Abstract
eart rate variability (HRV) is a non-invasive biomarker that reflects autonomic nervous system dynamics, providing valuable insights into physiological adaptation, stress, and recovery in athletes. Among the various HRV metrics, the root mean square of successive differences (RMSSD) has emerged as a robust and practical measure due to its strong association with parasympathetic activity, ease of calculation, and reliability in both short- and ultra-short-term recordings. This review examines the methodological considerations for using HRV to monitor training adaptations and recovery status in athletic populations. We highlight the superiority of routine, near-daily HRV measurements over isolated assess- ments, emphasizing the utility of weekly averages and the coefficient of variation (CV) to capture both chronic adaptations and acute homeostatic perturbations. Additionally, we discuss the selection of HRV devices, data recording procedures, and strategies to enhance athlete compliance. While RMSSD offers significant advantages for field-based monitoring, we also address its limitations, including its sole focus on parasympathetic activity and susceptibility to external confounders. Future directions include the integration of HRV data with other physiological markers and machine learning algorithms to optimize in- dividualized training and recovery strategies. This review provides sport scientists and practitioners with evidence-based recommendations to enhance the application of HRV in both research and real-world athletic settings. Keywords:HRV; autonomic modulation; wearable device; athlete performance; sport science 1. Introduction Heart rate variability (HRV) refers to the fluctuation
HRV data with other physiological markers and machine learning algorithms to optimize in- dividualized training and recovery strategies. This review provides sport scientists and practitioners with evidence-based recommendations to enhance the application of HRV in both research and real-world athletic settings. Keywords:HRV; autonomic modulation; wearable device; athlete performance; sport science 1. Introduction Heart rate variability (HRV) refers to the fluctuation in the time intervals between consecutive heartbeats, reflecting the dynamic interplay between the sympathetic and parasympathetic branches of the autonomic nervous system [1,2]. These fluctuations provide a non-invasive view into autonomic regulation, offering insights into physiological recovery, stress, and overall health. Accordingly, HRV has emerged as a foundational biomarker of cardiac autonomic modulation in both clinical and athletic contexts [2,3]. From a practical perspective, HRV serves as an objective physiological indicator that relates to how well the body is coping with physical stress. When measured at rest, high HRV indicates a predominance of parasympathetic activity, suggesting a relaxed and recovered state [4]. Conversely, low HRV reflects greater sympathetic activation or reduced vagal tone, often associated with stress or fatigue [4]. However, HRV responses Sensors2026,26, 3 https://doi.org/10.3390/s26010003
Sensors2026,26, 3 2 of 24 during prolonged or highly demanding activities can be more complex. While most work demonstrates a blunted HRV response under substantial physiological strain [5], other studies in military, tactical, and overreached athletic populations have reported paradoxical increases in HRV despite significant stress [6–8]. These findings emphasize that HRV should be interpreted within the broader training context and longitudinal trends, instead of isolated recordings. In sports contexts, a common approach to using HRV is through HRV-guided training, which involves making adjustments to a training session’s intensity based on waking HRV values [9,10]. While this model offers a practical method for day-to-day exercise prescription, HRV can also be used to evaluate broader physiological processes related to training adaptation and recovery [11–15]. Adaptation refers to the long-term autonomic and physiological adjustments that occur in response to repeated training stimuli [11], which are often reflected in gradual shits in weekly HRV trends [3]. In contrast, recovery relates to the short-term physiological response following physical stress and the body’s ability to restore homeostasis [12,14], which can be monitored through acute changes in HRV relative to an athlete’s individual baseline [3]. Recovery and adaptation are interdependent processes, as adequate recovery supports favorable adaptation, whereas impaired recovery may signal or contribute to maladaptation [13]. Thus, tracking HRV trends, particularly through weekly mean values and the coefficient of variation (CV), provides insight into both adaptation and recovery [3]. Rather than guiding single-day