How Nature Benefits Mental Health

The Health Burdens of Urbanization

While most of the global population already resides in cities (an estimated 56 % in 2020), the share of urban residents is expected to reach 68 % by 2050 (United Nations, 2018). In urban areas, natural elements are sparse, and many stressors are prevalent (e. g., noise, summer heat events, spatial confinement, high population density, social isolation, road traffic, etc.). In the presence of permanent stressors, the body experiences chronic stress, which can negatively affect cognition, decision-making, anxiety, and mood (McEwen, 2017), and increase the chance of cardiovascular conditions (Bravo, & Cappola, 2021; Liu et al., 2017), depression, autoimmune, and allergic flare-ups as well as tripling or quadrupling the chances of an adverse medical outcome (Miller et al., 2007). In line with the notion that the urban environment is often detrimental to physical and mental health, a meta-analysis of urban–rural health differences (Peen et al., 2010) found the prevalence of all psychiatric disorders increased by 38 %, mood disorders increased by 39 %, and anxiety disorders increased by 21 % in urban as compared with rural areas. A narrative minireview names social disparities, social insecurity, pollution, and the lack of contact with nature as recognized factors affecting urban mental health (Ventriglio et al., 2020).

Consequently, to prevent an increase in the urbanization-related burden of disease, well-designed and just urban living conditions are essential. While the high population density in urban areas is associated with various problems, it also yields the chance to reach numerous residents over extended periods with well-designed, nature-based public health interventions. With this goal in mind, natural areas are promising as they can reduce the exposure to stressors by increasing the distance to them (e. g., a green space between residential areas and large roads) and by decreasing their perceptual salience (Nilsson, & Berglund, 2006). The current paper aims to enhance the research guiding these interventions by calling for extended interdisciplinary collaboration in the field.

Benefits for Psychological Well-Being

Several findings indicate that nature is linked to positive effects on human mental well-being (for an extended overview, see, e. g., Hartig et al., 2014). The following presents selected evidence structured by study design.

Beginning with the evidence from correlative and qualitative studies, urban dwellers with more green spaces in their living environment report better mental (and physical) health than people with little green space in their living environment (Maas et al., 2006). Accordingly, urban greenery has also been linked to the residential environmental satisfaction of urban dwellers (Bonaiuto et al., 1999). Parents of children with ADHD report their better ability to concentrate after playing in a natural environment (compared to artificial play environments; Faber Taylor et al., 2001).

Exposure to increased residential green space in childhood is associated with better behavioral and mental health outcomes, and a study among children (aged 7 – 12) indicated that more residential green space is associated with a lower risk of anxiety, depression, and somatization problems at age 12 as well as a lower risk of conduct problems at age 7 (Madzia et al., 2019). A Danish nationwide study (N > 900,000) showed that high levels of green space presence throughout childhood are linked to a lower risk of developing a diverse spectrum of psychiatric disorders later in life (Engemann et al., 2019). Moreover, exposure to and spending time in green spaces is associated with better adolescent psychological health (risk for stress, positive affect, mental health, and resilience; see review by Zhang et al., 2020). Within a qualitative study on gardening therapy, participants reported associations of the garden with safety, positive sensory experiences, and symbolism of one’s own growth (Adevi, & Mårtensson, 2013).

While the above-mentioned correlational and qualitative evidence represents a good starting point, it is no indicator of causality. A range of experimental studies has also shown benefits across various mental health measures. Lyu et al. (2019) found improved mood in 3-day bamboo forest therapy compared to an urban control group. Gidlow et al. (2016) compared participants after walking in pleasant environments of either natural or urban character and found an advantage of the natural walks for cognitive outcomes as well as higher restorative experience (with improved mood and decreased salivary cortisol in both settings). Other studies found beneficial effects of nature contact on tension, fatigue, confusion, and anxiety (Elsadek et al., 2019) or anxiety, rumination, working memory span, and affect (Bratman et al., 2015).

