Such analyses provide greater knowledge of how natural environment interventions impact performance and by what mechanisms, which have theoretical and practical implications.

Findings from other studies assessing the effects of nature vs. control conditions on BDS performance and other objective measures of directed attention performance were also reviewed and integrated with the findings from the pooled data and analyses of the multiple experiments reported in this paper.

In summary, the specific purposes of this paper were to: 1 evaluate the effects of nature vs. urban exposures, and 4 compare the effects between studies using different exposure types. urban environment exposures were tested on directed attention in controlled experimental settings to ensure the validity of the cognitive performance measures.

All studies employed a randomized crossover trial RCrT or a randomized controlled trial RCT design and employed the same cognitive test to evaluate effects on directed attention—the BDS task. A summary of the studies and sample characteristics are provided in Table 1. For studies which have been published previously or which have been submitted for publication, the reference is included in the column of study names.

For more details on the study samples that have been published previously, see the original research articles Berman et al. All studies measured cognitive performance with the BDS, and affective state except one study: Bourrier et al.

Participants were randomly assigned to the experimental conditions, and the order of conditions was counterbalanced for within-participant designs. A summary of study characteristics, including the type and duration of environmental exposures and study design, are provided in Table 1.

For new and previously unpublished studies, additional descriptions of procedures and methods are provided in Appendix A in Supplementary Material.

For more details on the methods in studies that have been published previously, please see the original research articles Berman et al. Participants in each study were tested on the backwards digit span task, in which participants are presented with number sequences which they were required to repeat in backwards order.

Fourteen to eighteen depending on study separate number sequences trials were presented. Number sequences were a minimum of three digits in length and with a maximum of 9—11 digits depending on the study , with two trials at each digit string length.

This is a general performance measure that taps both working memory capacity and directed attention, since inconsistent performance is also reflected in this measure.

See Table 1 for information on test administration in the respective studies. The BDS task was used in the studies because this task places high demands on executive attentional control processes including encoding, maintenance, manipulation and updating of to-be-remembered items , and would therefore be sensitive to changes in this cognitive capacity caused by different environmental interactions.

Furthermore, the task has been used consistently across the different studies in order to be able to pool as well as compare results across studies. The Positive Affect and Negative Affect Schedule, PANAS Watson et al. Affect-related adjectives e. There were 10 positive and 10 negative affect items.

For each respective study sample, only participants with complete data for the BDS were included in the analyses. Statistical analyses were computed using SPSS Data sets were screened for univariate ± 3. Analysis of variance ANOVA models were computed in the general linear models procedure in SPSS to evaluate the effects of environmental condition nature vs.

urban across time from pre- to post environment exposure on BDS performance and affective state across the different studies. randomized controlled studies with environment as a between-subjects factor were analyzed in separate models—modeling the environmental factor as a within-subjects vs.

a between-subjects factor as appropriate, and the time point pre- vs. post environment interaction as a within-subjects factor. For the study samples in which environment was a within-subjects factor, the order of environmental conditions was included as a between-subjects factor in an additional model to evaluate the main and interaction effects of the order of environmental condition i.

The Linear Mixed effects Model LMM procedure in SPSS was utilized to analyze the same factorial models as described above, while also adjusting for affective state at each measurement point. Unlike repeated measures ANOVAs which cannot handle time-varying covariates , in the LMM models we included affect measured pre- and post each environment exposure as a time-varying covariate in the factorial model.

These models were computed in order to evaluate if the effects of environmental condition on cognitive performance were independent of or mediated by i. For studies with a randomized crossover design, the fixed factors time pre- vs. post environment exposure and environment nature vs.

urban were specified as repeated-measures fixed factors using unstructured covariance as the covariance structure, IBM, , and the order factor i.

the urban condition first was specified as a between-subjects factor. The full factorial models included the factors: time, environment, and order as the independent variables with BDS scores as the dependent variable and with affect scores included as time-varying covariates.

These models were computed using maximum likelihood model estimation. For data analyses where environment was tested as a between-subjects factor, only time pre- vs. post-environment interaction was specified as a repeated measure. Further descriptions and syntax for these analyses are provided in Appendix A in Supplementary Material.

Since the picture dose-response study did not test the contrast between nature and urban conditions, and only tested the effects of nature picture viewing in 2 sessions, this sample was not included in the pooled data-analysis models.

