Introduction
Relaxation videos gained in popularity not only on online streaming platforms like YouTube but in empirical research as well as indicated by a massive growth in publications related to the topic ‘‘relaxation video’’ over the past decades, with 8 publications listing ‘‘relaxation video’’ in the title in 1989, and 114 in 20201. To understand the psychological and physiological effects of relaxation videos, standardized measures to identify the various aspects of reactivity to relaxation are needed (Meier et al., 2020). In addition to subjective changes in affect and arousal, relaxation triggers changes in the autonomic nervous system (Chang et al., 2011). It can be described as a decrease in physiological arousal (Smith, 2007). Various physiological markers have been used to assess the changes associated with relaxation, such as a decrease in blood pressure and heart rate (HR). However, taking a closer look at the autonomic nervous system, which regulates the stress and relaxation response, shows that it is the increased activity of the parasympathetic nervous system that is associated with relaxation (e.g., Bertsch et al., 2012). Parasympathetic activity is related to decreased cardiovascular, respiratory, and electrodermal activity (Kreibig, 2010) and increase in positive mood (Shaffer and Ginsberg, 2017). A reliable measurement for parasympathetic activity is a vagally mediated marker of heart rate variability (HRV), high-frequency HRV (HF-HRV), which corresponds with respiratory sinus arrhythmia (RSA) (Acharya et al., 2006). Therefore, in this study we will look at relaxation as the increase in parasympathetic activity as indicated by RSA changes. While HF-HRV and RSA are the most specific indicators of parasympathetic activity (Thayer et al., 2012), HR is frequently reported in relaxation research and thus included as an additional parameter in this study.
Previous studies suggest an effect of relaxation videos including natural environments on autonomic regulation. For example, participants watching 360° nature videos (Irish countryside or Australian beaches) after a cognitive stressor showed significant stronger physiological relaxation than the control group, who viewed a video of an empty classroom, indicated by a reduction in electrodermal activity and an increase in HRV (Anderson et al., 2017). This line of research builds on several theories that explain why nature environments have a restorative effect, associated with an increase in parasympathetic activity, positive mood and replenish resources. For one, the Biophilia Hypothesis (Kellert and Wilson, 1993) describes biophilia as an inherent longing for connection to other living things, including flora and fauna. Further, the Attention Restauration Theory (Kaplan, 1995) focuses on the aspects of nature that help with recovery from mental fatigue, stating that the effortless preference for natural scenes is the needed counterpole to the stressful challenges of everyday life. Finally, the Psychoevolutionary Theory (Ulrich, 1983) emphasizes the esthetic preference for natural scenes and focuses on the evolutionary advantage of a quicker recovery from stressful situations. Recovery involves a positive affective reaction with a decrease in negative emotions like fear and reduction of physiological arousal. This reaction can be seen, for example, in the effects of a short walk in a forest which has been shown to reduce blood pressure and HR, increase HF-HRV (Lee et al., 2014), increase positive affect (Payne and Delphinus, 2019) and decrease negative affect, e.g., anxiety (Song et al., 2018). This effect of real-life nature seems to translate to sole representations of nature, like pictures or videos of rivers, forests, or even fireplaces (Dana Lynn, 2014). Watching nature pictures increased HRV in a resting state (Gladwell et al., 2012) and in a recovery phase after a stressor (Brown et al., 2013) but watching urban pictures did not. This hints at nature exposure leading not only to benefits in recovery phases but as well having buffering effects (Beute and De Kort, 2014). However, not all studies come to this conclusion. Van den Berg et al. (2015) found a greater recovery after a stressor while watching nature pictures compared to urban pictures, but no buffering effect. A 2 h long forest bathing experience (Yu et al., 2017) and looking at fresh roses (Song et al., 2017) had no significant effect on parasympathetic activity measures by HF-HRV. One study even found an opposite effect: HRV (indicated by RSA) decreased and heart rate increased when exposed to real-life nature during a 5-day field trip (Scott et al., 2021). However, most of these studies used real-life nature stimuli. While it can be assumed that real-life and virtual nature stimuli might elicit comparable effects (Browning et al., 2020) real-life nature stimuli lack standardization capability and accessibility across different research sites and sample composition. Moreover, these mixed results show, that further studies are needed to investigate, why exposure to natural stimuli only sometimes has relaxing effects. Keeping especially virtual nature interventions in mind when investigating this question is important since, video-based nature interventions could help scientific studies as they are easier to implement and standardize than a real-life outdoor setting that is subject to, e.g., weather conditions.
