In this post, I look at the long-term effects of deep breathing on heart rate variability (HRV) as measured during deep breathing practice
​
While there is plenty of data and published literature on the acute effect of deep breathing on HRV (basically the difference between resting conditions and practice), we know much less about long-term effects. Looking at this data might help us better understand the relationship between deep breathing and long-term physiological changes (if any!)

Enjoy the read

With the next update of HRV4Biofeedback, we are bringing to the app one of the most frequently requested features: the possibility to select other inhale to exhale ratios, apart from the standard 50 / 50

The update will be available in a few days and will provide three options:

• 50 / 50: inhale and exhale are of the same duration. This is the option currently available in the app
• 40 / 60: inhale is 40%  of the cycle, therefore exhale is a bit longer
• 1 / 2: inhale is half the exhale period

​In all cases, the base breathing frequency will be the one you have configured manually or determined via the resonant frequency test. This base frequency will be modulated according to your selection, to allow you to use different ratios

Enjoy

​In my latest blog post, I go over the basics of heart rate variability (HRV) biofeedback and show changes in baseline physiology (resting HR, HRV) potentially linked to practicing deep breathing at resonant frequency consistently for the past month

Learn more, here 

Get breathing!

We have added a real time panel to the app, showing the cumulative biofeedback score (which is based on the session duration as well as the breathing frequency), the average heart rate and HRV (rMSSD) of the session

In order to display the new panel, you need to select it in Settings. The data will be updated every minute during the session

We have added the power spectrum of your biofeedback session directly in the app. You can find it in the detailed view of the session, that you can access from the History menu

Normally, you should see the highest peak in the low frequencies, very close to the breathing rate used for your session. You can see here below an example of a session where the power spectrum peak is at about 0.09 Hz, which then the app translates into a breathing rate of 5.3 breaths / minute

Here is an example of a session without deep breathing, just to showcase how a resting breathing rate of about 10 breaths / minute is picked up in this case:

Enjoy the new feature!

​We have added integration with the Health app in the next version of HRV4Biofeedback, which will allow you to write Mindful minutes, heart rate and HRV (SDNN in this case) to the Health app

Thank you for your feedback. Enjoy!
​

​Quite a few of you have asked for the HRV (rMSSD) of the session to be displayed in the app. With the latest update, you will be able to find it in the detailed session view (under History) as well as under Summaries and Correlations

Enjoy
​

Before we start, I believe there is an important point to cover, concerning performance measures

In elite sport settings, performance is often the outcome of interest. However, in many sports (e.g. in teams settings), performance cannot be unambiguously measured (Cannon-Bowers and Salas, 1997; Richard et al., 1999; Wiseman et al., 2014), and is often estimated using different approaches

For example, in many situations, athletic performance is measured during isolated tasks (e.g. sprinting ability), which might have low fidelity with respect to the complexity of an actual game. Even when sport-specific tests are used (e.g. a passing test in soccer), it is difficult to extrapolate test results to in-game performance, as often shown by research on talent identification (Den Hartigh et al., 2018; Bergkamp et al., 2019). For these reasons, focusing only on studies that clearly report changes in performance can be limiting. On the other hand, it follows from the previous considerations on HRV Biofeedback (see parts 1–3 of this guide) that physiological and psychological parameters might be mediating the relation between HRV Biofeedback and performance

For example, HRV Biofeedback could increase parasympathetic activity at rest or improve emotional self-regulation during high anxiety tasks (Lehrer et al., 2003, Deschodt-Arsac et al., 2018), which might be beneficial for performance, as further discussed in the next sections. Hence, my goal here is to provide an overview of the effects resulting from the implementation of HRV Biofeedback methods in athletes. In particular, while it makes a lot of sense to first investigate sport-specific changes in athletic performance measures, I will also look at physiological and psychological changes that have been documented in the literature and can mediate athletic performance

In other words, in this post, I aim at answering the following questions: what is the effect of HRVB on athletic performance? How do physiological and psychological measures change following HRVB interventions in athletes?