Results for this study sample are presented individually for comparison purposes only see Tables 1 , 2. Table 2. Mean BDS scores by study, time pre- vs. post environment exposure , environment condition, and order test session. To compare the findings from our own studies with those of others, a review of the literature and summary of findings was performed for existing randomized controlled or randomized crossover studies that tested the effects of nature vs.

urban or control exposures on BDS performance, as well as on other executive cognitive performance tests. As such, only randomized controlled studies or randomized controlled crossover studies measuring cognitive performance both pre- and post-environment exposure were included.

Study populations included students undergraduate and graduate students and non-student adults. Since the data-analyses and results presented in this paper concern effects on BDS, the focus of the literature review was also primarily on findings for BDS.

In addition, study findings utilizing other cognitive tests were also reviewed and were summarized in terms of the test statistics reported in those papers to provide an overview for comparison to the BDS effects.

The literature search was performed using multiple search engines for academic journal articles and reports, including Google Scholar, PubMed, Scopus, and PsychInfo. Searches were concluded by the end of March Reference sections of the obtained papers were examined for additional studies, and a descendancy search was also conducted for studies that cited the obtained papers.

The review of the literature may however not be exhaustive. Mean BDS scores by time, environment condition, and order, for each study sample, are shown in Table 2 , and changes in BDS performance scores after nature vs.

Furthermore, aggregated BDS scores for the studies with environment as a within-subjects factor are shown by time, environment, and order in Figure 2.

Paired samples t-statistics for BDS change from pre- to post nature vs. urban environment interactions in the 1st vs. the 2nd test sessions, F -statistics for time pre, post by environment nature, urban interactions, and time by environment by order interactions, are shown in Table 3.

urban environment conditions, separated by whether the environment condition occurred in the 1st vs. the 2nd test sessions, are also shown in Table 3. Figure 1. A—D Mean BDS score changes pre- to post- nature vs. urban environment interactions, by order of environment conditions, for each study sample.

A BDS change after nature, 1st sessions; B BDS change after nature, 2nd sessions; C BDS change after urban exposure, 1st sessions; D BDS change after urban exposures, 2nd sessions. See results in Table 3. Figure 2. A,B BDS performance scores pre- and post-nature vs.

urban environment interactions, by order of environment conditions. Aggregated results for study samples with environment as a within-subjects factor. Mean BDS scores by time and environment condition, for the order of conditions nature 1st and urban 2nd; A and urban 1st and nature 2nd B.

Table 3. Effects on BDS performance from pre- to post environment exposures by study sample, and aggregated time pre- vs. post environment by environment conditions nature vs. urban by order effects on BDS across study samples. Table 4. Results from ANOVAs on time pre, post × environment nature, urban effects A , and time × environment × order nature 1st, urban 1st effects B , on BDS, for studies with environment as a within-subjects factor.

BDS performance scores pre- and post-nature vs. urban environment interactions, by order of environment conditions, are shown in Figure 2. See Tables 2 , 3 , and Figures 1A—D , for a summary of these effects. In sum, in the studies with a RCrT design i. There was also a strong time × environment × order interaction, whereby the pattern of performance changes after nature vs.

urban interactions differed depending on the order in which participants experienced the environmental conditions.

Specifically, BDS performance improved only after nature but not urban interactions if tested in the second session after initial practice effects occurred in the 1st test sessions.

Descriptive statistics for positive affect PA and negative affect NA are shown in Table 5. Table 5. Descriptive statistics for Positive and Negative Affect, for study samples with environment as a Within-subjects factor, by the order in which the environment conditions were administered.

Table 6. Results from ANOVAs on time pre, post × environment nature, urban × order nature 1st, urban 1st effects on positive affect and negative affect, for study samples with environment as a Within-subjects factor. Since there was a clear time × environment effect, and a clear time × environment × order effect on both BDS and PA, we wanted to examine if these interaction effects on BDS performance were preserved after adjusting for PA by including pre- and post-environment affect as a time-varying covariate in the factorial model, computed using the Linear Mixed Models procedure in SPSS.

Supplementary Material regarding these analyses and results are given in Appendix A in Supplementary Material , section 2. Table 7. Negative Affect, for study samples with environment as a within-subjects factor. In sum, for these studies with environment as a within-subjects factor, there were interaction effects of time × environment and time × environment × order on PA but not on NA.

PA increased after nature interactions but not after urban interactions, but this effect was restricted to the second sessions. Importantly, the time × environment and time × environment × order effects on BDS were hardly changed after adjusting for either PA or NA. An ANOVA was used to test the effects of nature vs.

urban environment interactions in the studies implementing a RCT design i. Table 8. Results from ANOVAs on time pre, post × environment nature, urban effects on BDS, for studies with environment as a between-subjects factor.

Descriptive statistics for PA and NA are shown in Table 9. The analyses of PA and NA exclude the Video study from UBC Bourrier et al.

