Physical activity has some of the better evidence of any behavioural intervention for mood, and reasonable evidence for aspects of cognition. It is also advice that reliably fails to land, partly because it is delivered in a form that assumes capacity nobody short of energy has.
The version worth having is more modest.
What is reasonably well supported
Mood. Regular physical activity has a genuine effect on depressive symptoms, and for mild to moderate depression the evidence is strong enough that it appears in treatment guidelines rather than in lifestyle sections. The effect size is meaningful, not a rounding error.
Anxiety. Reasonable evidence, particularly for the physical components — tension, restlessness.
Sleep. Regular activity improves sleep quality, which then improves next-day cognition. A good deal of the cognitive benefit may travel by this route.
Short-term alertness. A single bout of moderate activity produces a measurable improvement in attention and executive function for a period afterwards. This is the most immediately usable finding: a twenty-minute walk before demanding work is not a metaphor.
Long-term brain health. Population-level associations between activity across midlife and later cognitive outcomes are consistent. This is a risk claim over decades, not a treatment for this month.
What is oversold
Specific protocols. The confident claims about particular intensities, durations and timings producing particular neurological effects run well ahead of the evidence. The robust finding is that regular moderate activity helps; the details are far less settled than the internet suggests.
Exercise as a replacement for treatment. For mild to moderate depression it is a legitimate treatment option. For more severe illness it is an adjunct, and framing it as an alternative does real harm.
The idea that more is proportionally better. Most of the benefit appears at fairly modest levels. The gap between nothing and a little is much larger than the gap between a little and a lot.
The version for people with no energy
Standard advice assumes you can get to a gym three times a week. If cognitive fatigue or low mood is the problem, that assumption is the reason the advice does not work.
The useful unit is a ten-minute walk. Outside, ideally. That is enough to produce the short-term alertness effect, it requires no equipment, no change of clothes and no decision, and the transition cost is nearly zero — which matters enormously if starting things is the difficulty.
Attach it to something that already happens. After lunch. Before the demanding block. On the way back from the shop. A new standalone habit requires activation; an appendage to an existing one does not.
Put it before the work, not after. The alertness effect is most useful going into demanding work. Most people schedule exercise for the end of the day, when it competes with nothing left.
Consistency over intensity. Ten minutes daily beats an hour once a fortnight, both for mood and for the likelihood of continuing.
Outside counts double. Daylight helps sleep timing, and there is reasonable evidence for green space having an effect of its own. Two mechanisms for one walk.
Where the standard advice is wrong for you
One important exception, and getting it wrong causes harm.
If exertion reliably makes you worse afterwards — particularly a day or two later — then progressive increase is not the right approach, and being told to push through can make things considerably worse. That pattern of delayed worsening after exertion is characteristic of some post-viral and fatigue conditions, and the correct approach there is pacing: staying within a ceiling and increasing only from stability.
If that describes you, say it explicitly to a clinician. It changes what should be recommended, and it is advice people are still given in the wrong form.
The distinguishing question is not whether exercise is tiring — it is whether the following day or two are worse than they would otherwise have been.
How to tell whether it is doing anything
Two or three weeks, one line a day: whether you moved, and how the day went. That is enough to see whether the days you walked were better, and whether the effect was on the same day or the next.
And then the usual restraint: if they correlate, that is an association in your record. It might be that walking helped. It might be that the days you felt better were the days you were able to walk — which is the reverse arrow and is entirely plausible.
Distinguishing them is hard with observation alone. The nearest approach is to walk deliberately on some bad days and see, which is unappealing and is the only version of the experiment that answers the question.
The honest summary
Movement is one of the better-supported things on any list of lifestyle interventions for mood and sleep, with a real short-term effect on alertness. It is not a treatment for everything, it is not a substitute for treatment where treatment is needed, and the specific protocols are less established than the general finding.
Ten minutes outside before the hard part of the day is the version most likely to be done and most likely to help. That is a low bar on purpose.