Mental Health At Work

Who Is It Actually For?

On March 10, I attended the Health and Wellbeing at Work conference in Birmingham. Across a full day of sessions, government policy reviews, HSE presentations, talks on stress, burnout, and high performance, a pattern emerged that nobody named directly. Mental health at work, as it is currently framed, asks middle leaders to carry something they were not trained to carry, while quietly teaching employees that their wellbeing is someone else's job.

This is worth examining carefully, because policy writers' intentions are not wrong in and of themselves. However, they do emphasise problems and solutions while sidelining implementation .

The Scale of the Problem

The data is serious. Around 20% of people in the UK are currently not working due to ill health, a figure that is rising as the population ages. Work-related stress costs the economy an estimated £15 billion a year. GP appointments are being consumed, somewhere between 60 and 70% of them according to figures cited at the conference, by stress, depression, and anxiety. These are not marginal problems.

The policy response, most recently codified in the Mayfield Review and the Keep Britain Working initiative, is to make the workplace itself a site of health intervention. The logic is not unreasonable: people spend most of their waking hours at work, and good work, fairly paid, purposeful, relational, stable, does appear to support health. The Marmot evidence on social determinants is well established. Work is not just income; it is structure, identity, community.

Where the Load Lands

Across multiple sessions, the answer was consistent: the line manager. Build check-in culture. Notice changes in behaviour. Ask people how they are on a scale of one to ten. Model psychological safety. Hold boundaries around overworking. Create space for curiosity and divergent thinking. Make the case to your team; make the case to yourself. One session framed this explicitly as making space for people to talk to managers so they can intervene. Another described the manager's role in terms drawn almost directly from clinical practice: validation, needs assessment, titrated response depending on the score given.

This is not supervision. It is not management. It is something closer to embedded pastoral care, distributed across an organisational tier that is already, by most accounts, operating at capacity.

Middle leaders are a specific kind of worker. They carry the pressure from above, targets, strategy, accountability, while managing the human reality below. They are, structurally, the most exposed people in most organisations. The burnout literature is consistent on this point: the combination of high demand and limited control is the most reliable predictor of breakdown. Middle leaders have high demand almost by definition. Their control is often illusory. And now, incrementally, policy is asking them to become the primary interface between organisational life and employee mental health.

There is an argument that this is simply what good management looks like. That is partly true. Attentiveness, psychological safety, reasonable responsiveness to the people you work with: these are legitimate leadership competencies. The problem is not the principle; it is the scope creep. When a manager is trained to conduct check-ins using the language of therapeutic assessment, when their role is framed as one of intervention, when the measure of their effectiveness starts to include the mental health outcomes of their reports, the role has changed in kind, not just degree. And nobody has asked whether they want it, whether they are resourced for it, or what happens to them when it goes wrong.

The Under-Examined Side of the Equation

If a manager's job is to notice when you are struggling, create space for you to speak, validate your experience, and take action, then the implicit message is that your wellbeing is something that happens to you, and that the appropriate response is to surface it upward. This is not a neutral message. It trains a particular relationship with one's own internal experience: reactive, other-directed, contingent on being seen and held by someone with organisational authority over you.

Stress is not simply an external event that lands on a person. The neuroscience is reasonably clear on this: what we experience as stress is a construction, a prediction the brain makes based on prior experience, current context, and the body's internal state. That prediction is not fixed. It is not immune to intervention. But the intervention that changes it is not primarily managerial. It is in how a person learns to read their own experience, regulate their own physiology, and make meaning of pressure rather than waiting for pressure to be removed.

What the Policy Sessions Missed

None of this featured prominently at the conference. The dominant register across the policy sessions was structural: leading indicators, surveillance systems, compliance frameworks, escalation pathways. The question driving most presentations was not "what does the person under pressure actually need?" but "what can the organisation be required to do, and how do we measure whether it has done it?"

This is not an unreasonable place to start. Organisations do cause harm. Work conditions do shape health outcomes. The Marmot evidence on social determinants is not in dispute, and the HSE data on work-related stress, 1.9 million affected workers, £15 billion in annual costs, is serious enough to warrant serious structural attention. The problem is not that policy is addressing these things. The problem is that policy has come to treat structural intervention as the whole of the answer rather than part of it.

One speaker noted, correctly, that there are too many wellbeing measures, too little prevention, and too many testing measures. It was one of the sharper observations of the day. But the prevention being imagined was still structural: better organisational design, more accountable leaders, clearer frameworks for identifying risk. The possibility that prevention might also require something of the individual, a different relationship with their own stress responses, a developed capacity for self-regulation, a conscious choice to engage rather than wait, did not enter the frame.

The session on work and health outcomes illustrated this particularly clearly. The discussion of presenteeism circled for some time without arriving at a usable definition, which is telling in itself. What was consistent was the assumption that the employer is the primary responsible agent: physically, mentally, and in terms of the conditions that allow people to function. Somebody in the room asked whether anyone had thought to ask employees what they actually need. The question was noted and moved past.