A summary of the studies and outcomes can be found in the table below, while a thorough analysis of the studies organized by type of outcome (performance, physiological or psychological), is provided in the second part of this blog post

One of the first studies to use the standard HRV Biofeedback protocol previously introduced (Lehrer, Vaschillo, and Vaschillo, 2000) was carried out by Lagos et al. (2008). This was a case study of a junior competitive golfer where sport-specific performance measures improved after the HRV Biofeedback intervention (number of strokes required to complete a 18-hole competition). Additionally, the authors reported improvements for both physiological (increase in HF, a marker of parasympathetic activity) and psychological measures (Profile of Mood States (POMS) and Competitive State Anxiety Inventory (CSAI-2)). On the other hand, the same year, Tanis (2008) investigated the effect of HRV Biofeedback on volleyball performance, psychophysiological regulation and perception, with less promising results. In this case, sport-specific performance was defined as the individual rating of four sport-specific skills during each game, as determined by a coach. Psychophysiological regulation was measured as coherence, an artificial construct relating to an individual’s performance during biofeedback. From a physiological point of view, higher coherence should be reflected in higher HRV at the resonant frequency or higher overall HRV. The authors found no improvement in performance after six weeks of HRVB in a team of 13 athletes. Coherence (or HRV) as measured at baseline before each session, also did not improve during the intervention but only at the end of each session, highlighting a possible acute effect. In terms of psychological outcomes, athletes reported perceived improved mental performance

Additional studies in the following years were again able to show positive outcomes for HRV Biofeedback interventions. In particular, in Perry, Shaw, and Zaichkowsky (2011) the authors reported performance improvements both in gymnasts and hockey players following a 8–10 sessions biofeedback intervention. Anecdotal evidence on the positive psychological effects of biofeedback are reported also in two studies targeting elite athletes preparing for the Olympic games (Dupee and Werthner, 2011; Beauchamp, Harvey, and Beauchamp, 2012). In both studies however, no performance, physiological or psychological measures were quantified, and only anecdotal results were reported (two gold medals won and improved team ranking, Beauchamp, Harvey, and Beauchamp, 2012)

A common issue with studies targeting elite athletes is that often several interventions are carried out at the same time, as reported for example in Olympic speed-skaters undergoing a multi-year program targeting psychological training (e.g. PST), neurofeedback, as well as HRV Biofeedback (Beauchamp, Harvey, and Beauchamp, 2012). As a result, it can be difficult to isolate the effect of HRV Biofeedback

To sum up, most of these early studies reported positive outcomes in terms of psychological measures, and mixed findings in terms of physiological measures and performance improvements (Tanis, 2008). However, none of these studies included a control group, and some of the reported improvements in performance were either anecdotal (Beauchamp, Harvey, and Beauchamp, 2012) or involved a single athlete (Lagos et al., 2008)

In recent years, more and more studies have started including control groups to better account for the effect of HRV Biofeedback independently of other changes that might be occurring as training progresses

Two studies in particular (Pusenjak et al., 2015; Deschodt-Arsac et al., 2018), evaluated physiological responses to stress tests (STROOP and math tests as well as university exams), in high level and elite athletes of various sports. In both cases, control groups were present, and only the intervention group showed improved physiological responses during the stress tests, for example reduced heart rate, reduced breathing rate and increased HRV. Changes in physiological parameters during stress tests are not commonly measured, but might highlight improved emotional self-regulation due to HRV Biofeedback

Psychological measures also improved following the intervention, as reported in terms of standard tests (State-Trait Anxiety Inventory, STAI) or in short questionnaires evaluating subjectively the positive impact of the intervention

Unfortunately, none of the studies reported performance measures, which is a common issue especially when targeting athletes of different disciplines. Similar findings were also reported in another intervention targeting athletes of various sports (Dziembowska et al., 2016). In this case there were no measurements during stress tests, but HRV was measured before and after biofeedback training, showing improvements in the experimental group. In terms of psychological measures, both self-esteem (Rosenbert Self-Esteem Scale (SES)) and state anxiety (STAI) improved after the intervention in the experimental group, while no changes were reported in the control group

