What the forest-bathing research does and does not show

The short answer #

In short-term studies, people who spend time among trees tend to show lower stress measures and to report a better mood than people who spend similar time in a city. That result turns up again and again, from different teams in different countries. It is also small to moderate in size, usually measured within hours, drawn mostly from young or middle-aged volunteers in Japan and elsewhere in Asia,¹ and hard to separate from walking, quiet and simple expectation.

Nothing in this research shows that forest bathing prevents or treats an illness. The authors of the reviews say so themselves, usually in their final paragraphs.

How this was checked. Every figure below comes from the paper's own abstract or from its openly available text, which is linked in the notes. Where a number rests on a report of the study instead of the paper itself, the note says so. Nothing here is medical advice.

The studies at a glance #

Ten sources do most of the work in this essay. Each is discussed below, with its limits.

On a narrow screen, swipe the table sideways.

The founding experiment #

The word shinrin-yoku, usually translated as forest bathing, was coined in 1982 by Japan's Ministry of Agriculture, Forestry and Fisheries.² The experiment people most often cite came from Japanese researchers led by Bum-Jin Park and Yoshifumi Miyazaki.

In each of 24 forests across Japan, twelve young men walked for about sixteen minutes and then sat looking at a forest area for about fourteen. On another day, the same twelve did the same in a city area. Half went to the forest first and half to the city, and all of them stayed in identical rooms and ate identical meals. In total there were 280 men, all university students with an average age of about 22.³

After the forest session the men had lower salivary cortisol, a lower pulse and lower blood pressure than after the city session. Their heart-rate readings also pointed to more activity in the parasympathetic nervous system, the body's rest-and-recover side, and less in the sympathetic, or fight-or-flight, side.

The design has real strengths. Every man did both conditions, so differences between people largely cancel out, and the order was balanced. What it cannot tell us is just as important. It does not tell us whether the result holds for women, for older people or for people who are ill. It measured effects within minutes, so it says nothing about how long they last. And it cannot say whether the trees were the cause, or whether quiet, fresh air and half an hour away from traffic would have done much the same. The men also knew where they were.

What happens when the studies are pooled #

A single experiment can mislead. A systematic review searches for all the studies on a question by stated rules, and a meta-analysis then combines their numbers. Two such analyses are worth reading closely.

In 2019 an Italian team screened 971 articles on forest bathing and cortisol, reviewed 22 and pooled 8. Salivary cortisol was lower in the forest groups than in the urban groups after the visit, by an average of 0.05 micrograms per decilitre (the confidence interval ran from 0.04 to 0.06 lower).⁴ That is the expected result. The surprise is that cortisol was also lower in the forest groups before the visit, by 0.08. A difference that exists before anyone enters the trees cannot be caused by the trees. It may reflect differences between the groups, or the effect of looking forward to a forest trip, and the authors give expectation, the placebo effect, as a likely contributor. They also found that the studies disagreed with one another a good deal: a measure of that disagreement, heterogeneity, was 88 per cent after the visit.

A 2022 meta-analysis looked at blood pressure and cortisol together. Across 21 studies and 2,270 people, forest groups averaged 3.4 mmHg lower systolic and 3.1 mmHg lower diastolic blood pressure than urban controls, with a similarly small difference in cortisol across 13 studies and 1,786 people.⁵ Those are averages, and the authors were candid about how far apart the studies were. For systolic blood pressure, the prediction interval, the range in which a new study would be expected to land, ran from a fall of about 13 mmHg to a rise of about 6. In plain terms, a new study might find a larger effect, or none. The authors also point out that these studies cannot be blinded, because people know where they have been, and that nearly all were rated at high risk of bias. They advise caution.

Mood, anxiety and the quality of the reviews #

For mental health, a 2022 review and meta-analysis by Yasuhiro Kotera and colleagues concluded that shinrin-yoku can reduce symptoms in the short term, particularly anxiety, while calling for more rigorous studies, longer follow-up in a wider range of countries, and closer study of how any effect works.⁶ That is a careful way of saying that the result looks real and that the evidence is not yet strong.

A review of reviews gives a sense of how strong. A German team collected 11 systematic reviews, covering 131 separate studies, and graded each review's reliability with a standard tool. Two reviews earned moderate confidence, six low and three critically low.⁷ Reading the better-rated reviews, the authors judged that the evidence suggests benefits for high blood pressure, stress, depression and anxiety. They judged the evidence weak for metabolic measures, for eczema in children and for social skills in schoolchildren, and said the immune findings should be verified in each region's own native forests.

Most mood measures are questionnaires, and the people answering know what they have just done. That does not make them worthless. It does mean that expectation could account for part of the effect, and the studies cannot rule that out.

Is it the forest, or the walk? #

A forest visit includes a walk, fresh air, quiet, daylight and time away from work. It is reasonable to ask how much is the trees.

