Time-Domain Measurements of Heart Rate Variability

Normative HRV values only apply when clients breathe at average rates (Shaffer & Ginsberg, 2017).

BCIA Blueprint Coverage

This unit addresses IV. HRV Measurements: A. Time-domain measurements and their meaning, properties, and correlates.

Professionals completing this unit will be able to discuss the following HRV time-domain indices: SDNN, SDRR, SDANN, pNN50, NN50, HR Max - HR Min, RMSSD, CVRR, HRV-CV, and HRV triangular index.

🎧 Chapter Lecture: Time-Domain Measurements

How Time-Domain Indices Work

This section introduces the fundamental building block of time-domain HRV: the interbeat interval. Understanding how these intervals are measured and quantified is essential before exploring specific metrics. HRV time-domain indices quantify variability in the interbeat interval (IBI), which is the time between the peaks of successive R-spikes — the initial upward deflections in the QRS complex. Because the IBI is measured from one R-peak to the next, it is also called the R-R interval.

Software detects each heartbeat and measures the time to the next beat in milliseconds (ms). This process begins after the first detected beat and repeats continuously until the end of the epoch, which is the data collection period. The resulting series of IBIs forms the raw dataset from which all time-domain metrics are calculated.

Time-Domain vs. Frequency-Domain: Two Complementary Approaches

While time-domain indices measure how much interbeat intervals vary, frequency-domain measurements take a different approach: they decompose the heart rate signal into its underlying rhythmic components. Frequency-domain methods calculate the absolute or relative amount of signal energy, called power, in specific frequency bands — ultra-low frequency (ULF), very-low frequency (VLF), low frequency (LF), and high frequency (HF). These two approaches are complementary, and you will encounter correlations between them throughout this chapter.

SDNN: The Standard Deviation of Normal Heartbeat Intervals

This section covers the most widely cited time-domain metric and its role in cardiac risk assessment. The SDNN is the standard deviation of interbeat intervals for normal sinus beats, measured in milliseconds. The word "normal" here is critical: it means that abnormal beats, such as ectopic beats originating outside the sinoatrial node, have been removed through a process called artifacting. A related metric, the SDSD (standard deviation of successive RR interval differences), captures only short-term variability.

SDNN is calculated using data that are free of artifacts and abnormal heartbeats.

Both the sympathetic nervous system (SNS) and the parasympathetic nervous system (PNS) contribute to SDNN, and the metric is highly correlated with ULF, VLF, and LF band power, as well as total power (Umetani et al., 1998). In short-term resting recordings (≤ 5 minutes), the primary source of variation is parasympathetically-mediated respiratory sinus arrhythmia (RSA) — the natural acceleration and deceleration of heart rate with each breath — especially during slow-paced breathing protocols (Shaffer, McCraty, & Zerr, 2014). In 24-hour recordings, LF band power contributes more substantially to SDNN (Kusela, 2013). This distinction matters in practice: a 5-minute SDNN primarily reflects vagal tone, whereas a 24-hour SDNN captures the full range of autonomic influences.

The following table shows the correlations between time- and frequency-domain measures in 24-hour recordings and is provided courtesy of the Institute of HeartMath (Shaffer, McCraty, & Zerr, 2014).

Why 24-Hour Recording is More Accurate

The SDNN is more accurate when calculated over 24 hours than during the shorter periods typically used in biofeedback sessions. Extended recordings capture the heart's responses to a far wider range of demands, including changing workloads, anticipatory central nervous system activity involving classical conditioning, and circadian processes like sleep-wake cycles (Lehrer, 2012). In clinical terms, a 24-hour recording reveals contributions from the SNS that are largely invisible in a 5-minute sample (Grant et al., 2011).

While the conventional short-term recording standard is 5 minutes, researchers have proposed ultra-short-term (UST) recording periods ranging from 60 seconds (Salahuddin et al., 2007; Shaffer, Meehan, & Zerr, 2020) to 240 seconds (Baek et al., 2015). These abbreviated windows are increasingly relevant as wearable devices enter clinical settings, but they come with trade-offs in reliability that you will encounter later in this chapter.

24-Hour SDNN Predicts Mortality

The SDNN is the "gold standard" for medical stratification of cardiac risk when recorded over 24 hours (Task Force, 1996). Based on 24-hour monitoring, patients with SDNN values below 50 ms are classified as unhealthy, those between 50 and 100 ms have compromised health, and those above 100 ms are classified as healthy. These thresholds have direct implications for how you triage and prioritize your clients. The Task Force (1996) also established canonical normative 24-hour values for healthy adults: SDNN of approximately 141 ± 39 ms, SDANN of approximately 127 ± 35 ms, and HRV triangular index of approximately 37 ± 15. These values provide the reference benchmark against which clinical populations are compared.