Even though there are studies showing an increase in relaxation in reaction to either a real (e.g., Song et al., 2018) or virtual nature environment (e.g., Anderson et al., 2017) as described above, there is little research about individual predispositions influencing the relaxation response. A first review suggests that physiological relaxation (e.g., HRV, blood pressure) in reaction to indoor nature stimuli (e.g., pictures of forests, potted plants) is influenced by age, gender and personality (Jo et al., 2019). Jo et al. (2019) referred to interindividual differences in general, without taking a closer look at specific factors. One of the factors possibly influencing the relaxation response could be early life adversity, i.e., the experience of adverse and potentially harmful situations in childhood and adolescence, like poverty, neglect or child abuse (Smith and Pollak, 2021). Such experiences can influence the autonomic nervous system and lead to lower resting levels of HRV (Sigrist et al., 2020), which might moderate the influence of early life adversity as a risk factor for psychological disease (Shonkoff et al., 2012). Moreover, early life adversity could alter an organisms capability to relax and thereby counteract negative effects of stress (Bhasin et al., 2013) as well, which could play a role in the mechanisms behind early live adversity being a risk factor for psychopathology (Lima et al., 2010). This could, for example, be demonstrated through a decreased HF-HRV in reaction to a nature based relaxation intervention, as can be observed in depressed patients (Matsui et al., 2016). On the opposite endpoint of the physiological activation-deactivation continuum, early life adversity is a prominent research factor that is assumed to alter the psychophysiological stress response (Tarullo and Gunnar, 2006). It is often operationalized as a lack of parental care (Luecken, 2000; Engert et al., 2010).
In contrast, trait mindfulness, i.e., the general ability of being aware of one’s present surroundings, emotions and thoughts, is associated with life satisfaction, optimism and mental health (Brown and Ryan, 2003) and could thus positively contribute to the relaxation response. It is part of a mindfulness practice to concentrate on what one can see and hear, without letting the thoughts wander to different, potentially stressful topics. Therefore, trait mindfulness could be an advantage when exposed to relaxation videos and increase their effects on the parasympathetic nervous system.
In two independent experiments, we aimed to investigate autonomic responses to different video contents and their generic ability to induce physiological relaxation in two consecutive studies, while exploring the potential impact of interindividual differences. In Study 1, we investigated whether popular nature-based videos (including audio) showing either a river stream in a forest, rain in the thicket, the crackling of burning wood in a fireplace, or waves crashing on a beach induce a psychophysiological relaxation response. We hypothesized that HR decreases and HRV increases as a result of the (video-based) nature exposure alone in comparison to a questionnaire baseline controlling for sex, age, and depressive symptoms (cf. Laborde, 2017). Building on the results of Study 1, we conducted Study 2 to extend the findings by adding a physiological baseline and two more video conditions: a guided meditation video and a control condition using a movie clip of “The Lord of the Rings.” In Study 2, we hypothesized that HR decreases and HRV increases during the nature videos and during the guided meditation video but not during the movie clip. In addition to the effect of different video contents, in both studies we investigated the potential influence of two psychological factors that might modulate the relaxation response: early life adversity as a risk factor for psychopathology and trait mindfulness as a protective factor associated with metal health. We thereby aimed to investigate their influence on relaxation as proxies for psychological or personality-related factors as reported by Jo et al. (2019). We hypothesized that early life adversity would lead to less increase of HRV values in reaction to the relaxation videos and trait mindfulness would increase the relaxing effects of those videos, leading to stronger increases of HRV.
General Discussion
To investigate the effect of a nature-based relaxation video intervention on autonomic regulation, we measured HR and RSA throughout a 10-min video intervention in two consecutive studies. As intervention material, we used a nature-based relaxation video in Study 1 and compared this relaxation video to a meditation video and a control movie clip in Study 2. Both studies found an increase in RSA and a decrease in HR during the video segments in relation to the questionnaires before and after the video in the two independent samples as expected. However, contrary to our hypothesis on group differences in Study 2, the only interaction effect that we observed indicated a stronger HR decrease during the videos in the meditation video group.