Other studies in more homogeneous populations (e.g. including athletes of a single sport), provide more insights in terms of performance outcomes (Paul, Garg, and Sandhu, 2012; Paul and Garg, 2012; Choudhary, Trivedi, and Choudhary, 2016; Wakefield and Shipherd, 2017). Wakefield and Shipherd (2017 measured performance in a strength task (one repetition maximum) following an HRV Biofeedback intervention. In this case, performance was improved but was no different than the improvement in the control and alternate groups. This study clearly shows the importance of a control group, as improvements in performance can occur independently of HRV Biofeedback. On the other hand, other studies reported performance improvements only in the intervention group. For example, in Paul, Garg, and Sandhu (2012), the authors performed an intervention in basketball players and included placebo (watching motivational videos) and control (no intervention) groups, reporting improved reaction time, movement time and shooting performance in the HRV Biofeedback group. The control group also improved performance, but to a lesser extent. Similarly, physiological (HRV features and respiration rate) and psychological (concentration) parameters improved only in the HRV Biofeedback group. The same authors carried out an additional study but with an interesting twist, selecting only players with high trait anxiety (Paul and Garg, 2012). This is particularly relevant as the HRV Biofeedback intervention might have a different effect based on personal traits. In this second study, all measures (performance, psychological and physiological) improved following the intervention. The two studies just mentioned are also among the few that included follow-up analysis (Paul, Garg, and Sandhu, 2012; Paul and Garg, 2012). In particular, the authors re-assessed all measures one month post-intervention, and found consistent results. Performance improvements were also found in another study looking at HRV Biofeedback in track athletes (Choudhary, Trivedi, and Choudhary, 2016)

Needless to say, standard HRV Biofeedback protocols requiring athletes to spend significant time at the laboratory each week, and can be impractical for time-crunched elite or recreational athletes. As a result, sustained practice can become a challenge. Thus, in very recent studies, a variety of different approaches and methods have been employed in order to make HRV Biofeedback more practical, for example using mobile apps and mostly home-based practice as well as sessions as short as 3 minutes. However, mixed results have been reported

In a first study, HRV Biofeedback was used in professional soccer players (Rijken et al., 2016). In this case there was no control group, but interestingly various measures were reported not only post-intervention but also at 5 weeks follow-up. In terms of performance, each player self-reported performance using a numeric scale, and reported no change throughout the study or at follow-up. Physiological measures (HRV analysis in the frequency domain) showed results consistent with previous literature (Dziembowska et al., 2016), such as increased LF and reduced HF. Among a series of psychological parameters collected with standard questionnaires (Recovery-Stress questionnaire for athletes and Sport Improvement Measurement-60), improvements were reported for emotional stability and concentration capacity only. In a recent study (Perry, 2018) used a combined HRV Biofeedback and mindfulness protocol in 4 junior soccer players. Performance was assessed using a sport-specific test (Loughborough Soccer Passing Test), and improved over time. No changes were reported in physiological (heart rate, respiration rate, skin conductance and HRV) and psychological (CSAI-2) parameters

Finally, Mueller et al. (2019), carried out a HRV Biofeedback intervention on an elite triathlete and provides additional insights on baseline changes in HRV. In particular, in this study for the first time resting physiology (heart rate, HRV) was measured in unsupervised free-living settings using a commercially available mobile phone app. Performance was assessed in terms of time spent training as well as subjectively, and improved during the study. Morning HRV, a marker of parasympathetic activity and overall physiological stress was improved during the intervention. In terms of psychological measures, school stress was assessed with a numerical scale and remained unchanged