Experimental studies are much closer to actually assessing causal relationships than correlational designs. However, one issue that restricts the derivation of causality from such experimental studies is that the effects might be partly due to people associating natural environments with relaxation and leisure. Exploring this idea, a Swiss study (von Lindern et al., 2013) found overall negative associations of having a forest-related paid occupation with restoration. They profited less from nature contact in the sense of “being away” from everyday life. This might help add nuances to urban green space planning by combining different types of natural environments where possible (a forest ranger might find more restoration in a grassy field, a farmer might relax best in the forest) and is also a caveat for assuming causality even from experimental studies.

It is noteworthy that, while experimental study designs are closer to investigating causal relations than correlative designs, it is impossible to control all variables but one (in this case, nature contact). To assess the robustness and the size of primary study effects, we need meta-analytic results, which are presented in the following.

A meta-analysis on the effects of forest bathing (in Japanese “shinrin-yoku”; Kotera et al., 2022) summarizing 20 studies with 2,257 participants found buffering effects for depression (significant; labeled small to medium), anxiety (large and insignificant for RCT; large and significant for pre-post designs), and anger (significant; labeled medium). However, there were indications of publication bias for most results, so the true effects might be smaller.

A 2017 meta-analysis (Soga et al.) summarized various mental and physical health outcomes (e. g., anger, anxiety, body mass index, mood, quality of life) from comparisons between gardening intervention groups and control groups. Of the 76 comparisons, 4 had negative results, all of which were nonsignificant. The overall estimate standardized mean difference reached 0.42 with a 95 % CI of 0.36 to 0.48. All subgroup estimates remained positive and significant for gardeners and nongardeners, for a health vs. well-being dichotomy, and for the therapeutic vs. the nontherapeutic context.

Regarding positive affect, brief exposure to natural environments had a mean effect of r = .31 (95 % CI ranging from .24 to .37) in a meta-analysis summarizing 31 studies with 2,284 participants (McMahan, & Estes, 2015). Substantial heterogeneity was moderated by measurement scale, type of exposure, mean sample age, and study location. The same report also analyzed negative affect in 20 studies, with the mean effect estimated as r = -.12 (95 % CI from -.17 to -.07).

The results presented here provide support for the potential of nature to enhance mental health. While it might be possible that publication bias or the self-selection of researchers and participants interested in this topic (possibly scoring high on connectedness to nature) distorted the results of these studies toward the positive, we believe it unlikely that this would account for all effects found. Thus, while we acknowledge that nature is not the almighty cure for any mental health complaint, we still believe that its broad potential should be systematically harvested and understood.

Benefits for Physical Health

While our main focus is on the mental health effects of nature contact, we want to briefly touch upon the benefits for physical health. These are important because physical and mental health are tightly intertwined. The effects of natural environments on physical health are measurable in the very short term in some cases, for example, based on lower blood pressure in natural (vs. urban) environments (Hartig et al., 2003a). An early study by Ulrich (1984) found that patients recovered more quickly after surgery if their hospital room window provided a view of natural surroundings (compared to patients with a window facing an adjacent building). These patients were hospitalized for shorter periods of time after surgery, required less analgesia, and their condition was rated as more positive by nursing staff. For adolescents, contact with nature is associated with a lower risk of obesity and increased physical activity (Zhang et al., 2020). Mitchell and Popham (2008) studied the UK population in terms of income distribution and causes of death and found that health disparities between rich and poor people were significantly lower in neighborhoods with good access to green nature. Hence, natural environments might also play a role in buffering the effects of socioeconomic status on health. Wong and colleagues (2021) found that gardening therapy reduced the levels of the proinflammatory cytokine interleukin 6 and improved the number of naïve cytotoxic (CD8+) T cells, a reduction of which is associated with aging. In addition, it reduced CTLA4+, a parameter that inhibits T-cell-based immune responses, so gardening therapy might improve the hosts’ defense against viruses and cancer cells.