The overall decrease in HR and increase in RSA in response to the videos suggests a successful activation of the parasympathetic system as a result of video exposure (Acharya et al., 2006; Shaffer and Ginsberg, 2017) in general. This is backed by the majority of subjective rating of the videos as pleasant and relaxing. Thus, the focus of the general discussion shifts from nature-based relaxation videos to relaxation videos in general. Our findings are in line with recent empirical evidence on the effects of relaxation videos. For example, watching videos at the workplace is helpful to recover from the stress and demands of work, increasing wellbeing, work satisfaction and relaxation (Janicke et al., 2018). Additionally, a video-based relaxation program successfully reduced anxiety, depressive and somatic symptoms in elderly participants with an anxiety disorder (Gould et al., 2019). Because of this and the convenience to standardize video interventions, they appear as a seminal method to induce relaxation in research settings and psychotherapy. Beyond that, experiencing relaxation is beneficial and linked to wellbeing and mental and physical health (Yu et al., 2010). However, standardized protocols with adequate control conditions and longitudinal investigations of the long-term effects of relaxation videos are missing.
While we hypothesized that nature and meditation videos would lead to a stronger relaxation response than the movie clip, we found a general response to any video that participants watched in both of our studies. First and foremost, this could be a general effect of sustained attention to a pleasant and non-threatening video stimulus. In contrast to the physiological baseline, looking at a blank classroom wall, or the questionnaire phases, looking down on a paper with the different questionnaires, the video intervention required participants to direct their attention to the screen which by itself might affect autonomic regulation (Holzman and Bridgett, 2017). Looking closer at movie clip content reveals another possible explanation for the comparable increase in RSA in this condition. The movie scene from “The Fellowship of the Ring” is set in the Shire with green hills, widespread meadows, and a blue sky underpinned with beautiful calm music. The contents are similar to those of the relaxation videos: a calm nature scene accompanied with instrumental music. When planning this study, we attempted to align the control condition as much as possible with the relaxation video, e.g., in colors and general atmosphere, yet without the specific aim of a relaxation video but with a distracting movie content similar to other studies (Zeidan et al., 2010). However, these features that made the movie clip comparable to the nature relaxation videos might have induced a comparable level of relaxation in the control condition compared to the other videos. According to the Attention Restoration Theory (ART, Kaplan, 1995), an environment needs four different aspects to trigger relaxation: Fascination, an environment’s ability to capture involuntary attention, Being Away, a physical and mental distance to everyday life, Extent, the sufficiency of an environment to engage the viewer’s mind, and Compatibility, the fit between the viewers’ intentions and the possibilities the environment offers. All four aspects of the ART can be identified in the nature-based relaxation videos, the meditation video, and the movie clip alike, giving a possible explanation of why RSA as an index of relaxation increased in all conditions. As such, our choice of a control condition using this particular excerpt from the movie was perhaps not a control condition, but a relaxing intervention as well.
After analyzing the effect of the different video conditions on autonomic regulation, we included parental care and overprotection, and trait mindfulness measures as secondary predictors to the mixed models and found inconclusive results with regard to HR. While higher maternal and paternal care predicted lower HR during the video segment in Study 1, we found no significant influence of parental care or overprotection on HR in Study 2. This speaks perhaps for a smaller effect, however our statistical power and hence our chances to detect a smaller effect were actually improved in Study 2 with a larger sample size. Results appear more consistent for RSA and measures of early life adversity. In Study 1, higher paternal care predicted higher RSA before and after but not during the video. In Study 2, maternal care was related to higher RSA, while maternal overprotection was associated with lower RSA in response to the video. In addition, in Study 2, higher paternal overprotection predicted lower RSA independent of time.
While the link between HRV and psychopathology, e.g., in post-traumatic stress disorder or depression, is well established (Beauchaine and Thayer, 2015), the discussion to which extent early life adversity, particularly careless or overprotective parenting, in particular, contribute to autonomic dysregulations or an overall lower HRV is still ongoing. For example, lower resting HRV was predicted by higher inconsistent discipline, corporal punishment, and lower parental involvement (Graham et al., 2017) and a blunted HR stress response is linked to the experience of early life adversity (Voellmin et al., 2015).
