You have almost certainly been told that some small daily practice will change how you feel, by someone with no obligation to tell you how well it works. This page is the obligation. Every entry carries two grades, and where the evidence is thin we print that in the same size type as everything else.
How to read a grade
Two questions get asked about any technique, and fusing them is how the field misleads people. Does doing this shift mood? is one question. Do we understand why? is a different one. A practice can work reliably for reasons nobody has pinned down, and a practice can come with a confident brain-chemistry story that turns out to be false. So each entry is graded twice.
- Outcome runs A to E: Strong, Moderate, Preliminary, Theoretical, Contradicted. It answers only whether the effect shows up.
- Mechanism is Established, Plausible, Speculative, or Misattributed. That last one is the interesting case: the practice may help while the popular explanation for it is simply wrong.
- Our confidence is confidence in our own grade, not in the technique. Where it reads Medium, treat the grade as provisional. We would rather show you the wobble than hide it.
Nothing here is a prescription, and none of it is treatment. It is a map of what the research currently supports, drawn so you can see the thin ground as clearly as the solid.
-
Outcome A · Strong Mechanism Plausible
Structured exercise
About as reliable as this library gets. Walking, running, strength work and yoga all reduced depressive symptoms across a large network of trials, and more of it tended to help more.
Why it is graded this way: The honest “why” is that there isn’t one why. It runs through several routes at once — neurotrophic, inflammatory, sleep, a sense of mastery, often company. The single endorphin rush of popular telling is not what the evidence describes.
-
Outcome A · Strong Mechanism Established
Sufficient, regular sleep
The clearest two-way relationship in this whole library. Short or irregular sleep degrades mood and the ability to regulate it; improving sleep improves both. Protect this before adjusting anything downstream of it.
Why it is graded this way: Well characterised, and unusually so for this field. This is one of the few entries where the mechanism is not the weak link.
-
Outcome A · Strong Mechanism Plausible
Behavioural activation
Scheduling and then doing something that matters to you, before the motivation to do it arrives. It is a frontline component of depression treatment and holds up comparably to CBT in several analyses.
Why it is graded this way: The framing that makes it work is counterintuitive and worth stating plainly: action comes first, mood follows. Waiting to feel like it is the part that keeps people stuck.
-
Outcome A · Strong Mechanism Plausible
Social connection
Among the strongest and most consistent links in this library, and almost entirely an association. The headline evidence follows people who are more connected; it does not test what happens when someone sets out to become so.
Why it is graded this way: Plausible rather than settled, and the causal arrow runs both ways: low mood withdraws people, and withdrawal lowers mood further.
-
Outcome B · Moderate Mechanism Established
Cutting late-day caffeine
Reliable, but read the route carefully: the mood benefit is mostly downstream of sleeping better, not a direct effect on mood.
Why it is graded this way: Caffeine’s half-life runs to roughly five hours and it blocks adenosine, so an afternoon cup is still present at bedtime and still degrading the sleep that carries the mood effect. A clean example of an indirect causal chain, which is why we file it as a sleep lever.
-
Outcome B · Moderate Mechanism Plausible
Time in nature
A real benefit to mood and stress. A graded umbrella review rates the evidence for improved positive affect and reduced negative affect as high quality, and for depression as moderate.
Why it is graded this way: Loosely specified. Attention restoration and stress reduction are the usual accounts, and both are hard to separate from the exercise, daylight and company that tend to arrive with a walk outdoors.
-
Outcome B · Moderate Mechanism Plausible
Mindfulness meditation
Real reductions in anxiety and depressive symptoms, of small-to-moderate size, which shrink substantially when compared against an active alternative rather than a waiting list. Useful and low-risk. The cultural volume runs well ahead of the data.
Why it is graded this way: Plausible, with publication-bias concerns documented in the literature itself.
-
Outcome C · Preliminary Mechanism Speculative
Reducing or structuring social-media use
Plausibly helps some people. The trial evidence is genuinely mixed: some studies find small wellbeing gains from cutting back, others find nothing. This is an honest “it depends”, and a good candidate for testing on yourself rather than taking on authority.
Why it is graded this way: Speculative. The effect varies by person and by platform, and the field has not settled why.
-
Outcome C · Preliminary Mechanism Speculative
Expressive writing
Writing about a specific stressful experience helps some people process it. Across the literature the benefits are inconsistent and depend heavily on context.
Why it is graded this way: Speculative. Several accounts compete and none has settled the matter.
-
Outcome D · Theoretical Mechanism Misattributed
“Dopamine detox”, as marketed
The signature case for grading two axes rather than one. The underlying behaviour, cutting back on compulsive overstimulation, has a reasonable rationale. The mechanism sold alongside it is false.
Why it is graded this way: You cannot drain, fast, or reset dopamine, and dopamine is not the pleasure chemical of popular telling — it tracks wanting and prediction error. Keep the behaviour if it helps you. The brain-chemistry story attached to it does not survive contact with the research.
-
Outcome D · Theoretical Mechanism Misattributed
“Serotonin foods”
The popular version of this claim is wrong, though not in the way “debunked” usually implies. Eating tryptophan-rich foods does not lift mood the way the claim describes; what evidence exists, mostly on tryptophan-enriched preparations rather than dinner, points to modest and conditional effects rather than none.
Why it is graded this way: This is the axis that carries the entry. Tryptophan competes with other large neutral amino acids for transport into the brain, so the chain from mood food to serotonin to happiness breaks at its first link. What would matter is the ratio between them, not the amount of tryptophan, and an ordinary meal does not shift that ratio. Diet quality matters to mood for other reasons. This particular explanation is the misattribution.
DRAFT — eleven entries at launch, and this is deliberately not everything we have graded. The rest are in the queue rather than in a drawer, and they arrive as each one earns its citation. Every grade carries the date we last reviewed it. Three of those dates are older than the others, because three of these entries have not been re-read since they were first graded, and rounding that up would have been the easy lie.