Results in terms of performance outcomes have been conflicting so far. Part of the issue has to do with the difficulties of measuring performance. While for certain sports, in particular individual endurance sports, performance testing is often done routinely, this is not the case for team settings where performance testing is hardly performed or standardized. Additionally, when performance testing is carried out with a test that aims at simulating certain skills required during a game, it is debatable how such tests relate to in-game performance (Den Hartigh et al., 2018; Bergkamp et al., 2019). Another issue with standard tests, especially if the athletes have not performed them before the study, is the learning effect that would make them better at the task regardless of the intervention. This is the case in a few studies that showed improved performance in literature (Perry, Shaw, and Zaichkowsky, 2011; Paul, Garg, and Sandhu, 2012; Perry, 2018)

In elite settings where a control group was lacking (e.g. Canadian Olympic athletes), anecdotal evidence has been positive, but it is difficult to determine if HRV Biofeedback was particularly beneficial in the athlete’s quest for a gold medal. Other studies that have shown improvements also relied on subjective assessment of athletic performance, or on metrics that might have little to do with performance, for example time spent training in an elite triathlete (Mueller et al., 2019). Similar criticisms can be made for studies that showed no improvements (Tanis, 2008; Rijken et al., 2016), as performance was assessed subjectively in both volleyball players (from their coaches) and elite soccer players (self-assessed)

From the performance outcomes reported in HRV Biofeedback interventions so far, it seems difficult to determine if there is any direct positive effect of HRV Biofeedback on performance. While some of the aspects just covered can be better accounted for (e.g. including an active control group, using standard performance tests or preferably in-game performance, as well as using standard protocols for interventions), in certain sports it can still be very challenging to effectively determine performance changes following an intervention, for example in team sports or sports with a strong tactical component (Cannon-Bowers and Salas, 1997; Richard et al., 1999; Wiseman et al., 2014; Den Hartigh et al., 2018; Bergkamp et al., 2019)
​
For these reasons, and considering the important physiological and psychological changes that can be impacted by HRV Biofeedback, and how physiological and psychological changes can mediate performance, it can be particularly interesting to determine if there are more consistent findings when it comes to such measures

Psychological measures following HRV Biofeedback interventions are probably the most consistent in terms of positive outcomes. In particular, the various studies investigating effects on anxiety (both trait and state) as well as on self-esteem and self-efficacy, often found improvements in most measures (Lagos et al., 2008; Paul, Garg, and Sandhu, 2012; Paul and Garg, 2012; Dziembowska et al., 2016) with only one study reporting no changes in a small sample of younger athletes (Perry, 2018). While a few studies reported only anecdotal evidence or qualitative feedback from the athletes, a number of studies used standard questionnaires to assess anxiety, stress and mood

In general, psychological parameters seem to benefit from HRV Biofeedback interventions, especially for what concerns anxiety, similarly to what has been reported in literature outside of sports

In terms of physiological measures, results are also quite consistent across studies. However, an important caveat here needs to be considered. In particular, I have stressed before how physiology can be measured at different times, and in the case of HRV Biofeedback, how measurements taken during the session are reflective of acute changes as a result of deep breathing and RSA, and how these changes are not necessarily linked to baseline physiological changes in parasympathetic activity. Hence, while various studies have consistently shown increases in HRV, and in particular in the LF band (due to breathing at 0.1Hz), these changes do not imply any adaptation in the ANS (Dziembowska et al., 2016; Rijken et al., 2016). These are important aspects to consider as results differ greatly, as we can see from the few studies that measured both before and during HRV Biofeedback, and often found no improvement in HRV before the session, but increases during and right after each session (Tanis, 2008)

When it comes to baseline changes, very few studies have investigated HRV outside of the HRV Biofeedback session

In a few cases, baseline measurements were taken before the session, resulting in conflicting results (Tanis, 2008; Lagos et al., 2008). While this protocol might not be ideal, as other confounding factors are most likely present (e.g. physical activity beforehand, food or caffeine intake, etc.), a pre-session measurement is at least not affected by RSA as much as the actual HRV Biofeedback session. In this case, the study that found no change was shorter (6 weeks instead of 10), which might point to a required minimum dose for baseline physiological changes to take place. Only one study has measured changes in resting HRV from data collected first thing in the morning, showing an increase over time. However, this was a case study with only one participant, no control, and the change in HRV was associated with a change in training load, hence it is not possible to determine causality