It is important to consider the various means of observing HRV. While many studies look at RSA at baseline, or resting condition, there is also a number of studies who have looked at RSA changes in response to stress. These have not shown systematic effects, however. For example, RSA change in reaction to the Trier Social Stress Test (Busso et al., 2017) could not be associated with experiences of interpersonal violence or poverty. Similarly, in healthy women with adverse childhood experiences RSA did not mediate the physiological reaction to the Montreal Imaging Stress Task (Winzeler et al., 2017). It is conceivable that a stressor is not the right experimental manipulation to observe systematic associations between the parasympathetic nervous system and variables of personality, or early life adversity. The parasympathetic system is not being stimulated by a stressful situation, but will withdraw; thus, in these cases the experiment will investigate not whether the system can be activated, but if it can be shut down. The null findings in these studies could be a consequence of this distinction.
Our results support the findings that lower parental care and higher parental overprotection might lead to parasympathetic dysregulation that could be associated with psychopathological development later in life (Meyer et al., 2016). Especially the observation that higher paternal overprotection was associated with an overall lower HRV sheds light on the often-overlooked role of fathers’ parenting behavior in psychophysiological development. For example, it was found that fathers appropriate mind-related comments (e.g., “validation of infant’s internal state”) increased the baseline HRV of their one-year old child, independent of the mothers influence (Zeegers et al., 2018). In other aspects the influence of fathers’ parenting behavior on children is similar to the influence of mothers parenting, for example, the same behaviors promote secure attachment (Fagan et al., 2014). Additionally, the effect of mothers and fathers parenting style on their children influence each other (Fagan et al., 2014). Because fathers influence is less well understood and researched as mothers influence (Unternaehrer et al., 2021), it seems highly recommended to include both aspects of parenting in our research on the influence of early life adversity and parenting behavior.
On the other hand, for trait mindfulness, again mixed findings were observed: contrary to our expectations, we found higher scores on the observing subscale (indicating better mindful observing skills) to predict higher HR and lower RSA during the video in Study 1 only. This finding that higher scores of trait mindfulness, especially mindful observing skills, are associated with higher HR and lower RSA seems counterintuitive at first since it suggests less relaxation during the video. We assessed trait mindfulness using the FFMQ, including observing as one subscale, which can be seen as a cognitive function linked to attention. It was found that mindfulness positively influences executive control (Verhaeghen, 2021), self-regulation (Tang and Posner, 2009), and attention (Jha et al., 2007; Tang and Posner, 2009; Verhaeghen, 2021). This link between mindfulness and cognitive functions in general, and attention in particular, might explain why parasympathetic activity was lower during the videos for participants who scored high on the FFMQ-observing subscale. When a cognitive challenge, like focusing on something, arises, it decreases parasympathetic activity and, therefore, HRV (Giuliano et al., 2018). Thus, it is possible that scoring high on mindful observing skills goes along with increased attention to one’s surroundings and internal processes. This association between mindful observing and cognitive effort could affect innervation of the parasympathetic nervous system. Additionally, higher FFMQ scores are associated with better performance in sustained attention (Rice and Liu, 2017) and higher flexibility in changing one’s attention focus, for example directing attention from unimportant stimuli to those needed to accomplish a task (Sørensen et al., 2018). These links suggest that mindfulness contributes to the ability to willingly direct attention, which can be interpreted as voluntary attention, as mentioned in the Attention Restoration Theory (Kaplan, 1995). According to this theory, voluntary attention stands opposite to involuntary attention, which is needed to achieve restoration. Therefore, it could be concluded that the aspects of mindfulness focusing on voluntary directing attention counteract the achievement of restoration. This counteraction could inhibit relaxation and therefore explain the low HRV while viewing the videos. In line with this argumentation, Watford and colleagues reported overall lower HF-HRV in persons scoring higher on a trait mindfulness questionnaire. These results seemed to be linked to heightened vigilance and cognitive load in more mindful participants (Watford et al., 2020). In our data, this effect was only visible in Study 1 and could not be replicated in Study 2 and should thus be interpreted with caution.
The relationship between HRV and mindfulness experience is a complex one. We found that less mindfulness experience is associated with a lower HRV in general but scoring high on trait mindful observing skills leads to reduced HRV during the video. It is important to note that there is a fundamental difference between resting-state HRV and HRV in reaction to a stimulus. Therefore, it is not counterintuitive that mindfulness skills might lead to an increase in baseline HRV but prohibit a strong increase in reaction to a relaxation video.
