracts, supervision development); a funded, findable clinical pathway; sustained, measured, skills-based training for the strained majority; and only then awareness events and apps, labelled as the last layer, not the strategy. A budget that lives entirely in the last layer is posters. The numbers, one in two PhD students in distress, deserve better than posters.
Is resilience training the same as therapy?
No, and the boundary is a selection criterion. Therapy treats clinical conditions through the clinical pathway; training builds skills in a non-clinical population before problems become clinical. A provider who blurs the line, or sells training as a counselling substitute, has failed the most basic test. Ask any provider where their clinical boundary sits and what their referral practice is.
References
1. Levecque, K., Anseel, F., De Beuckelaer, A., Van der Heyden, J., & Gisle, L. (2017). Work organization and mental health problems in PhD students. Research Policy, 46(4), 868–879.
2. Dreisoerner, A., Goetz, V., Frohnmayer, D., Tran, U. S., Voracek, M., & Nater, U. M. (2026). Prevalence and severity of mental health problems in early-career researchers: a systematic review and meta-analysis. Nature Human Behaviour.
3. Joyce, S., Shand, F., Tighe, J., Laurent, S. J., Bryant, R. A., & Harvey, S. B. (2018). Road to resilience: a systematic review and meta-analysis of resilience training programmes and interventions. BMJ Open, 8(6), e017858.
4. Vanhove, A. J., Herian, M. N., Perez, A. L. U., Harms, P. D., & Lester, P. B. (2016). Can resilience be developed at work? A meta-analytic review of resilience-building programme effectiveness. Journal of Occupational and Organizational Psychology, 89(2), 278–307.
Last Updated on August 3, 2026 by Dr Nadine Sinclair
Dr Nadine Sinclair
Dr. Nadine Sinclair is a molecular biologist, strategy consultant and co-founder of Mind Matters. A scientist by training and at heart, she conducted her doctoral research at the Max Planck Institute for Biophysical Chemistry. She has 30,000+ hours of hands-on project management experience, built across 18 years as a strategy consultant at McKinsey and as an independent consultant, working with research institutions, foundations, pharmaceutical and biotech companies and governments. She is the co-creator of the Personal Resilience Indicator and the author of On Track. Since 2018, over 1,500 researchers from more than 30 institutions have trained with Mind Matters.