In other studies HRV is reported as improved, similarly to respiration rate being reduced, but it is often unclear when these parameters were measured (Paul, Garg, and Sandhu, 2012, Choudhary, Trivedi, and Choudhary, 2016). Given how short some of these studies were (down to 10 days), it seems unlikely that baseline changes in physiology were assessed. It is also highly unlikely that a person would modify respiration rate from an average of 12–14 breaths/minute to exactly 6, again highlighting how despite the lack of clarity in terms of the protocol, changes reported are most likely acute changes during the HRV Biofeedback intervention, and not baseline assessments of ANS activity
​
While these acute changes are unrelated to baseline effects, they are still very important. In particular, consistent results form a variety of studies show that athletes are able to learn HRV Biofeedback and that breathing at their resonant frequency does synchronize breathing and heart rate in a way that is quantifiable and can have beneficial effects. I have introduced earlier the various pathways that try to explain the relation between HRV Biofeedback and positive changes in physiological and psychological variables (see part 1 of this guide), and we have seen how certain psychological aspects are consistently improved following HRV Biofeedback interventions, for example anxiety. Reduced anxiety could be a beneficial effect of HRV Biofeedback due to deep breathing and changes in baroreflexes regardless of baseline changes in the ANS. Yet, the relationship between acute changes in HRV, baseline changes in HRV, and psychological measures following an intervention is complex and requires further investigation

In this series of posts, we have provided an overview of research on HRV Biofeedback and athletic performance, as well as of the main effects on physiological and psychological parameters that can mediate performance outcomes

While researchers have been quite active in the last two decades, trying to determine if there is a positive impact on performance due to HRV Biofeedback interventions, the quality of most studies is fairly low and methodology is inconsistent. Case studies or anecdotal evidence is often reported, with very few studies using active control groups. Additionally, performance is almost always quantified subjectively or using simple tests that might not be reflective of actual performance

Based on the available evidence, HRV Biofeedback can be considered an effective tool to reduce anxiety as well as acutely improve HRV, and therefore can be considered valuable in the context of emotional self-regulation even in an athletic population
​
Hopefully, the availability of tools such as HRV4Biofeedback will make it easier for scientists, athletes and coaches to further investigated these aspects

For each breathing frequency, we report the low frequency power as well as SDNN, a feature that captures the overall oscillations in instantaneous heart rate, so that you can see differences between various rates for your tests and your clients'

The resonant frequency as determined by the app, is also highlighted in red:

Finally, the power spectrum will report the main frequency of the session:

HRV Biofeedback has been first formally defined by Lehrer, Vaschillo, and Vaschillo (2000). The authors proposed a protocol consisting of a 10−session program. According to this original protocol, a biofeedback device is needed, which is a device able to display heart rate instantaneously as well as respiration rate, at least in terms of a pacing signal which allows the user to follow a given breathing frequency. The instrument should ideally be available also for home use and not only at the laboratory
​
The protocol starts with a first introductory session in which the participant is instructed about how heart rate changes with respiration (e.g. heart rate increasing when breathing in), and how biofeedback can help exercise reflexes in the body and therefore control more efficiently the autonomic nervous system (Lehrer, Vaschillo, and Vaschillo). The participant is informed that sustained practice can help improve health outcomes and manage stress

During the first session, the resonant frequency of an individual needs to be established. The resonant frequency is typically around 6 breaths/minute, but can change by 0.1–1.5 breaths/minute between individuals. Hence, each participant should try a range of frequencies and the practitioner should determine the optimal frequency by analyzing the power spectrum of each different test. Each test in this phase lasts 3 minutes, and the resonant frequency will be the one with the highest amplitude

In HRV4Biofeedback, we have introduced a resonant frequency protocol that works similarly, and guides you through a series of breathing frequencies to determine which one is optimal for you, as described here