adjustments, these longitudinal patterns help monitor overall training status across weeks or training blocks and support decisions related to load management, performance readiness, and recovery strategies [14,15]. Over the past few decades, a variety of HRV metrics and devices have emerged, en- abling the monitoring of recovery and long-term adaptation in a variety of settings [16]. While this diversity may prove advantageous to consumers and researchers, it also high- lights potential pitfalls, particularly if selected procedures do not align with specified goals. Given this methodological heterogeneity, clearer guidance is needed regarding the appropriate use of mobile HRV tools. However, existing reviews on this topic have generally focused on laboratory-based HRV methods or
variety of settings [16]. While this diversity may prove advantageous to consumers and researchers, it also high- lights potential pitfalls, particularly if selected procedures do not align with specified goals. Given this methodological heterogeneity, clearer guidance is needed regarding the appropriate use of mobile HRV tools. However, existing reviews on this topic have generally focused on laboratory-based HRV methods or broad conceptual overviews, with limited attention to the unique methodological challenges associated with mobile plat- forms. Currently, there is no single resource that integrates device-specific considerations, ultra-short and weekly HRV metrics, and practical issues related to data interpretation and athlete compliance. Thus, a clear, integrated synthesis that consolidates device-specific considerations, ultra-short and weekly HRV metrics, and practical issues related to real- world data interpretation is still lacking. The aim of this narrative review is to address key methodological considerations for using mobile HRV devices in athlete monitoring and to provide practical recommendations for optimizing their application in assessing training adaptation and recovery. Topics discussed include the selection of HRV metrics and mobile devices, best practices for recording procedures, and strategies for interpreting HRV data in applied sport settings. 2. Methods This review was conducted as a narrative synthesis of conceptually relevant evidence to provide practical guidance for applied athlete monitoring via wearable HRV sensors and mobile technologies. The aim was to contextualize emerging practices, highlight methodological considerations, and connect empirical findings to real-world application rather than exhaustively summarize all available studies. The authors’ intentions were to align the review with established recommendations outlining when narrative reviews are appropriate within scientific scholarship [17–19]. https://doi.org/10.3390/s26010003
Sensors2026,26, 3 3 of 24 Literature searches were conducted between January and August 2024 using the databases and example search terms summarized in Table. Searches were further supple- mented by manual screening of reference lists from key empirical papers and HRV-focused reviews. No publication date limits applied to the search criteria. Table 1.Summary of literature search procedures for this narrative review. Databases Data Range Searched Example Search Terms Used PubMed Scopus Google Scholar January–August 2024 “heart rate variability”, “HRV”, “RMSSD”, “ECG”, “time-domain”, “frequency-domain”, “electrocardiography”, “PPG”, “photoplethysmography”, “smartphone”, “ultra-short-term”, “athlete”, “athletic”, “training”, “adaptation”, “recovery”, “training load”, “readiness”, “fatigue”, “overreaching”, “overtraining”, “monitoring”, “performance”, “exercise”, “mobile”, “wearable”, “sports”, “coefficient of variation”, “weekly HRV”, “training adaptation”, “autonomic modulation”, “vagal”, “parasympathetic”, and device-specific terms (e.g., “Oura”, “Whoop”, “HRV4Training”, “FitBit”, “Polar”, “Apple”) Study selection was guided by methodological rigor, ecological relevance to athlete monitoring, and applicability of HRV assessment using mobile devices. Evidence was synthesized qualitatively by prioritizing peer-reviewed studies that: evaluated HRV under resting conditions and in response to acute exercise; explored longitudinal HRV responses to training; validated ultra-short-term recordings and wearable devices; addressed method- ological considerations relevant to field-based monitoring; and had direct relevance to using HRV in applied sport settings. Studies that focused primarily on areas outside of the main purpose of this review, such as in the realms of clinical disorders or pharmacological interventions, were generally excluded unless they offered essential mechanistic insight. Because this was a narrative review, predefined inclusion/exclusion criteria and formal risk-of-bias scoring were not applied. The authors acknowledge the potential for selection bias inherent to narrative syntheses [17]. However, every effort was made to reflect a balanced cross-section of the most influential and methodologically sound work related to the use of mobile HRV devices and recording procedures for athletic contexts. 3. Choice of HRV Metric 3.1. The Common Parameters HRV analysis begins with the detection of R-R intervals, also referred to as inter- beat intervals, which represent the time between successive R-waves in the QRS complex of an electrocardiogram (ECG). While modern wearable technologies now allow for the recording of these intervals using validated mobile devices (discussed later