A meta-analysis on outdoor greenspace exposure covering 143 studies (Twohig-Bennett, & Jones, 2018) reported statistically significant reductions in diastolic blood pressure, salivary cortisol, heart rate, diabetes incidence, and mortality. However, the study quality was limited, and the study heterogeneity was substantial. Summarizing, also the outcomes of a more physical nature point toward nature contact being beneficial for human health.

Caveats for Nature-Based Interventions

While overall contact with nature is highly promising for human health, several caveats should be considered when applying nature-based interventions. In the designing of public green spaces, nature may also bear health risks such as venomous animals, allergies, and disease vectors. The appearance of these risks differs greatly around the globe and is typically underestimated in the scientific literature, which does not adequately represent biodiverse tropical regions (Keniger et al., 2013). In the global North, especially tree pollen allergens and biogenic volatile organic compounds (related to ground-level ozone) are an issue (Wolf et al., 2020). Noteworthy is that roadside vegetation might also reduce the ventilation necessary for diluting traffic-emitted pollutants (Vos et al., 2013).

An additional challenge of spatial greenery planning is social justice: As mentioned before, nature may have the potential to buffer the effects of socioeconomic disadvantage on health (Mitchell, & Popham, 2008). Nevertheless, unintended social processes (e. g., loss of affordable housing as a consequence of greening measures making the area more attractive) need to be carefully considered (Cole et al., 2017).

Summary

The lion’s share of the research reports point toward the same outcome: Nature is beneficial to human mental health. Also, the multitude of the reported outcome variables and stimuli overall, the interplay between several of these variables as well as the diversity of the methodological approaches all point toward a multifaceted nature-well-being relationship. The next section presents the most popular explanatory approaches to this relationship.

General Pathways

Serving as an umbrella theory, a group of general pathways provides an overview of the positive effects of nature on human well-being (Hartig et al., 2014). A recent approach (adapted from Beute et al., 2020) names these pathways as follows:

1. 1.
   Mitigation (e. g., natural spaces mitigate high temperatures, floods, and air pollution);
2. 2.
   Restoration (e. g., contact with nature restores resources after stress and fatigue);
3. 3.
   Instigation (e. g., natural spaces enable physical activities, for which urban environments are otherwise inappropriate (Maas et al., 2008) and facilitate social cohesion by serving as a meeting place).

Exposure to nature can thus act against stress both at the causal level (by reducing the number or intensity of urban stressors) and at the symptom level (by helping to reduce the effects of existing stressors). More concrete than this umbrella theory is previous, influential environment-psychological work that attempts to explain said effects: the biophilia hypothesis, the attention restoration theory, and the stress reduction theory, which are briefly presented below.

The Biophilia Hypothesis

The biophilia hypothesis proposed by Wilson (1986) and Kellert (Wilson, & Kellert, 1993) and states that humans hold an inherent and adaptive love of life and life-like processes. Fromm (1964) had previously used the term biophilia term to describe an affinity for nature. The hypothesis is widely used to explain nature-related behaviors, including conservation attitudes. For example, a study analyzing many photos on a social media platform found that contact to nature was associated with positively connotated activities and positive emotions (Chang et al., 2020). Despite the widespread use of this hypothesis across fields such as developmental psychology, preventive medicine, and architectural theory, one major criticism concerns the lack of explanatory depth (Joye, & De Block, 2011). It appears plausible that, because human evolution took place in natural surroundings, we are adapted to these environments, feel at ease there – and may even express a love for nature. However, the utility of the biophilia hypothesis is limited because it does not provide answers as to exactly how this happens: If we assume the observed positive effects do not come about metaphysically, then certain psychological and biological processes within the body must be responsible for them. Yet such processes do not lie within the scope of this theory. Later theoretical efforts tried to fill this gap, such as the theories presented next.