A critical practical point is that SDNN is mathematically and physiologically tied to recording length and nonstationarity — the tendency of a signal's statistical properties to shift over time. A 5-minute SDNN and a 24-hour SDNN are not "more or less the same thing"; they are different summaries of different mixtures of biology, including respiratory sinus arrhythmia, baroreflex activity, thermoregulation, circadian effects, and activity patterns. This is why the standards explicitly warn that comparing HRV indices from time series of different lengths is inappropriate (Task Force, 1996).

Heart attack survivors whose 24-hour measurements placed them in a higher category had a greater probability of living during a 31-month mean follow-up period. Patients with SDNN values over 100 ms had 5.3 times lower mortality risk at follow-up than those under 50 ms (Kleiger et al., 1987). This landmark finding raises a compelling clinical question: could training patients to increase SDNN to a higher category reduce their mortality risk?

A Firstbeat Bodyguard 2, which is designed for ambulatory 24-hour monitoring, is shown below.

SDRR: Standard Deviation Including All Beats

This section covers a metric that is closely related to SDNN but includes a critical difference. The SDRR is the standard deviation of interbeat intervals for all sinus beats, including abnormal or ectopic beats — heartbeats originating outside the sinoatrial node — and artifacts, measured in milliseconds. Unlike SDNN, which excludes artifacts, the SDRR retains these irregular beats in the calculation. Like the SDNN, the SDRR is more accurate when calculated over 24 hours.

The inclusion of abnormal beats matters clinically because these beats may reflect underlying cardiac dysfunction or sensor noise that masquerades as genuine HRV. When you see a notable discrepancy between a client's SDRR and SDNN values, it suggests the presence of artifacts or ectopic beats that warrant further investigation. In practice, the SDNN is generally preferred precisely because it filters out these confounding signals. In a clean sinus-rhythm record, SDRR and SDNN will converge; as soon as ectopy or artifact enters, SDRR can increase for reasons that have nothing to do with autonomic modulation. That sensitivity is useful if your question is "how irregular is this rhythm?" but becomes a liability when you want to track autonomic control — one reason contemporary standards lean heavily on NN-based metrics for autonomic interpretation (Sammito et al., 2024; Task Force, 1996).

Below is a BioGraph® Infiniti heart rate variability display. The roller coaster accelerates as SDRR increases.

SDANN: Standard Deviation of Average 5-Minute Segments

This section introduces a 24-hour metric that captures slow-changing rhythms in heart rate. The SDANN is the standard deviation of the average NN intervals (normal-to-normal intervals — "clean" IBIs calculated after artifacting) for each of the 5-minute segments during a 24-hour recording. By averaging within each 5-minute window first, the SDANN estimates heart rate changes produced by cycles longer than 5 minutes, such as circadian rhythms and physical activity patterns.

Like the SDNN, the SDANN is measured and reported in milliseconds. It correlates strongly with SDNN and is generally considered redundant with it (Shaffer, McCraty, & Zerr, 2014). Minimum heart rate is more strongly associated with Ln SDANN (where Ln means natural logarithm) than with Ln RMSSD, while maximum heart rate is weakly and inconsistently correlated with these time-domain measures (Burr et al., 2006). For most clinical and biofeedback applications, the SDNN provides the same information.

SDNN Index (SDNNI): Average of 5-Minute Standard Deviations

This section covers a metric that complements the SDANN by focusing on variability within rather than across 5-minute windows. The SDNN Index (SDNNI) is the mean of the standard deviations of all NN intervals for each 5-minute segment of a 24-hour recording. It is calculated by dividing the 24-hour record into 288 five-minute segments, computing the standard deviation within each segment, and then averaging those 288 values.

Because the SDNNI captures variability within short windows, it primarily reflects autonomic influences on heart rate rather than the slower circadian and behavioral rhythms captured by SDANN. It correlates with VLF power over 24 hours (Shaffer, McCraty, & Zerr, 2014). Think of SDANN and SDNNI as complementary lenses: SDANN reveals how your client's average heart rate shifts across hours, while SDNNI reveals how variable each 5-minute window is on its own.

