Shinrin-yoku, generally translated as forest bathing, describes spending time in forest environments attentively, without exercise as the objective.
The term originated in Japanese public health policy in the 1980s, and a substantial research literature has developed around it.
The core finding
Studies comparing time in forest environments with time in urban environments have consistently reported differences in physiological and psychological measures.
Commonly reported: lower cortisol, lower blood pressure, lower heart rate, reduced sympathetic and increased parasympathetic activity, and improvements in self-reported mood and anxiety.
Reviews of this literature generally support the conclusion that time in natural environments produces measurable short-term effects on stress markers and mood.
That much is reasonably robust across studies and populations.
The mechanism question
Where the evidence becomes contested, and where the popular claims concentrate.
Phytoncides. The most cited proposed mechanism. Volatile compounds released by trees, claimed to have effects on human immune function, particularly on natural killer cell activity.
Studies reporting increases in these cells following forest visits exist, generally with small samples and without adequate control for other differences between forest and urban environments.
Some studies have exposed participants to these compounds in hotel rooms and reported effects, which addresses one confound and leaves others.
The honest summary: an interesting hypothesis with suggestive evidence and considerable uncertainty.
Attention restoration. The proposal that natural environments engage attention effortlessly — soft fascination — allowing recovery of directed attention capacity depleted by demanding tasks.
This theory has a substantial literature and reasonable support, and it makes predictions that have been tested.
Stress reduction through evolutionary preference. The proposal that humans have an evolved preference for certain environment types, and that these produce reduced stress response.
Plausible, difficult to test, and frequently stated more confidently than the evidence supports.
The mundane explanations. Forests are quieter, cooler, involve walking, involve being away from work, and involve reduced sensory demand.
Any of these could account for a substantial portion of the effect, and studies rarely control for all of them.
Which is the main methodological problem in this area — the comparison is between two environments differing in many respects, and attributing the effect to any particular difference requires controls that are hard to construct.
Dose and duration
Practically useful findings.
Research examining the relationship between time in nature and wellbeing has suggested thresholds around two hours per week associated with better reported health and wellbeing, with effects appearing to plateau beyond a certain point.
That figure has circulated widely and comes from observational research with the usual causal limitations.
Studies of single visits generally find effects from durations of twenty minutes to a couple of hours, with longer exposures not producing proportionally larger effects.
Which suggests the practical guidance is modest amounts regularly rather than occasional long expeditions.
What's not required
Worth stating because the practice has been commercialised.
Guided sessions, specific exercises and paid programmes are available and the research does not suggest they're necessary.
Studies generally involve participants simply being in the environment, walking or sitting.
Nor does it require a forest specifically. Studies in various natural environments — parks, coastal areas, green space generally — have found effects, and urban green space appears to produce benefits at a scale accessible to people who can't reach forests.
That last point is the practically important one, since access to natural environments is unevenly distributed and the interventions that matter are those available to people in cities.
What appears to help
From the research and from the practice traditions.
Attention rather than exercise. The effects on stress appear more consistent for attentive presence than for vigorous activity in the same environment, though exercise has its own benefits.
Sensory engagement. Deliberately attending to what you can see, hear and smell, rather than walking while thinking about something else.
Without a device. Studies examining phone use in natural environments have found it reduces the observed benefits, which is unsurprising and worth stating.
Regularity over intensity. Consistent with the dose findings.
The policy dimension
The more consequential implication.
If access to green space affects health, then its distribution is a health equity issue, and it's distributed very unevenly — with lower-income urban areas typically having less.
Research examining green space and health outcomes at population level has found associations, including in some studies with reduced health inequality where green space is more available.
Which makes urban green space provision a health intervention rather than an amenity, and a considerably cheaper one than most.
That's arguably the most useful thing to take from this literature, and it receives less attention than the phytoncide claims.
Windows and images
A body of research relevant to anybody without easy access to natural environments.
A frequently cited study found that hospital patients with a window view of trees had shorter recoveries and required less analgesia than those with a view of a wall. The study is small and has been influential, and subsequent work has broadly supported that views of nature affect recovery and stress measures.
Studies using images and video of natural environments have also reported effects, generally smaller than actual exposure.
Which suggests something available to people in hospitals, offices and homes without gardens: a view, a plant, an image. The effects are modest and they are not nothing, and they are considerably cheaper than the alternative.