Table 9. Descriptive statistics for Positive and Negative Affect, for study samples with environment as a between-subjects factor.

Table Results from ANOVAs on time pre, post × environment nature, urban effects on Positive vs. Negative Affect, for study samples with environment as a between-subjects factor.

Since the analyses on affect excluded the video study from UBC as this study did not measure affect , the time × environment effect on BDS was also computed after excluding this study, to compare these results with those adjusted for affect.

Therefore, additional models testing time × environment effects on BDS after adjustment for PA and NA were not performed separately for this subset of studies with an RCT design. In sum, in the studies with an RCT design i.

Unlike in the within-subject designs, there was no significant time × environment effect on PA, nor on NA. Since the effects of the order in which environment conditions were administered were significant in predicting BDS performance changes in the studies implementing a RCrT design, we also analyzed the environment × time effects on BDS separately for all first and second sessions.

That is, these analyses included data from the studies which tested environment as a within- or between-subjects factor, comparing the effects of nature vs. urban conditions on BDS in all first sessions, vs. in all second sessions see Table In order to compare the effects on BDS with and without adjustment for affect, the linear mixed effects model procedure was used in SPSS to compute the factorial models fixed effects models with environment as a between-subjects factor and time pre- vs.

post exposure as a within-subjects factor. Maximum likelihood estimation was used and unstructured covariance type for the repeated factor. Again, these results illustrate the importance of delineating the effects of practice primarily occurring in first sessions, as evidenced by the large main effect of time across the environment conditions in the first but not in the second sessions in order to adequately evaluate the effects of the environment exposures in the second sessions on cognitive performance.

The significant time by environment interaction effect in the second sessions did not change much after adjusting for PA, F 1, In line with this reasoning, BDS changes and affect changes were not correlated in any other environment × session order cell, except in the nature condition tested in the 2nd session.

There was no observed correlation between BDS change and NA change. To further evaluate any potential mediation of BDS changes through PA changes, after nature interactions in the second test sessions, supplementary mediational path analyses were also performed see Appendix C in Supplementary Material.

The results of the path analyses are presented in Diagram 1. E-F and in Appendix C in Supplementary Material , and showed that out of the total improvement of 0. These results further indicate that the positive effect of nature interactions i.

Diagram 1. A—D Illustration of general effects and potential mechanisms of action on cognitive performance and affect from pre- to post nature vs. urban environment interactions, and the order of participation in the environment conditions.

A Nature environment interaction, Nature condition is tested in session 1; B Nature environment interaction, Nature condition is tested in session 2; C Urban environment interaction, Urban condition is tested in session 2; D Urban environment interaction, Urban condition is tested in session 1.

Depending on the order in which individuals participate in the nature and urban conditions, the changes observed on cognitive performance and affect also differ. In first test sessions A,D , there are clear practice effects on cognitive performance in addition to the effects that are caused by the environment condition per se.

Second test sessions B,C are instead devoid of the initial performance improvement due to practice, and the cognitive performance improvements observed constitute more clean effects of the environment conditions.

Thus, the second sessions provide a better evaluation of the effects of the environment conditions per se on cognitive performance. Changes in affect also differ depending on the order of environment conditions. That is, positive affect increased after nature interactions performed in the second but not first sessions, which could be due to different expectations on the second session, depending on the experience environment condition in the prior, first session.

That is, if the urban condition was done first, the expectations on the second condition may be low and the actual experience after the environment interaction in session 2 may have exceeded expectations.

urban environment interactions, in 2nd test sessions. E Nature condition, 2nd test sessions: Total effect, c: 0. Studies from outside of our laboratories and our collaborators laboratories were reviewed that employed a RCT or RCrT design to test the effects of nature exposures, compared to urban or other control conditions, on BDS performance or other executive cognitive performance tasks evaluated pre- and post the environment exposure conditions.

Twenty-seven studies were identified in the literature which met the inclusion criteria and the characteristics and results of these are summarized in Supplementary Table 12 in Appendix B. Out of these 27 studies, 12 studies included BDS measurement pre- and post the environmental exposure conditions.

Mean BDS scores pre- and post each environmental exposure for different subgroups, where applicable are reported in Supplementary Table 13 in Appendix B if reported in the original reports.

Five of these 12 studies reported a statistically significant effect of environment condition on BDS performance in favor of the nature condition Cimprich and Ronis, ; Lin et al.

The other 7 studies reported no favorable effects of the nature conditions on BDS performance Bodin and Hartig, ; Stark, ; Perkins et al. The results for environment effects on BDS in the 12 different studies identified in the literature review were heterogeneous.