The Attention Restoration Theory

The attention restoration theory (ART; Kaplan et al., 1998) posits that natural environments are, to a greater extent than artificial environments, fascinating to us, create a sense of being away, and fit our needs. The theory distinguishes between directed and undirected attention. While the former requires cognitive resources and can be exhausted (causing cognitive performance to decline), the latter is also referred to as effortless attention because it does not require cognitive resources. According to ART, natural environments attract effortless attention because of the “fascinating” environmental elements they contain. Consequently, this frees up one’s thoughts (clearing away random thoughts). According to the theory, this circumstance leads to the restoration of directed attention, just like a muscle that can recover from strain through pausing.

Several studies demonstrate the effects proposed by ART. A meta-analysis of 31 studies examining the effect of natural environments on outcomes linked to attention restoration found positive effects for specific cognitive measures (i. e., Digit Span Forward, Digit Span Backward, and Trail Making Test B), but not others (Ohly et al., 2016). A later systematic review and meta-analysis pooled 49 outcome measures in eight cognitive domains (Stevenson et al., 2018). The domains most affected by natural environments (low to moderate effect sizes) were working memory, cognitive flexibility, and attentional control (with lower reliability).

ART works on a psychological level of analysis, i. e., it uses constructs like directed and undirected attention or attentional control. While this certainly is more specific than the biophilia hypothesis, the constructs used by ART cannot be observed directly. Instead, research must rely on experimental procedures like attention tests as proxies. The theory is increasingly subject to criticism. Most notably, it has been questioned whether the effects measured really pertain to attention and not to motivation (Joye, & Dewitte, 2018) because the widely-used experimental proxies are also sensitive to the participants’ motivation to complete these tasks. Joye and Fischer (2022) argue that work motivation would be a more elegant and more rigorous explanation to the observed effects than the depletable cognitive resource proposed by ART – which itself has been seriously questioned in recent meta-analyses (Friese et al., 2018).

In summary, while there is some evidence for ART, the theory does not afford the study of its phenomenon at the level of its underlying mechanisms – which would allow to significantly broaden the scope of potential conclusions for research and applications in environmental design.

The Stress Recovery Theory

The stress recovery theory (SRT; Ulrich, 1983; Ulrich et al., 1991) is a psycho-evolutionary theory that states that the human environment is evaluated rapidly and without major cognitive effort. As such, the environment elicits a specific affective response assumed to be anchored in evolution, i. e., evolutionarily advantageous environments are evaluated as more positive. SRT argues that many “preferenda” are present in the natural environment, i. e., preferred environmental elements that lead to a more positive evaluation. These include intermediate levels of visual complexity, spatial depth, certain structural configurations (e. g., symmetries), or the absence of threats. An environment that contains a sufficient number of such beneficial elements contributes to reducing anxiety and, ultimately, stress. According to SRT, natural environments are more likely than artificial environments to contain such preferential elements, making them more conducive to stress reduction. Numerous studies have linked contact to nature to stress-related outcomes and found the effects of nature contact on variables like fear, anger, or blood pressure (e. g., Ulrich et al., 1991). Conceptually, SRT suffers the same weakness as ART (see above) because neither of its components is directly observable.

The Perceptual Fluency Account (PFA)

In recent years, efforts have tried to reconcile SRT and ART. For example, extending on the idea of rapid and effortless processing or “soft fascination” of natural stimuli, the perceptual fluency account (PFA) explains attention restoration and stress reduction as a product of more fluent processing (Joye, & van den Berg, 2011). Its central assumption is that natural settings are processed with more ease than urban scenes. It is assumed that the human brain is attuned to the visual structure of natural stimuli, with its fractal-like, self-similar patterns. This ease of processing the natural environment is assumed to elicit both positive affect and efficient attention restoration. On the one hand, experimental studies using computer-generated and natural stimuli varying in their fractal characteristics showed connections with stated preferences and with relaxation as measured in EEG responses (Hägerhäll et al., 2004, 2015). On the other hand, Menzel and Reese (2022) found no support for the predictions of the PFA on restoration in phase-scrambled nature images. While the PFA may offer greater explanatory depth than ART and SRT because it allows to employ more objective variables, it has not yet been used in a significant number of studies.