At this point, the participant is instructed to practice at home for two 20 minutes periods each day. During the second session, the participant is instructed to look at heart rate and to try to maximize oscillations in heart rate while synchronizing breathing and heart rate. The participant can be instructed about abdominal breathing and pursed lips breathing, and is asked to practice for two 20 minutes sessions per day again. The following eight sessions are mostly used to acquire further experience with the technique. The protocol was used by the same authors in one of the first HRV Biofeedback interventions (Lehrer et al., 2003), where participants were instructed to practice twice per day for 20 minutes, for one week


The study population, in this case, was not athletes, however I have included the protocol in Table 1 as this is the original protocol employed in most of the following studies on athletes

In Paul, Garg, and Sandhu (2012), a slightly different approach was used. In this case, the protocol was more compact, still providing 10 sessions, but on consecutive days. Each session lasted 20 minutes and included breathing at the athlete’s resonant frequency. The protocol, in this case, was similar to what I have described earlier (Lehrer, Vaschillo, and Vaschillo, 2000; Lehrer et al., 2003), with participants breathing at slightly different frequencies during the first session, so that the resonant frequency could be determined (4.5 to 6.4 breaths/minute). A pacing stimulus was used to aid the process, and the participants were instructed to breathe abdominally and exhale with pursed lips. The same protocol was also used in another study by the same authors (Paul and Garg, 2012). In these studies, no home practice was reported, possibly because of practical reasons

Another variant of the same protocol was also used by Choudhary, Trivedi, and Choudhary (2016), consisting of a 10-week long intervention where paced breathing was practiced at the athlete’s resonant frequency. The supervised sessions lasted 30–40 minutes including baseline measurements. Additionally, participants were instructed to practice twice per day at home, for 20 minutes each session. Two 20 minutes sessions per day at home plus 10 supervised sessions in the laboratory were also reported in a few other HRV Biofeedback interventions (Lagos et al., 2008, Dziembowska et al., 2016). During these interventions, a standard HRV Biofeedback protocol with deep abdominal breathing following a pacing stimulus was used, as also employed in previous studies (Lehrer, Vaschillo, and Vaschillo, 2000; Lehrer et al., 2003; Paul, Garg, and Sandhu, 2012; Paul and Garg, 2012). Before the intervention, the resonant frequency of each participant was also determined

Finally, other authors relied on the original protocol but shortened the amount of home-based practice. In one case, the authors used a combined HRV Biofeedback and mindfulness program lasting 10 weeks but with only one 20 minutes session per day (Perry, 2018), while in another study the authors used a rather short 4 weeks intervention comprising of only five 15 minutes sessions per day (Mueller et al., 2019)

A few other studies used different non-standard protocols, often consisting of a single weekly session of varying duration (between 6 and 60 minutes). In particular, HRV Biofeedback was part of a multi-year program targeting Olympic speed-skaters in Canada (Beauchamp, Harvey, and Beauchamp, 2012). While this study included several interventions (psychological skill training or PST, neurofeedback, etc.), HRV Biofeedback was an important element, with weekly sessions of 45 minutes. Weekly sessions were also used in Tanis (2008). The authors administered a protocol which consisted of 30 minutes sessions once per week in supervised settings, combining biofeedback with positive emotional focusing. While the supervised session was 30 minutes, only 5 minutes were used for HRV Biofeedback, and the remaining part of the session was used for baseline measurements and to provide information or feedback on previous sessions. Participants were also encouraged to practice at home, but no data was reported on home use. The intervention lasted 6 weeks

Two biofeedback interventions with weekly sessions are reported in Perry, Shaw, and Zaichkowsky (2011), one on gymnasts lasting 10 sessions and 5 weeks and one on hockey players and lasting 8 sessions. In the first study, biofeedback was combined with neurofeedback and no details are reported on the exact protocol. The duration of the sessions is reported only for the second intervention, where they lasted 12 minutes. No information is provided on how the biofeedback sessions were delivered, however the authors report a constant breathing frequency of 6 breaths/minute for each athlete, hence it seems that in this case biofeedback was taught in terms of synchronization between heart rate and breathing, but without determining the optimal resonance frequency for each athlete. Short weekly sessions were also reported in another intervention, where they lasted 10 minutes for 6 weeks (Wakefield and Shipherd, 2017)