Importantly, these studies did not parcel out practice effects from environment effects. The varying findings from these studies could thus stem from practice effects contaminating the effects of the environmental condition to varying degrees in the different studies.

This is especially true for those studies employing an RCT design such that all test sessions are likely contaminated by practice effects. For example, Bratman et al. However, these studies also both employed a RCT design, which means that all test sessions were first sessions and were thus very likely to be confounded by practice effects.

Gable et al. reported practice effects with a mean BDS improvement of 0. To compare with the results across the studies from Berman and colleagues, the difference between the nature and urban conditions regarding changes in BDS performance were clearly smaller in the first sessions than in the second sessions.

In the first sessions, mean changes in BDS performance were 0. The mean changes in BDS performance in the second sessions, devoid of initial practice effects, were instead 0. As such, the effects of nature compared to urban interactions on BDS performance improvements could be underestimated, and the statistical power reduced, in those studies due to such methodological limitations.

Regarding the effects of nature exposures on other executive cognitive tests, 17 different cognitive tests of executive and attentional processes some including multiple performance measures were used in addition to the forward and backward digit span, across the 27 studies that were identified in the literature 28 studies when including Berman et al.

These studies included the following tests, with the number of studies using the test indicated within brackets : ANT with different test components assessing executive 4 , orienting 3 or alerting attention 2 , Necker cube pattern control NCPC 6 , Sustained attention to response task SART 3 , Trail making test A TMTA 2 and B TMTB 3 , Operation span Ospan 1 , Symbol digit modalities test 1 , Change detection 1 , Logical memory 1 , Error scale 1 , Category matching 1 , Colored number pictures 1 , Reading span task RST 2 , Stroop 1 , Search and Memory test 1 , Proofreading task 1 , and Symbol substitution test 1.

These include the single study testing Ospan Bratman et al. The latter found greater performance improvements after viewing nature pictures compared to day-time city pictures but not compared to viewing nightscape city pictures. The diverse findings on NCPC and ANT-E may also suggest that these tests do not optimally capture the effects that nature exposures might have on directed attention processes possibly due to not being demanding enough on directed-attention processes, or having weaker reliability.

However, null findings may also be related to study designs rather than the cognitive tasks employed, such as limited effectiveness and contrast between the environment conditions tested, limited sample sizes and power, varied testing procedures, and confounding effects of test practice and order effects in studies employing a crossover design with the environment conditions being a within-subjects factor.

This design choice could abolish time by environment effects on some tests like ANT-E and BDS due to fatigue affecting performance on different tests for different participants both pre- and post the environment exposures and thereby confounding the assessment of cognitive performance changes invoked by the environment exposures.

The significant effect on Ospan despite this test randomization procedure and long testing session, observed in the same study Bratman et al. In fact, Ospan is also the task that is most similar to the BDS tasks in that both are demanding WM tasks that heavily tap several cognitive control processes dependent on executive attention.

Based on the findings of both the studies reported in this paper and other studies identified in the literature, such WM tasks seem to capture at least some of the effects that different environment exposures have on cognition, which was also observed by Stevenson et al.

In this paper we analyzed the effects of nature vs. urban environment interactions on directed attention, measured by the BDS task, across 12 experimental studies.

Furthermore, we assessed how affect and the order of environmental conditions influenced the effects of environment on cognitive performance. Across the studies presented and analyzed in this paper, cognitive performance was found to improve significantly more after nature interactions compared to urban interactions.

Overall, BDS performance improved on average by 0. However, we found a strong interaction between time pre- vs. post environment exposure , environment type nature vs. urban , and the session order of environmental conditions for studies with environment as a within-participants factor.

Specifically, cognitive improvements were generally larger in first sessions than second sessions across different environmental conditions e. On the other hand, in the second testing sessions without the initial practice effects , the effects of the different environment conditions on cognitive performance were magnified, whereby performance continued to improve after the nature interactions by 0.

These results help to clarify which effects are attributable to environmental conditions and demonstrate that initial practice effects must be parceled out to isolate cognitive effects from environmental interactions that are not confounded by practice effects see Diagram 1.

While these results show that the nature interactions tended to improve performance significantly, they also show that urban interactions may even cause declines in directed attention performance in some cases, as indicated by the decline in performance after interacting with urban environments in second testing sessions.

With regards to affect, although nature interactions generally have a positive effect on affect when compared to urban interactions, the differential effects of environmental exposure on directed attention as measured by BDS could not be explained by changes in affect.