The economic framing ran through most sessions as an unexamined assumption. Health at work was repeatedly anchored to income-based outcomes: employment status, productivity, reduced sickness absence, return-to-work rates. These are legitimate concerns. They are also a particular way of defining what wellbeing is for, and it is worth noticing what that definition excludes. A person can be in work, productive by measurable standards, and experiencing sustained internal distress that no surveillance system will detect and no compliance framework will reach. The metric captures the behaviour. It says nothing about the person producing it.

What was conspicuously absent across the policy sessions was any serious account of individual agency. Not in the reductive sense of telling people under pressure to simply try harder or be more resilient, a framing that has rightly attracted criticism for letting organisations off the hook. But in the more fundamental sense that a person's experience of pressure is not only a function of their external conditions. It is also a function of how they have learned to construct that experience, what meaning they make of it, what stories they carry about themselves and their capacity to manage difficulty. Policy cannot reach that. It can create better conditions around a person. It cannot change what happens inside them. That work is irreducibly personal, and the conference had almost nothing to say about it.

Almost nothing. One session was the exception.

What Dame Laura Kenny Said Instead

This was thrown into relief by one of the conference's more unexpected sessions, a keynote from Dame Laura Kenny, Britain's most decorated female Olympian and five-time Olympic gold medallist in track cycling. What she described was not a wellbeing programme. It was a practitioner's account of what psychological self-management actually requires when the pressure is real, sustained, and operating at the furthest extreme of human performance.

Several things she said deserve attention in this context.

She opened by questioning the definition of high performance itself. Working until you can give no more, she suggested, is not high performance. It is a path to breakdown. Human beings are not machines. The F1 car analogy does not hold. This matters because the same assumption, that sustained output is the measure of a committed employee, runs silently through most high-pressure professional environments, and nobody in the policy sessions named it as a problem.

The more searching observation was about identity. She drew a clear line between "I am the athlete, I am the winner" and "I am the person that wins." The difference is not semantic. When a person fuses their identity with a performance role, any disruption to performance becomes a threat to the self rather than a problem to be solved. The person who maintains that separation has access to a different quality of response under pressure. That separation cannot be granted by a manager or encoded in a check-in protocol. It has to be constructed from the inside, over time, through deliberate self-examination.

She was equally direct about the perils of believing oneself to be indestructible or indispensable. At the level of elite sport, the trap is identifiable: the athlete who confuses their value as a person with their value as a performer, who cannot step back because stepping back feels like disappearing. The damage this causes, she made clear, goes deeper the longer it is sustained. Knowing when to stop and what you are stopping for requires a self that exists independently of the performance. Without that, the outcome orientation that drives high achievement becomes the thing that prevents recovery.

On care, she was direct. Viewing it as a drain on resources, as something that conflicts with being the kind of person who performs at the highest level, is self-destructive physically and psychologically. The refusal to seek support is not toughness. It is a cognitive error rooted in misidentification. This is recognisable well beyond elite sport. The professional who does not engage with the wellbeing initiative may not be unaware of their own stress. They may have built an identity in which acknowledging it feels like professional failure.

She also addressed communication within teams, and the distinction between free speech and blame. Psychological safety, in her framing, does not mean saying whatever one feels like saying. It means creating the conditions in which honest feedback can be given without framing it as accusation. Free speech within a team is not disrespectful speech. It is not blaming speech. The two are frequently conflated, and the conflation is what prevents the kind of open exchange that actually supports both performance and wellbeing. Boundaries matter here too: protecting one's own boundaries is not defensiveness. It is a precondition for sustainable functioning.

What connects all of this is that none of it is deliverable by policy. It requires a person to look at their own patterns of thought, their own identity constructions, their own relationship to pressure and care, and make deliberate choices about how to engage differently. The conference's architecture had no room for that conversation. The assumption running through most of the sessions was that the right combination of managerial attentiveness, structural support, and compliance frameworks would produce wellbeing outcomes. Dame Laura Kenny's session, perhaps unintentionally, made the case that the most important work happens somewhere else entirely.

The Cost of Misplaced Architecture

Middle leaders are not therapists. Employees are not patients waiting to be seen. The conflation of these roles is understandable given the scale of the problem and the genuine desire to do something about it. But good intentions distributed through the wrong architecture do not produce good outcomes. They produce overloaded managers, undertrained in what they are being asked to do, and employees who have quietly learned that their mental health is something that gets managed at them rather than by them.

That is not wellbeing. It is the appearance of it.

Where Change in Mental Health at Work Actually Begins

Real change in how one experiences pressure at work does not begin with a policy, a check-in, or a manager who has been trained to ask the right questions. It begins with a decision made by the person under pressure, for themselves. This matters because of how change actually works: the brain constructs experience from prior learning, expectation, and the stories we have rehearsed about ourselves. Those stories do not update because someone else noticed they were causing problems. They update when the person living them chooses to look at them directly, gets curious about them, and decides to build something different. That decision is not a small thing. It is the point at which suffering stops being something that happens to you and becomes something you are actively working on. The investment that follows, in therapy, in practice, in the slow work of changing habitual patterns of thought and behaviour, is real precisely because you chose it. Nobody can construct that for you. The architecture of workplace wellbeing, however well-designed, can only ever create the conditions around that choice. The choice itself is yours.