athletic contexts. 3. Choice of HRV Metric 3.1. The Common Parameters HRV analysis begins with the detection of R-R intervals, also referred to as inter- beat intervals, which represent the time between successive R-waves in the QRS complex of an electrocardiogram (ECG). While modern wearable technologies now allow for the recording of these intervals using validated mobile devices (discussed later in this paper), ECG remains the gold standard due to its superior temporal resolution and ability to precisely detect the fiducial point of each R-wave [2]. Once extracted, these R-R intervals form the basis for calculating the various HRV metrics, which are typically categorized into three primary domains: time-domain, frequency-domain, and non-linear measures. Table provides an overview of the most widely used HRV indices, outlining their physiological significance and implications for athlete monitoring. https://doi.org/10.3390/s26010003
Sensors2026,26, 3 4 of 24 Table 2.Common heart rate variability metrics. Please note that this is not an exhaustive list. For a comprehensive review of available HRV indices, see Reference [2]. HRV Metric Definition Physiological Significance and Practical Implications Time-Domain Parameters Derived from Statistical Analyses RMSSD Root Mean Square of Successive R-R Differences Parasympathetic marker; widely used and validated in athletes, especially in ultra-short-term mobile recordings, relatively easy to interpret SDNN Standard Deviation of Normal-to-Normal Intervals Reflects overall autonomic activity; usable in ultra-short-term mobile recordings, though links to training status remain unclear pNNx Proportion (%) of R-R Intervals That Differ by >x milliseconds (e.g., >50 ms for pNN50, etc.) Primarily parasympathetic; rarely used in athlete monitoring, with limited support for ultra-short-term use SDNN/RMSSD Ratio between SDNN and RMSSD Sympathovagal balance; higher values suggest sympathetic shift. Emerging as a complementary athlete monitoring metric Frequency-Domain Parameters Derived from Power Spectral Density HF Power High frequency power (0.15–0.40 Hz) Parasympathetic marker; typically requires ECG, specialized software, and≥3 min lab recordings, often normalized as HFnu LF Power Low frequency power (0.04–0.15 Hz) Controversial marker of mixed autonomic input; Requires ECG, specialized software, and≥3 min recordings, often normalized as LFnu LF/HF Ratio between LF and HF Sympathovagal balance, though debated; higher values = sympathetic shift; Requires ECG, specialized software, and≥3 min recordings Non-linear Parameters Derived from Poincare Plotting SD1 Short-term variability perpendicular to line of identity Captures rapid parasympathetic changes; may require lab setup and specialized software, interpretation may be challenging SD2 Long-term variability along the line of identity Represents slower autonomic changes (both branches); lab-based, requires specialized software, and can be difficult to interpret SD1/SD2 The ratio between SD1 and SD2 Purported marker of sympathovagal balance, though debated; may require lab setup, software, and can be difficult to interpret The selection of the appropriate HRV metric depends on the context of measurement and the goals of research or practice [1,2]. Time-domain metrics are often preferred in mobile settings due to their ease of computation and interpretation [3,20]. Frequency-domain metrics may offer richer insight into autonomic balance, especially when assessed under tightly controlled laboratory conditions
