The confusing part of a concussion is rarely the injury itself. It is the weeks afterwards, when the event is over, any imaging that was done came back unremarkable, and the headaches, the fog and the sensitivity to light are still there. The question that follows — how can something be wrong when nothing showed up and the injury has healed — rests on treating the injury and the symptoms as the same thing on the same timeline. They are not.

Stage One: The Injury

A concussion is a biomechanical event. Force is transmitted to the head, the brain moves within the skull, and a cascade of physiological changes follows immediately.

The important structural fact is that this is a functional disturbance rather than a visible one. Standard clinical imaging in mild traumatic brain injury is typically unremarkable — which is precisely why it is classified as mild, and why a normal scan is an expected finding rather than a reassuring one about symptoms.

Stage Two: The First Forty-Eight Hours

Care is organised around defined windows, and a 2026 clinical review divides them explicitly: the acute stage covering the first 48 hours, the subacute stage from two days to four weeks, and the persistent symptom stage beyond four weeks (Neurologic Clinics, 2026).

Those boundaries are not arbitrary administration. They mark where the expected trajectory and the appropriate management change — and where a symptom stops being part of ordinary recovery and becomes something with its own name.

Stage Three: Where Most Recoveries Go

The majority resolve. The framing that matters for the minority who do not is that persistence is common enough to be expected in a proportion of cases rather than anomalous.

A 2025 systematic review and meta-analysis states the figure directly: up to 30% of individuals who sustain a mild traumatic brain injury develop persisting symptoms after concussion (JAMA Network Open, 2025).

Nearly one in three is not a rare outcome. It is a routine one that the standard story about concussion — brief event, short recovery, back to normal — does not accommodate.

Stage Four: What Persists, and Why the Distinction Matters

The 2026 clinical review places particular focus on post-traumatic headache, describing it as the most common symptom to persist after concussion.

That specificity is useful. ‘Post-concussion symptoms’ as a category covers headache, light sensitivity, sleep disturbance, cognitive slowing and autonomic changes, and treating them as one undifferentiated cloud makes each of them harder to address.

It also separates two questions that get conflated. Whether the original injury has healed is a question about tissue. Whether the systems disrupted by it have returned to their previous functioning is a question about function — and the second can remain open after the first has closed.

Stage Five: What Predicts a Longer Course

The 2025 meta-analysis was designed around exactly this question. It searched six databases — Ovid MEDLINE, Embase, PsycINFO, CINAHL, SPORTDiscus and the Cochrane Central Register — for studies published between 1970 and February 2024, looking for clinical factors measurable within 28 days of injury that were associated with symptoms persisting beyond a month.

The design choice is the point. Restricting to factors collected in the first four weeks makes the results usable prospectively rather than only in hindsight: they are things a clinician could observe early enough for the information to matter.

The output is summary odds ratios rather than a mechanism. This is prognostic work — identifying who is more likely to have a longer course — and it does not by itself explain why.

What Management Actually Looks Like

Here the clinical review is worth quoting in substance, because it contradicts a durable piece of folk wisdom.

Management focuses on early incremental return to activities of daily living, light aerobic activity with a graded return to full play, and symptom-targeted strategies.

The instruction is early and incremental, not extended rest in a dark room. Graded return means activity is added deliberately and progressively rather than resumed all at once when symptoms happen to lift.

None of that is a self-directed protocol. Return-to-activity decisions after a head injury belong with a clinician who has assessed the person, and this article is a description of how the field frames the problem rather than guidance for managing one.

What This Evidence Does Not Establish

It does not explain the mechanism of persistence. The meta-analysis identifies factors associated with a longer course; association is not explanation, and the underlying reason symptoms persist in some people is not settled.

‘Up to 30%’ is an upper bound across a varied literature. Studies from 1970 to 2024 differ in how they define both mild traumatic brain injury and persisting symptoms, and a figure quoted as a ceiling should be read as one.

It does not diagnose anyone. Prognostic factors describe likelihood across groups. They do not determine what will happen to a particular person, and treating them as a personal forecast misuses them.

What it does support is worth holding onto: symptoms outlasting an injury is a recognised, common, staged clinical picture with defined management — not a sign that the person is imagining it, and not evidence that nothing happened. Where the pattern includes disturbed sleep, that has consequences of its own well beyond tiredness.