One review addressed a related question: does exercise outdoors do more than the same exercise indoors? It pooled 11 studies and 833 adults, who walked or ran in a natural setting on one occasion and indoors on another. Outdoors, people reported more revitalisation, energy and positive engagement, and less tension, confusion, anger and depression. They also enjoyed it more and said they were more likely to do it again. One result cuts the other way: there was a trend towards less calmness after the outdoor sessions, though it was statistically clear in only one study.⁸

A well-known experiment in California assigned 38 healthy adults, by chance, to a 90-minute walk in a natural area or along a busy street. After the nature walk they reported less rumination, the habit of dwelling on negative thoughts about oneself, and a brain scan showed less blood flow in a region linked to it. The urban walkers showed no such change.⁹

Look at the comparison in that experiment. Most forest studies set trees against a city, and the city is often a loud one. Such studies show that a forest beats a busy road. They do not show that a forest beats a quiet park, a canal path or a calm town. This is a reading of the study designs, not a finding, but it matters. If a quiet green space does much of the same work, the barrier to trying it is much lower.

The immune claim #

The most quoted forest-bathing result concerns natural killer cells, white blood cells involved in the body's early defence against infection. Qing Li and colleagues took 12 men aged 37 to 55 from Tokyo companies on a three-day, two-night trip to forests. Eleven of the 12 showed higher natural killer cell activity afterwards, by about 50 per cent, compared with measurements taken before the trip on an ordinary working day.¹⁰ A later study compared a forest trip (12 men) with a city trip (11 men) and found the rise only after the forest.¹¹ A third had 12 men spend three nights in a city hotel while the oil of hinoki cypress was vaporised in the room. Their natural killer cell activity rose and their adrenaline fell.¹² From this, the researchers suggested that tree aromas, called phytoncides, and lower stress hormones may both play a part.

Treat this evidence with care, for several reasons. All the studies described here come from one research group; in his own 2010 review, Li noted that apart from his studies there were no published reports on the topic at that time. The groups were small and the volunteers were middle-aged men. A three-day trip differs from ordinary workdays in many ways besides trees, such as rest, no work and different surroundings. A higher count in a blood test is not the same as better protection against disease, and no study shows that. The review of reviews says the immune findings should be verified in each region's own native forests, and a Swedish university summary notes that it is not clear whether the amount of aroma breathed in a forest has the same effect as the aromas used in laboratory experiments.¹³

Time in nature across a whole population #

Another kind of evidence asks how much time in nature goes with better health across ordinary life. A study led by Mathew White analysed a national survey of 19,806 adults in England, covering parks, woodland, beaches and countryside, but not people's own gardens. Compared with people who had spent no time in nature in the previous week, those who reported 120 to 179 minutes were more likely to rate their health as good (an odds ratio of 1.59) and their wellbeing as high (1.23). Longer visits went with no further gain, and it made no difference whether the time came in one visit or several. The pattern held across many groups, including older adults and people with long-term illness.¹⁴

This is an observational study. People reported their own health and their time outdoors for one week, so it cannot show that nature caused the better health. People who feel well may simply get out more. The authors say as much, and call for follow-up and intervention studies. An odds ratio is also not a percentage chance: 1.59 means the odds were about one and a half times higher, not that the chance of good health rose by 59 per cent. The study is the source of the ‘two hours a week’ figure, which is best read as a promising signal and not as a prescription.

An old study that still carries weight #

In 1984 Roger Ulrich published a two-page paper in Science comparing 46 patients who had had gallbladder surgery at a suburban Pennsylvania hospital between 1972 and 1981. Each of 23 patients whose window looked onto a small stand of trees was matched with a patient of similar characteristics whose window faced a brick wall. The tree-view patients stayed in hospital for a shorter time, by a little under a day on average, and took fewer strong painkillers.¹⁵

It deserves its place in the story and also its limits. It was a review of old hospital records, not an experiment in which patients were assigned to rooms; it involved one hospital and one operation; and 46 is a small number. It is an early and intriguing sign, and it has been widely cited ever since.

What we still do not know #

• Whether benefits last. Most studies measure within hours or days, and few follow people for months.
• Who benefits. Volunteers have often been young or middle-aged healthy men, and the studies come mostly from Asia.
• Whether trees are special. Few studies compare a forest with other calm places, so the effect of trees is hard to separate from the effect of quiet.
• How much is enough. The two-hour figure comes from an observational study, not from a trial.
• Why it might work. Three explanations are on the table, and they may all contribute: stress reduction, attention restoration, and the aromas of trees. None has been settled.
• Whether published results are the whole picture. Reviewers cannot rule out publication bias, the tendency for positive results to appear in print more often than null ones.

Five questions for the next headline #

The next time a headline says that nature cures something, these questions will usually find the weak point quickly.

1. Who took part, and how many? A dozen young men in one country is a start, not a conclusion.
2. What was the comparison? Trees against a noisy road, a quiet park, or nothing at all? The answer changes what the result means.
3. What was measured, and when? A blood test an hour later and a questionnaire six months later are very different claims.
4. Could people tell what result was expected? If so, expectation may be doing some of the work.
5. Has anyone else found the same thing? A result from one research group is a lead. A result from several, using different methods, begins to be evidence.

What this means for a walk #

None of this is a reason to hold back. A woodland walk carries little risk, costs little and is enjoyed by many people, and it needs no evidence to justify it. What the research supports is modest: regular, unhurried time among trees or in green space seems to leave most people calmer in the short term. It supports nothing stronger than that, and it is best seen as an addition to medical care, not a replacement for it.

If you have a health condition, talk to a doctor or nurse before changing how you manage it. Beyond that, take what helps and leave the rest. If you would like to notice what a walk does for you, the journal is a simple place to keep a before-and-after record, and Build your own practice sets out how to begin. A shorter summary of this essay is on What the research says.