On the opposite side of the spectrum, either very long weekly sessions or very short multiple times per day sessions have been proposed in a few studies. Biofeedback was combined with neurofeedback in Dupee and Werthner (2011), where the authors report 60 to 90 minutes sessions, for a total of 20–40 hours per athlete. In this case, paced breathing lasted between 6 and 20 minutes, and it increased over time. Similarly, Pusenjak et al. (2015) used an 8-week intervention with two one hour sessions each week. It is unclear what were the specifics of these sessions in terms of biofeedback and instructions to the participants. However, the authors report asking the participants to perform breathing exercises at home (without using a device, in my understanding). On the other hand, Rijken et al. (2016), used a home-based protocol with daily sessions and guided paced breathing. The intervention lasted 5 weeks and consisted of about three daily sessions as short as 3 minutes, for an average of 100 sessions per athlete

Similarly, Deschodt-Arsac et al. (2018) used a 5 weeks protocol with two sessions per day of 5 minutes in duration

While most protocols, especially in the early days, consisted of supervised sessions in laboratory settings (Perry, Shaw, and Zaichkowsky, 2011; Beauchamp, Harvey, and Beauchamp, 2012; Pusenjak et al., 2015), there is a tendency in doing more home-based unsupervised practice in recent studies (Rijken et al., 2016; Perry, 2018; Mueller et al., 2019)

The only formally defined protocol for HRV Biofeedback (Lehrer, Vaschillo, and Vaschillo, 2000) includes both supervised sessions once per week and daily practice at home, and can therefore benefit in terms of practical applicability from recent advancements in technology (e.g. availability of smartphone apps for HRV Biofeedback)

Considering these factors, and the demands of modern-day life, small variations of the original protocol that require less time from the athlete (from 3 to 20 minutes instead of two sessions of 20 minutes per day) are currently more often considered by researchers and practitioners working in applied settings (Rijken et al., 2016; Perry, 2018; Deschodt-Arsac et al., 2018; Mueller et al., 2019)

To make  the most of your measurement, make sure to follow these guidelines for all your measurements:

• Make sure the flash turns on. If the flash doesn't turn on, the app won't be able to measure with high accuracy (which is the reason why you cannot measure using an iPad, unless you use a bluetooth sensor). If the flash doesn't turn on and you have an iPhone, go under your iPhone's settings (in your phone, not in the app) and make sure that HRV4Biofeedback is allowed to access your camera
• Make sure to cover the back camera completely. In some phones the flash might be quite far from the camera. Try to cover entirely the camera and also the flash as much as you can, using only one finger. Here you can find more instructions for iPhones with 3 cameras
• Make sure there is limited light (apart from the flash) in the room you are taking the measurement in. Measuring in dark places provides better measurement as natural light does not interfere with our algorithms, while we measure blood flow using light reflection. It should still be possible to measure accurately when there is sunlight, but in case of trouble, you might want to try a darker room
• If you use a phone cover, remove it when taking the measurement
• Place your hand on the camera before tapping the "Start session" button. The app will discard the first few seconds of data regardless, but placing the finger before starting the measurement can still be beneficial during the first few seconds
• Try not to move your finger and not to change the amount of pressure on the camera. Keep your finger still and in a comfortable position. This is the most important point. While we have techniques in place to cope with motion artifacts, movement will most likely mess up your data beyond what can be fixed
• Do not apply excessive pressure, which might impede blood flow and therefore prevent the app from measuring it correctly
• Do not use earphones while measuring (iPhone only). Unfortunately due to some incompatibilities with the techniques we use, it has been reported that the app might freeze if you measure using earphones on iPhones.

As first time users, try to take 2-3 measurements following the advice above, and you should be able to acquire high quality data rather easily.