Specifically, in the nature condition, devoid of practice effects on the BDS task, changes in positive affect could only explain 2. However, it is also possible that limitations in the measurement of affect could prevent the detection of existing covariation in affect and cognitive performance changes resulting from environmental influences.

Such limitations could be the result of the nature of affect rating scales with possible floor vs. ceiling effects for negative vs. The boost in cognitive performance following nature interactions may be caused by a replenishment of cognitive resources mediated by a rest of executive cognitive-control processes, as suggested by ART.

The characteristic perceptual features of natural stimuli, in terms of containing statistical fractal patterns, may contribute to this effect, since such patterns have been found to induce brain signals related to a wakefully relaxed state Hägerhäll et al. Furthermore, the soft fascination that comes from perceiving natural stimuli may also facilitate a form of restful distraction from, and reduction in, other effortful cognitive processes that compete for attentional resources and which can thereby have a negative impact on executive cognitive performance.

Examples of these phenomena are the negative effects of proactive interference whereby recently activated items in working memory interfere with current to-be-remembered items and task goals , and ruminative thinking, which similarly can interfere with cognitive task performance Brinker et al.

Evidence in support of the hypothesis that interactions with nature may facilitate a distraction from other cognitions, and as such a stilling of the mind, were reported by Bratman et al.

The nature-related performance boost could also stem from an increased motivational state, modulating the deployment of neural resources and functioning of neural networks involved in executive cognitive-control processes. It has also been suggested that natural environments and stimuli tend to have a positive effect on well-being because these are the environments in which humans evolved, and they contain the resources that enable human survival e.

According to this reasoning, environments that signal that resources for survival are in abundance such as vegetation, water, raw material, shelter, biodiversity , while threats to survival are absent or low, should have the most positive effects and be the most preferred, for which there is some support McMahan and Estes, ; Wyles et al.

This way, environmental stimuli are plausibly linked to the up- and down-regulation of motivational systems McMahan and Estes, which signal reward and modulate approach behaviors. These systems include e. Importantly, these motivational systems are intimately related to and subserve the regulation of executive cognitive-control processes and associated neural networks, as well as the experience of effort and cognitive fatigue Botvinick and Braver, If one of the underlying effects of natural stimuli is to induce a motivational boost, this would also be a plausible explanation for the positive effects that can occur from very brief exposures to natural stimuli, such as viewing images of nature for a short duration.

Future research should investigate such possible underlying mechanisms of the observed behavioral effects of interactions with nature. A summary of the observed effects of environment exposure and test order on cognitive performance changes, and the potential mechanisms discussed above, are illustrated in Diagram 1.

Generally, among all the studies by Berman et al. presented and analyzed in this paper, the effects of virtual reality nature and picture exposures were smaller, compared to walks in real nature as well as nature exposure via a video.

It is noteworthy that the walks in real environments at the University of Chicago, which were shorter and less contrasted in terms of nature and urban built elements than those at the University of Michigan, showed smaller effects. This is understandable and suggests that the extent of naturalness vs.

urban-ness plays a role for the positive effects on cognitive performance that are gained. This is in line with other recent research findings of dose-response effects on stress recovery from experimentally controlled environments with different amounts of natural elements Jiang et al.

However, not all data are consistent on this point. In a different healthy sample screened to be free from mental illness , the effects of the same nature walk on BDS performance was not superior compared to the urban walk at the University of Michigan.

Moreover, participants showed significant improvements across both nature and urban conditions in both the first and second sessions. It is possible that this sample consisted of individuals who were more robust to fatigue effects, were less affected by the environment, and had a high cognitive learning capacity.

At the other end of the fatigue spectrum, the results for the participant sample who were clinically diagnosed with major depression exhibited the largest effect of the nature walk compared to the urban walk at University of Michigan.

These effects were due to clear improvements in the nature condition, while notable deterioration in BDS task performance occurred after the urban condition, which is especially evident in the second sessions where the initial practice effects were not present.

The present paper reported the results of all experiments conducted by Berman and colleagues, which met the inclusion criteria, with the purposes of analyzing and delineating the effects of session order as well as affect data which are not available for other studies from the effects of environmental exposures on BDS.

A strength of this paper is the large amount of data that is analyzed, and the absence of publication bias for the individual experimental studies non-significant results are less prevalent in published samples, but are included here. Another strength of this multiple-experiments study was the use of the same cognitive test measure across all studies, allowing for both comparability between studies and a more robust analysis across different studies of the effects of nature vs.

control exposures on executive cognitive performance as measured by the BDS task. When reviewing the results of other studies in the literature, the qualitative, and quantitative summaries were limited to the information presented in those study reports.