On the Application of Theories

A systematic review of mental health benefits of nature-based recreation (Lackey et al., 2021) counted the use of 15 different theories, frameworks, and models in 51 articles. Remarkably, 59 % of the articles (30 in total) did not rely on any theoretical framework. Within the remaining 21 articles that did, the most popular theories were ART (8 studies), the biophilia hypothesis (4), the ecopsychological model (2), and the Restoration Outcomes Scale (2), which builds on both ART and SRT. The authors conclude that theory usage and further development building on existing work should be a future focus.

In light of the unrealized explanatory potential, a new generation of theories seems called for, which incorporates the biopsychological processes underlying the relationship between nature and mental health.

The Developmental Potential of the Current Theoretical Approaches

On the one hand, most of the above-mentioned findings suggest that the effects are robust and comprehensive and largely point in the same direction. On the other hand, the different study designs and results raise questions about possible differences between different manifestations of natural exposure and their health effects as well as about the (bio–)‌psychological mechanisms involved in the generation of these effects in humans. The available studies reported above primarily originate from environmental psychology and the health sciences, with most reported effects based on data from surveys, quasiexperiments, or pre-existing statistics.

To investigate the underlying mechanisms in the body, however, requires the skilled use of biopsychological on-site measures that go beyond tests and questionnaires. The following section presents research that has already used such measures.

Summary

All of these approaches operate on a psychological level, i. e., they use constructs that themselves cannot be observed. Instead, proxies must be measured, which makes empirical studies of the theories difficult. Efforts to do so are few and not always conclusive, as previously described. There has possibly not only been a publication bias toward significant results, but also against studies that contradicted the mainstream theories.

In sum, while the incumbent theories successfully guided many research undertakings, it is now time to move toward a larger framework incorporating a more biopsychological perspective. Luckily, many research undertakings already integrate biopsychological measures; a short review of reviews is provided in the following.

A Review of Reviews of Biopsychological Evidence

This section can provide only a first overview of the evidence available. Numerous studies on nature and mental health have applied biopsychological measurements. A narrative review on the biopsychological effects of outdoor nature (Haluza et al., 2014) identified 17 papers reporting 20 different measures of brain activity, the cardiovascular system, the endocrine system, and immune function. Mixed results were found for prefrontal cerebral activity and cardiovascular activity. Most studies covered by the review focused on the endocrine system with singular significant results for blood glucose, dehydroepiandrosterone, dopamine, glycated hemoglobin A1c, and noradrenaline. There were mixed results for adrenaline, cortisol, and salivary amylase activity. Regarding immune function, significant singular results were found for CD3+ cells, granulysin, natural killer cells, and perforin. Effects on immunoglobulin A and white blood cell counts were insignificant.

However, many studies relied on small, predominantly male samples and cross-sectional designs. A review based on 52 Japanese studies on the biopsychological effects of nature therapy (Song et al., 2016) found promising evidence across the same body systems. The main study limitations were short intervention durations and an overrepresentation of young and healthy participants. A narrative review (Antonelli et al., 2020) of forest volatile organic compounds (VOC; e. g., pinene, limonene) described a multitude of biopsychological pathways invoking mostly positive effects on, for example, inflammation, nausea, anxiety, or pain perception. However, much of the evidence was preclinical and limited to mice studies.

A recent review covering 33 studies on nature and immune function (Andersen et al., 2021) found positive effects on antiinflammatory, antiallergic, and antiasthmatic parameters alongside increased natural killer cell activity. However, some study designs lacked proper baseline measurements or control groups. The authors proposed the expression of proinflammatory molecules, the infiltration of leukocytes, and the release of cytotoxic mediators (perforin, granzyme A / B, and granulysin) as baseline outcomes for future studies.

