The therapy boom filled waiting rooms. It didn’t clarify who belongs in them

The word “anxious” has colonized the English language. We are anxious about deadlines, anxious about commutes, anxious about what we said at dinner, anxious about the future of work and the shape of the planet and whether our friends actually like us. Mental health awareness campaigns spent the last decade teaching people to put language around their inner lives. They succeeded. What they did not build was the vocabulary for the distinction that follows.

Stress, anxiety, and clinical anxiety disorders are three different phenomena. They are not three points on a single dial, moving seamlessly from mild to severe. They operate through different mechanisms, require different interventions, and carry different prognoses. But the cultural conversation around mental health has flattened these distinctions into a single spectrum — one that runs from “a bit stressed” through “definitely anxious” to “probably should see someone,” with the clinical threshold somewhere in the blur between the second and third points. The result is a paradox: we have more words for inner distress than any previous generation, and less capacity to tell when those words describe something that requires clinical care.

The distinction starts with mechanism. Stress, as the American Psychological Association defines it, is a response to an identifiable external demand. A deadline, a confrontation, a medical diagnosis, a layoff notice: stress exists in conversation with a specific trigger and resolves, largely, when that trigger resolves. The physiological response is adaptive — the cortisol spike, the raised heart rate, the narrowed focus — because these states evolved to help organisms navigate a real threat. The problem stress creates, clinically speaking, is not what it does while it operates but what prolonged exposure does to the system that hosts it. Chronic stress, sustained over months, sensitizes the threat-response circuitry in ways that lower the threshold for the next state in the taxonomy.

Anxiety is what emerges when the threat-response system has learned to operate without a trigger. The APA defines it as a response to a perceived rather than a concrete danger — internal, anticipatory, and disproportionate to any identifiable external cause. Anxiety can exist without a stressor; it does not resolve when the external situation resolves. What makes it clinically interesting is also what makes it culturally difficult to locate: it is subjective, invisible, and capable of mimicking common human experiences like worry, vigilance, and social self-consciousness. An anxious person and a person under normal situational stress can report almost identical inner experiences. The difference lies in duration, persistence, and what happens when the external trigger disappears.

A clinical anxiety disorder requires something beyond anxious feelings: it requires that those feelings impair function, that they persist across contexts and time, and that they cannot be explained by another medical condition or life circumstance. A 2025 analysis in Depression and Anxiety, drawing on the Global Burden of Disease Study, counted 359 million people worldwide meeting criteria for anxiety disorders — roughly 5 percent of the global population, not the majority. That number almost certainly does not include everyone who is anxious, or everyone who is significantly distressed. It is a count of people whose anxiety meets the clinical bar.

The confusion between these three states is not merely semantic. The treatments that work for clinical anxiety disorders — cognitive behavioral therapy targeting avoidance loops, medication for generalized anxiety or panic disorder, exposure work for phobias — are not the same interventions that help someone manage a stressful work period or learn to sit with existential worry about the future. And applying clinical-grade intervention to subclinical distress may not only fail to help; it can, in some cases, entrench the problem.

A finding from Australia’s Better Access initiative illustrates this with uncomfortable precision. The government mental health program increased access to subsidized therapy. Analysis of its outcomes found that patients presenting with mild anxiety or depression who entered clinical treatment were more likely to worsen than to improve. This finding, cited by Nick Haslam of the University of Melbourne in his 2026 paper “Concept Creep and the Mental Health Crisis,” is not an argument against therapy. It is an argument that clinical structures designed for moderate-to-severe disorder may not serve — and can actively harm — people whose distress falls below the clinical threshold.

Haslam’s concept creep theory offers a structural explanation for how the confusion arose. The idea, developed in work published in 2025 in SAGE journals, is that mental health concepts undergo historical expansion: their definitions grow broader, the threshold for application drops, and what once described severe clinical conditions gradually comes to encompass milder experiences. While the public has become better at recognizing when someone has a mental health condition, it has simultaneously become worse at recognizing when someone does not. Mental health literacy expanded one side of the ledger and contracted the other.

The counter-argument deserves its full weight: before the vocabulary expanded, millions of people with genuine clinical disorders went unrecognized — by themselves, by their families, by their doctors. The expansion of mental health language has unquestionably pulled some people toward care they genuinely needed. According to GrowTherapy’s 2026 survey of therapist practices, anxiety and stress now drive 34 percent of all therapy-seeking in the United States. For the people in that 34 percent who have clinical disorders, the awareness movement was a route to treatment they might otherwise never have reached.

But the same GrowTherapy data shows access to mental health services declined to 47.4 percent of people who sought care in 2026, down from 50 percent the prior year. The waiting list crisis is partly a mismatch problem. Clinical resources are finite. If those resources are disproportionately absorbed by subclinical stress, then people with disorders severe enough to disable function wait longer, arrive later, and present in worse shape. Research published in Frontiers in Psychology distinguishes between state anxiety — transient, adaptive — and trait anxiety, a stable predisposition carrying genuine clinical risk. Someone with chronic trait anxiety may encounter the cultural message that “anxiety is just stress” and conclude their daily functional impairment is normal.

What we know / What remains in dispute

What we know: Stress, anxiety, and clinical anxiety disorders are mechanistically distinct and require different responses. The Global Burden of Disease Study counts 359 million people globally meeting clinical criteria — roughly 5 percent of the world population. Nick Haslam’s research documents measurable concept creep: the public has grown better at identifying mental illness and worse at identifying its absence. Australia’s Better Access data shows mild-symptom patients in clinical programs fared worse, on average, than those who did not access clinical treatment.

What remains in dispute: Whether mental health awareness has helped more people than it has confused. Whether the correct response to overcrowded waiting rooms is better triage or expanded capacity — or both. Whether concept creep is a necessary expansion of who can claim legitimate suffering, or a dilution of a vocabulary that only functions when it is precise. And whether the DSM threshold itself is drawn correctly: critics argue the line between disorder and ordinary distress was set by professional consensus, not by biology.

What the debate converges on: the distinction matters practically. Mislocating stress as disorder contributes to clinical overcrowding and, under some conditions, harm. Mislocating disorder as stress leaves people in functional impairment they have learned to call normal. The vocabulary project was necessary. The triage project that should have accompanied it is still waiting to be built.