Incomplete reporting of descriptive statistics for each experimental condition and dependent variables, in some studies, limited the review of and comparisons to other's findings in studies testing the effects of nature and control exposures on BDS performance.

Pooled data-analyses were performed on a total of participants across 12 studies with different types of interactions with nature vs.

urban environments. Significant environment × time interactions were found whereby BDS performance improved more after nature compared to urban environmental interactions.

Importantly, this effect was magnified after parceling out initial practice effects on the BDS task. In this case, BDS performance instead declined after urban environment interactions in some studies, indicating a fatigue effect, while BDS performance continued to improve after nature interactions.

Furthermore, the cognitive performance improvements after nature interactions were found to be largely independent of changes in positive and negative affect. These results suggest that the mechanisms through which nature interactions alter cognitive performance vs.

affect may differ and be independent. Other studies in the literature examining the effects of nature vs. urban or other control environment interactions on BDS performance showed mixed results where some found clear practice effects that could have overshadowed the detection of any environmental effect.

Effects of order and practice should thus be handled carefully in future studies to obtain more accurate estimates of the effects that different environmental interactions have on cognitive performance.

All studies were approved by and carried out in accordance with the recommendations of the Institutional Review Boards for psychology at the University of Chicago, the University of Michigan, and the University of British Columbia. All subjects gave written informed consent in accordance with the Declaration of Helsinki.

Conception and design of the present paper: CS and MB. Data-base construction, data-analyses, and writing of the manuscript: CS.

Conception, design, and data collection in individual experimental studies: MB, SV, KES, FM, KELS, GN, SB, JE, OK, and JJ. Critical review of the manuscript: MB, SV, KES, FM, KELS, GN, SB, JE, OK, JJ, and CS. This work was supported in part by a grant from the TKF Foundation to MB, two grants from the John Templeton Foundation the University of Chicago Center for Practical Wisdom and the Virtue, Happiness, and Meaning of Life Scholars Group, an NSERC Canada Discovery Grant to JE, grants from the National Science Foundation BCS to MB and to KES DGE , as well as an internal grant from the University of Chicago to MB.

Given the continued growth in urbanization and a move to an indoor lifestyle, our results highlight the importance of spending time in nature, especially when exercising. All participants gave their informed written consent, approved by the Human Research Ethics Board at the University of Victoria HREB: BC Human Research Ethics Board approved all experimental protocols at the University of Victoria HREB: BC The experiment conformed to the ethical standards prescribed by the Declaration of Helsinki and subsequent revisions.

All participants were given a comprehensive set of instructions regarding the procedure and tests, agreed both verbally and in the consent form to the testing procedures, and were given course credit in exchange for their participation.

Previous work in our laboratory 43 conducted an ERP experiment with a sample size of and found that detecting an ERP elicits a large effect size of 0. We conducted a power analysis for a repeated measures t test using this standardized effect size, an alpha of 0. Moreover, our laboratory follows a protocol wherein ERP studies include a minimum of 30 participants, corresponding to a power of 0.

To avoid conducting underpowered research 45 , we kept testing participants until we had achieved our a priori set size of Further, each participant was instructed to abstain from eating or consuming caffeine two hours before testing.

All participants had normal or corrected-to-normal vision and no known neurological impairments. Participants completed a standard visual oddball task on an Apple iPad Apple Inc. At the same time, EEG data were recorded from a Muse EEG system Interaxon Inc.

The task and recordings were collected prior to and after min indoor and outdoor walks. The first participant walked inside on day one and outside on day 2. Subsequent participants alternated the walk locations on a participant-to-participant basis.

The study was held at one of three times: am, 12 pm or pm—in an effort to limit the effects of the daily circadian rhythm. Here, we choose the oddball task, rather than the previously used Stroop task 19 , 20 , to expand the current scientific literature on this topic.

Both tasks are highly well-known in cognitive neuroscience, and both have been previously used to measure selective attention capacity and working memory 1 , 2 , 3 , 4 , 19 , 36 , Therefore finding the same results with a different yet equally valid task would only provide additional power to the result.

Stimulus order was randomized, with the constraint that the stimulus presentation software ensured that no more than two infrequent oddball circles appeared consecutively. Participants were instructed to quickly press the bottom left or right of the iPad screen when they saw one of the blue circles oddball and not respond when they saw the green circles control.

The circles were presented for — ms, and the trial ended automatically on oddball trials in which the participants did not respond.

Each trial began with a fixation cross as soon as the previous circle disappeared. Participants completed four blocks of trials. After completing the pre-walk oddball task, participants completed a brief walk inside the Engineering Lab Building or outside the Alumni Chip Trail at the University of Victoria.