Limitations and Future Research
Several limitations should be considered when interpreting these results. First and foremost, we planned Study 1 with more liberal exclusion criteria and included a questionnaire baseline only. Adding a physiological baseline (Laborde, 2017) in Study 2 allowed for a more rigorous design but at the same time limited comparability between the two studies. Since the baseline measure of Study 1 was taken while participants filled out questionnaires it is possible that this activity influenced the HRV, making it difficult to compare the baseline levels of HR and HRV with the video response. In Study 2, however, neither HR nor HRV changed significantly between the physiological baseline and the first paper-pencil questionnaire segment. Nevertheless, it was not possible to combine data from Study 1 and Study 2. Thus, sample size might have been too small to detect small effects for secondary predictors and replications in bigger samples are needed to investigate the influence of different psychological characteristics like early life adversity or trait mindfulness on autonomic regulation.
Second, as discussed previously, the selected video material might not have been appropriate to investigate differential effects of a nature videos for several reasons: (a) content of the control video, (b) video presentation on laptop screens, and (c) unbalanced nature sceneries in the nature video condition. The control video, as discussed before, might have been too similar to the experimental condition and thus, we failed to investigate group differences even though we attempted to create comparable control intervention with the movie clip. Many studies researching the relaxing effects of a natural environment have used urban environments to search for group differences. This can be seen in real-life experiments, for example, the effect of a walk in a forest compared to a walk through a city (Song et al., 2018). The forest walk shows a positive effect on anxiety and negative affect. Some experimenters used virtual representations of nature and urban environments to compare their effects on physiological and psychological parameters (Gladwell et al., 2012; Tang et al., 2017). Intervention material displaying urban environments in comparison to nature environments might thus be better suited for this line of research. Furthermore all videos were presented on rather small laptop screens (17-inch in Study 1 and 13-inch in Study 2). As we cannot exclude the possibility that screen size might have influenced relaxation effect, a more systematic investigation of effects of presentation mode (e.g., screen size and resolution, seating distance, acoustic properties) would be helpful for future studies. Moreover, we abstained from analyzing differential effects of nature video content (e.g., forest vs. fireplace) due to unbalanced groups even though. As soundscape studies have revealed, not only visual but also acoustic characteristics of rural videos elicit a psycho-physiological relaxation effect (e.g., subjective ratings as pleasant and restorative, HR decrease) with differential effects, e.g., of water sound compared to bird songs (Ratcliffe, 2021). While we werent́ able to compare different video characteristics within the nature video condition, future research should focus on this particular comparison or determine only one scenery at the start, for example “rain falling in the thicket” which was the most prominent video in this study.
The observation of overall lower RSA during the video condition in Study 2 was unexpected, but as this main effect was independent of the video intervention, it is most likely linked to group differences that exist beyond the experimental manipulation of our study. Though we tried to form comparable groups by quasi-randomly assigning participants to the different conditions, a more comprehensive pre-screening might be necessary in future studies to prevent such an effect. This should be applied especially when psychological characteristics that could constitute group differences are part of the research focus.
Fourth, both study samples consisted of relatively young participants, with an age range of 18 to 49 and only n = 4 participants older than 30 years. Taking age into account when measuring HRV is especially important since HRV decreases with age in our and other studies, mainly between 20 and 35 years (Hirsch and Bishop, 1981). Therefore, our study findings are limited to young adulthood, and future studies might want to recruit a sample with a wider age range to be able to measure the reaction to the intervention over the whole lifespan. Additionally, the sample consisted of mostly healthy participants. Only in Study 1 were some participants with a BDI-II score above 18 (n = 9) which suggests clinically relevant depressive symptoms (Hautzinger et al., 2006). Like age, mental and physical diseases influence HRV (Acharya et al., 2006; Shaffer and Ginsberg, 2017) and should be taken into account. Only adding healthy participants to the sample was necessary to maximize the comparability of the different groups and limit additional factors influencing the HRV. Beyond age and health, various factors influence HRV, which future studies should take into account: for example, the menstrual cycle (Vallejo et al., 2005), sleep (Glos et al., 2014), and smoking (Barutcu et al., 2005). In addition, it would be interesting to not only look at sex effects but take gender effects into account as well. Furthermore, we measured BMI in Study 2 only and cannot draw any conclusions about its influence in the sample of Study 1. The impact of between-subject differences is reflected by the inter-individual variance as indicated by random effects. While this points toward individual factors that might explain additional variance, the control for random effects in the multilevel model approach is a strength of this data-driven approach (Curran et al., 2010), Those factors could be measured to control their influence or a within-subject design could ensure that those factors do not influence the differences between two interventions. Yet, when using a within-subject design, sequence effects must be taken into account.