can be difficult to interpret The selection of the appropriate HRV metric depends on the context of measurement and the goals of research or practice [1,2]. Time-domain metrics are often preferred in mobile settings due to their ease of computation and interpretation [3,20]. Frequency-domain metrics may offer richer insight into autonomic balance, especially when assessed under tightly controlled laboratory conditions [2,21–23]. However, controversy exists concerning the physiological underpinning of the low-frequency domain, and the complexity and respiratory sensitivity of these metrics can limit field-based applications [23–25]. Non-linear HRV metrics evaluate the complexity and unpredictability of heart rate patterns, which can be advantageous during non-steady state assessments [26–29]. Notably, the Poincaré plot-derived SD1 and the time-domain metric RMSSD are mathematically equivalent and, hence, both reflect vagally mediated HRV [30]. In addition, because most HRV variables follow a non-normal, right-skewed distribution, it is standard practice to apply a natural logarithmic transformation (e.g., lnRMSSD) before statistical analysis to stabilize variance and meet assumptions for parametric testing [20]. 3.2. RMSSD as the Standard for Athletic Monitoring Among the available HRV metrics, RMSSD is arguably the most common for field- based monitoring [3]. Calculated as the root mean square of the successive R-R interval https://doi.org/10.3390/s26010003
Sensors2026,26, 3 5 of 24 differences, RMSSD primarily reflects parasympathetic (vagal) activity [31–33]. Numerous characteristics position RMSSD as the preferred HRV metric among athletes [32–38]. For instance, it is easier to compute and interpret than frequency-domain and non-linear metrics [1,4]. Furthermore, it remains relatively stable across a range of spontaneous breathing rates [39,40]. It reflects parasympathetic activity at rest and in response to acute exercise [41]. In addition, RMSSD demonstrates superior reliability and is less sensitive to external factors compared to frequency-domain metrics [42–44]. Therefore, the need for strict control during data collection is minimized with RMSSD. Perhaps its most advantageous characteristic related to athlete monitoring in mobile or field-based settings is the utility of RMSSD under ultra-short timeframes. Traditional standards for short-term HRV assessment recommend a 5 min stabilization period followed by a 5 min recording period [2], a protocol often considered impractical in dynamic sport environments. However, because RMSSD is less susceptible to respiratory variations and non-physiological influences on signal drift that often confound other HRV indices [1,45], it appears to be suitable when assessed under ultra-shortened recording durations. Indeed, several studies have shown that RMSSD maintains strong agreement with standard recording durations (i.e., 5 min stabilization preceding a 5 min recording) when derived from a 1 min segment that follows a 1 min stabilization period [31,46], and that this reliability holds irrespective of body position [47]. Similarly, Munoz et al. [48] compared multiple HRV indices across durations ranging from 10 s to 5 min and found RMSSD exhibited the strongest correlation (r > 0.90) with standard-length recordings, even at durations as short as 30 s. Further research has demonstrated that 1 min RMSSD recordings among athletes produced values nearly identical to conventional 5 min recordings across acute bouts of exercise [31] and long-term training periods [49]. 3.3. Section Summary Based on the collective evidence, RMSSD can be recommended as the primary HRV metric for field-based monitoring in athletes. Its strong association with parasympathetic activity and minimal sensitivity to respiratory fluctuations make it an ideal choice for tracking physiological trends over time. Additionally, RMSSD’s ease of calculation and
bouts of exercise [31] and long-term training periods [49]. 3.3. Section Summary Based on the collective evidence, RMSSD can be recommended as the primary HRV metric for field-based monitoring in athletes. Its strong association with parasympathetic activity and minimal sensitivity to respiratory fluctuations make it an ideal choice for tracking physiological trends over time. Additionally, RMSSD’s ease of calculation and its accuracy in ultra-short-term recordings across various body positions and training conditions enhances its practicality in real-world athletic settings. Therefore, this metric appears to be an asset for athletes, coaches, and practitioners seeking valid physiological feedback with minimal disruption to training or competition routines. However, it should be noted that RMSSD is not without limitations, and several practical and physiological considerations relevant to its interpretation are outlined in Section. 4. Choice of Mobile HRV Device As previously mentioned, HRV is most accurately measured by collecting heart rate data via ECG and calculating it using specialized acquisition software. However, this approach is practically limited, requiring the need for laboratory equipment, specialized expertise, and controlled testing conditions, making it inconvenient for routine HRV moni- toring in field settings. Although mobile-based HRV devices have gained popularity due to their convenience and accessibility, a key concern relates to their accuracy relative to ECG. 4.1. Accuracy of Various Mobile HRV Devices A meta-analysis by Dobbs et al. [16] compared HRV data from 23 studies using portable sensors, including chest straps, wrist-worn monitors, and smartphone-based photoplethys- mography (PPG) applications, and found small but acceptable discrepancies relative to ECG for all devices. Among the available HRV metrics, RMSSD consistently demonstrated https://doi.org/10.3390/s26010003