This trail was a green and lush forested path around the campus. For both locations, the distance of the walk was 2 km. Participants were asked to walk at their normal pace—above leisurely but not hard breathing.

In addition, participants were asked not to speak to anyone, use their cell phones or listen to music for the entirety of the walk. A research assistant timed each walk and walked approximately 10 m behind the participant, pacing them to ensure they maintained the same walking intensity.

Aside from the experimenter's directional instructions to ensure the participant kept a set pace, there was no conversation between the experimenter and the participant.

After completing the walk, participants completed the oddball task again. All EEG recording was done in a quiet room within the Engineering Lab Wing building. com for full technical specifications, see Fig.

The Muse EEG system has electrodes located analogously to FPz, AF7, AF8, TP9, and TP10, with FPz utilized as the reference electrode during recording. Note that the PEER application does not send event markers to the EEG headset as per a traditional ERP study 46 but instead reads the EEG data with known Bluetooth lag and jitter.

Specifically, we have written MATLAB code that allows two-way communication via the OSC protocol to send markers to mark a continuous EEG recording.

Due to the Bluetooth lag, the EEG samples corresponding to this point in time did not arrive for 18—20 ms on average with a jitter of approximately 5 ms 47 , It is important to note that this jitter only impacted the initial signal locking between the MUSE system and our software and did not vary over time.

In addition, the random delays in the temporal onset of our data collection would have a Gaussian distribution, and thus the lags would average out. The Muse EEG system was made by InterAxon Inc with electrodes labelled AF7, AF8, TP9, and TP Reference electrode is labelled at FPz.

Simply put, this means that all data points are guaranteed to be in the same order they were continuously collected in. As such, the averaging of temporal onset did not impact the present data as much as one might assume.

Thus, the signal did not differ from trial to trial but from participant to participant. Signal quality was then inferred by examining the variance per second on each EEG channel, and data collection began when all channels had a variance per second less than Data were processed offline in MATLAB using EEGLAB 49 and custom code.

We did not re-reference the continuous EEG data offline as our ERP analysis focused on the two posterior Muse electrodes TP9 and TP10 referenced when recording electrode FPz. Continuous EEG data were filtered with a dual-pass Butterworth filter with a 0.

A preliminary analysis of the data revealed no lateralized effects; further, we wanted to improve the signal-to-noise ratio of the ERP measures 50 , so we created a pooled frontal and a pooled posterior virtual electrode by averaging across the frontal AF7 and AF8 and the rear TP9 and TP10 electrodes, respectively.

Our ERP analysis only focused on the new average posterior virtual electrode based on our previous work 47 , html for exploratory analyses examining this issue that provided the rationale for the choices we made here. After filtering, epochs of data from ms before to ms after stimulus onset oddball, control were extracted from the continuous EEG data.

Segments were then baseline corrected using the ms preceding stimulus onset. Segments were then averaged for each participant's oddball and control trials, and a difference waveform was constructed by subtracting the average control from the average oddball ERP waveform.

N and P ERP component amplitudes and latencies were quantified at the participant level by finding the local minimal N — ms and local maximal P — ms voltage amplitudes within the windows mentioned above around the grand average component peaks.

ERP peak amplitude data were statistically analyzed using a two walk location: indoor, outdoor by two time: pre-test, post-test fully repeated measures analysis of variance.

Post-hoc decomposition of the interaction was done via dependent samples t tests. Reaction time was calculated as the time it took participants to press the screen after the stimulus circle was presented.

Accuracy was defined by the number of errors made during the task responses to control stimulus or no response to oddball stimulus. Basso, J. The effects of acute exercise on mood, cognition, neurophysiology, and neurochemical pathways: A review.

Brain Plasticity 2 , — Article Google Scholar. Brisswalter, J. Effects of acute physical exercise characteristics on cognitive performance. Sports Med.

Chang, Y. The effects of acute exercise on cognitive performance: A meta-analysis. Brain Res. Article CAS Google Scholar. Antecedent acute cycling exercise affects attention control: An ERP study using attention network test.

Hillman, C. Be smart, exercise your heart: Exercise effects on brain and cognition. Lambourne, K. The effect of exercise-induced arousal on cognitive task performance: A meta-regression analysis. Ludyga, S. Acute effects of moderate aerobic exercise on specific aspects of executive function in different age and fitness groups: A meta-analysis.

Psychophysiology 53 , — McMorris, T. Differential effects of differing intensities of acute exercise on speed and accuracy of cognition: A meta-analytical investigation. Brain Cogn. Sibley, B. Effects of an acute bout of exercise on cognitive aspects of Stroop performance.