In the last years, multiple reviews have been performed on the psychophysiological backgrounds of stress and nature (Corazon et al., 2019; Jones et al., 2021; Mygind et al., 2021). A broad review identified 18 studies on greenspace interventions and cortisol while calling for more rigorous research (Jones et al., 2021). A systematic review and meta-analysis covering 26 peer-reviewed studies (Mygind et al., 2021) found only limited support for a causal relationship of public greenspace on psychophysiological stress response measures (mostly serum and salivary cortisol, salivary amylase, and heart rate-related indicators). Once more, methodological quality was criticized, i. e., small sample sizes within the studies and the considerable likelihood of small-study publication bias across the studies. Corazon and colleagues (2019) also found rather weak and inconclusive biopsychological results of stress recovery outdoors.

This short review of reviews shows that research is already investigating the nature-mental health relationship with the help of biopsychological measurements, while study quality is criticized relatively often. The knowledge gained from such studies is partially inconclusive and has not yet been properly integrated into a more comprehensive theoretical framework. To summarize, there is still a lot to uncover regarding the complex interplay of biopsychological processes.

Integrating Biopsychology and Environmental Psychology

How might the integration of biopsychological parameters lead to an increased understanding? First of all, the nature-health-relationship may contain both physiological processes and (physiologically unobservable) psychological processes (see, e. g., Hartig et al., 2003b). To separate these processes from each other, the best starting point might be exploring as much as possible the physiologically observable processes before declaring any others as unexplainable by bodily processes.

Second, it has become evident that multiple reviews clearly criticized issues of study quality (see, e. g., Corazon et al., 2019; Jones et al., 2021; Mygind et al., 2021; Twohig-Bennett, & Jones, 2018). Thus, the correct and rigorous usage of said parameters decreases the risk of biased results and detects whether previous inconsistencies of different measures (e. g., an insignificant overlap between a cortisol curve and questionnaire results) were partially brought about by said biases. Besides, more results with small or at least unrelated biases could fuel valuable future reviews and meta-analyses with increased certainty.

Third, some measurements (e. g., pulse-related measures) permit continuing during the intervention with low-invasive techniques. This has two specific advantages: (a) Individuals may be influenced less by the measurement process itself, and (b) it allows procurement of a real-time curve rather than relying on singular data points. This means, for example, if individuals report feeling relaxed after a short walk through the park, an alternate measure is available for the magnitude and the onset time of said process of relaxation – given correct assumptions of how the questionnaire relates to the biomarker.

Certainly, there are substantial reasons why these gaps still exist: While questionnaires usually pose little or no burden on the research budget, the equipment to measure biomarkers ranks among the highest nonpersonnel project expenditures. Furthermore, besides requiring extended knowledge of the procedures, from machine –handling to data analysis, reliable measurements depend on a multitude of factors such as external temperature, humidity, the participants’ specific food, beverage, and medication intake, all of which decreases the participant database and the suitable time slots for data recording (see, e. g., Strahler et al., 2017 for confounders of the measurement of salivary cortisol and alpha-amylase).

While acknowledging barriers to the widespread use of biopsychological measures, we believe that the untapped research potential as outlined above deserves to be harvested.

In the future, we envision an integrated, complex model incorporating (1) the different aspects of nature contact, (2) the internal biopsychological mechanisms, and (3) the various outcomes. Such a model does not yet exist, and not enough is known to formulate it in a top–down fashion. Instead, we would like to encourage the community to build it together. By focusing on different potential biopsychological pathways in different empirical studies, we can create a more complete picture: The integrated model (or versions thereof) emerge by connecting the knowledge from the different studies at their intersections – i. e., at the variables or mechanisms they share.

This approach requires considerably more effort than the continued use of the current psychological theories. It cannot be presumed that, within the complexity of the human mental and bodily processes, only one mechanism (or a few mechanisms) can be identified to explain all effects. Instead, any number of mechanisms might interact in a complex fashion.

Altogether, although the level of complexity of such an approach will initially be higher, it offers fundamental advantages: It allows for less speculative studies, where more deterministic effects can be observed more easily, using more straightforward and more objective operationalizations. In the end, it can be expected that it will be less challenging to communicate such studies to the public, and that it will be easier for political decision-makers to act upon them because they are based on “hard” facts.