Sensors2026,26, 3 6 of 24 the lowest error across devices, regardless of sensor type or body position [16]. Subsequent validation studies have similarly shown strong agreement between ECG and mobile HRV devices, including smartphone cameras [50–52], wristwatches [53,54], rings [55], chest straps [56], and ambulatory blood pressure devices [57]. Accuracy may also depend on recording duration, as a one-minute measurement is generally sufficient for RMSSD [16], whereas frequency-domain parameters typically require at least three-to-five minutes [51]. Furthermore, PPG-based devices perform best under resting conditions but lose accuracy during or immediately after exercise because of motion artefacts and vascular changes [58–60]. This is likely due to PPG and ECG being physiologically distinct signals. PPG reflects peripheral pulse dynamics rather than cardiac electrical activity [61]. Recent work suggests that these fundamental differences may introduce systematic variability between modalities, particularly during exercise, when vascular and pulse transit time changes can distort PPG signals relative to ECG [61]. Therefore, mobile devices appear to be suitable for measuring resting HRV, particularly for ultra-short RMSSD. However, their use during or immediately following exercise should be approached with caution due to the potential for reduced signal accuracy. Nevertheless, when applied under appropriate conditions, mobile HRV devices provide a practical and valid means of monitoring autonomic trends in athletic settings. 4.2. Real-World Practicality of Different Mobile HRV Devices Beyond accuracy, practical considerations such as comfort, convenience, and cost may influence device selection. Wrist-worn and ring-based devices allow continuous data collection and are particularly effective for nocturnal recordings [54,62–64]. Many also provide additional health-related metrics such as heart rate, sleep, and oxygen saturation. Chest strap monitors yield highly accurate ECG-like signals [65] but require proper place- ment/contact and removal, which some users may find inconvenient and not practical for long-term continuous recordings. Wrist, ring, and chest-worn devices also tend to be more expensive than smartphone-based alternatives. In contrast, smartphone applications that rely on PPG, either through a connected finger sensor [66] or the phone’s camera [58,67], are inexpensive, portable, and ideal for consistent morning recordings [68,69]. These applica- tions may also enhance athlete compliance because they are low-cost, easy
long-term continuous recordings. Wrist, ring, and chest-worn devices also tend to be more expensive than smartphone-based alternatives. In contrast, smartphone applications that rely on PPG, either through a connected finger sensor [66] or the phone’s camera [58,67], are inexpensive, portable, and ideal for consistent morning recordings [68,69]. These applica- tions may also enhance athlete compliance because they are low-cost, easy to use, and often integrate wellness questionnaires, providing a practical solution for tracking recovery and adaptation through daily HRV measures. Ultimately, the optimal device depends on the athlete’s needs, monitoring frequency, and preferred measurement context. Table rizes commonly used mobile HRV devices, including example manufacturers, advantages, limitations, and validation findings. It is also important to note that many commercially available HRV mobile devices (e.g., smartphone applications, wearables, etc.) automatically apply a log transformation to RMSSD. However, because lnRMSSD values are not intuitively interpretable by consumers, several systems multiply the transformed value by a constant (such as 20) to rescale the output to a more user-friendly format [30,66]. For example, a raw lnRMSSD value of 4.2 mightbe presented as an HRV “score” of 84. While these adjusted, arbitrary unit values are useful for tracking trends, practitioners and consumers should understand how such scores are derived to ensure consistency when interpreting HRV data across different platforms. In addition, because transformed scores do not retain the original time-domain scale, they may not be directly comparable to raw RMSSD values in milliseconds. https://doi.org/10.3390/s26010003
Description
This review examines HRV monitoring methods for training adaptation and recovery in athletes.