Sport Exerc. Tomporowski, P. Effects of acute bouts of exercise on cognition. Acta Physiol. Oxf , — Google Scholar. Lahart, I. The effects of green exercise on physical and mental wellbeing: A systematic review. Public Health 16 , Pretty, J. The mental and physical health outcomes of green exercise.

Health Res. Thompson Coon, J. et al. Does participating in physical activity in outdoor natural environments have a greater effect on physical and mental wellbeing than physical activity indoors? A systematic review. Article ADS CAS Google Scholar. Berto, R. Exposure to restorative environments helps restore attentional capacity.

Bowler, D. A systematic review of evidence for the added benefits to health of exposure to natural environments. BMC Public Health 10 , Bratman, G. The benefits of nature experience: Improved affect and cognition. Urban Plan. Araújo, D. The empowering variability of affordances of nature: Why do exercisers feel better after performing the same exercise in natural environments than in indoor environments?.

Aspinall, P. The urban brain: Analysing outdoor physical activity with mobile EEG. Bailey, A. Cognitive benefits of walking in natural versus built environments.

World Leisure J. Berman, M. The cognitive benefits of interacting with nature. Brito, H. Benefits to performance and well-being of nature-based exercise: A critical systematic review and meta-analysis.

Article ADS Google Scholar. Gidlow, C. Where to put your best foot forward: Psycho-physiological responses to walking in natural and urban environments. Rogerson, M. Effects of the visual exercise environments on cognitive directed attention, energy expenditure and perceived exertion.

Public Health 12 , — Prince, S. Correlates of sedentary behaviour in adults: A systematic review. Byun, K.

Positive effect of acute mild exercise on executive function via arousal-related prefrontal activations: An fNIRS study. Neuroimage 98 , — Hyodo, K. Acute moderate exercise enhances compensatory brain activation in older adults.

Aging 33 , — Yanagisawa, H. Acute moderate exercise elicits increased dorsolateral prefrontal activation and improves cognitive performance with Stroop test.

Neuroimage 50 , — Dalsgaard, M. A reduced cerebral metabolic ratio in exercise reflects metabolism and not accumulation of lactate within the human brain. Developing the catecholamines hypothesis for the acute exercise-cognition interaction in humans: Lessons from animal studies.

Skriver, K. Acute exercise improves motor memory: Exploring potential biomarkers. Winter, B. High impact running improves learning. Kashihara, K. Positive effects of acute and moderate physical exercise on cognitive function. Ranjbar-Slamloo, Y. Dopamine and noradrenaline in the brain; Overlapping or dissociate functions?.

Acute exercise has a general facilitative effect on cognitive function: A combined ERP temporal dynamics and BDNF study.

Psychophysiology 54 , — Acute cardiovascular exercise and executive control function. The effect of acute treadmill walking on cognitive control and academic achievement in preadolescent children.

Neuroscience , — Kamijo, K. Differential influences of exercise intensity on information processing in the central nervous system. The interactive effect of exercise intensity and task difficulty on human cognitive processing.

Kaplan, S. The restorative benefits of nature: Toward an integrative framework. Stevenson, M. Attention Restoration Theory II: A systematic review to clarify attention processes affected by exposure to natural environments. Health Part B 21 , — Basu, A. Attention restoration theory: Exploring the role of soft fascination and mental bandwidth.

McMahan, E. The effect of contact with natural environments on positive and negative affect: A meta-analysis. Williams, C. The ERP, frequency, and time—frequency correlates of feedback processing: Insights from a large sample study.

Psychophysiology 58 , e Cohen, J. Statistical Power Analysis for the Behavioral Sciences L. Erlbaum Associates, MATH Google Scholar. Ioannidis, J. Why most published research findings are false. PLoS Med. Luck, S. An Introduction to the Event-Related Potential Technique MIT Press, Krigolson, O.

Choosing MUSE: Validation of a low-cost, portable EEG system for ERP research. Using muse: Rapid mobile assessment of brain performance. Delorme, A. EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis.

Methods , 9—21 Oken, B. Statistical issues concerning computerized analysis of brainwave topography. Download references. Theoretical and Applied Neuroscience Laboratory, University of Victoria, STN CSC, PO Box , Victoria, BC, V8W 2Y2, Canada. Faculty of Health, York University, Keele St, Toronto, ON, M3J 1P3, Canada.

You can also search for this author in PubMed Google Scholar. designed the study, wrote the manuscript, and analyzed the data.

was responsible for data collection, assisted with data analysis, and helped with manuscript preparation. and O.