Age exerts a strong influence on all three Holter-derived metrics. In a large normative sample spanning ages 10–99, SDNN and SDANN declined gradually across decades, while the SDNN Index showed a more linear decrease across the lifespan (Umetani et al., 1998). This pattern has a direct practical implication: these are not static traits but age- and timescale-dependent summaries, and a "low" score in an older adult may be developmentally normal rather than clinically concerning. Applying young-adult norms to clients over 60 can lead to unnecessary alarm.

NN50 and pNN50: Counting Large Beat-to-Beat Changes

This section covers two related metrics that quantify abrupt shifts between consecutive heartbeats and link directly to parasympathetic activity. The NN50 measures the number of adjacent NN intervals that differ by more than 50 milliseconds. At least a 2-minute sample is required. The pNN50 expresses this count as a percentage of total NN intervals, making it easier to compare across recordings of different lengths.

While the conventional minimum recording for pNN50 is 5 minutes, researchers have proposed UST periods of 10 seconds (Salahuddin et al., 2007), 30 seconds (Baek et al., 2015), and 60 seconds (Shaffer, Meehan, & Zerr, 2020). The pNN50 is closely correlated with PNS activity (Umetani et al., 1998) and also tracks with both RMSSD and HF power. However, the RMSSD typically provides a better assessment of RSA, especially in older adults, and most researchers prefer it over pNN50 (Otzenberger et al., 1998).

A practical limitation of pNN50 is the floor effect that emerges with aging and elevated heart rates: as the cardiac cycle shortens, successive intervals rarely differ by more than 50 ms, so the threshold becomes increasingly difficult to cross and the metric collapses toward zero even in physiologically healthy individuals.

Normative data from healthy adults using 5-minute recordings illustrate this clearly: pNN50 averaged approximately 0.17 ± 0.18 in women and 0.15 ± 0.16 in men aged 25–49, but fell to only about 0.05 ± 0.09 in women and 0.04 ± 0.07 in men aged 50–74 (Voss et al., 2015). Note that pNN50 is sometimes reported as a proportion (0–1) and sometimes as a percent (0–100%), so values of 0.17 and 17% are equivalent. This fourfold drop across age groups means fixed cutpoints can over-label healthy older adults as "abnormally low" if age is not taken into account (Umetani et al., 1998).

Despite its limitations relative to RMSSD, pNN50 may be a more reliable index than short-term SDNN for the brief samples commonly used in biofeedback sessions. This makes it a practical backup metric when RMSSD is unavailable in your software platform. In summary, NN50 and pNN50 capture parasympathetic activity through a simple threshold-based approach, but RMSSD offers superior sensitivity, particularly with aging clients.

HR Max - HR Min: The Most Client-Friendly Metric

This section covers the time-domain metric that is easiest to explain to clients and most intuitive for real-time biofeedback displays. HR Max – HR Min is the average difference between the highest and lowest heart rates during each respiratory cycle. Where RMSSD and SDNN require explaining standard deviations and root mean squares, HR Max – HR Min simply measures a wave's height — a concept clients grasp immediately (Moss, 2022). At least a 2-minute sample is required, and physically active individuals typically show wider peak-trough differences than those who are sedentary.

HR Max - HR Min is Affected by Breathing Rate and Measures RSA

An important nuance for practitioners is that this index is susceptible to the effects of respiration rate, independent of vagus nerve traffic. Rather than directly indexing vagal tone, HR Max – HR Min reflects RSA. This distinction matters clinically: since longer exhalations allow greater acetylcholine metabolism, slower respiration rates can produce higher RSA amplitudes that are not mediated by changes in vagal firing (Lehrer, 2012). In other words, a higher score during slow-paced breathing does not necessarily mean the vagus nerve is more active — it may simply reflect the biomechanics of slower breathing.

Booiman (2017) reported values in the 30- and 40-bpm range for Dutch clients in their teens and twenties during slow-paced breathing. For example, the screen capture below is from a 16-year-old female client, 2 weeks post-concussion, who achieved a HR Max – HR Min value of 30 bpm while breathing at 5.5 breaths per minute. HR Max – HR Min can reach 50 beats per minute for elite athletes, making it a useful benchmark in optimal performance work with military personnel and competitive athletes.

This measure is widely used for HRV assessment during paced breathing protocols and is highly correlated with both SDNN and RMSSD (Shaffer, McCraty, & Zerr, 2014). Its accessibility makes it especially valuable in settings where client engagement is a priority, such as VA clinics working with veterans who may be new to biofeedback concepts.

However, practitioners should recognize that HR Max–HR Min is fundamentally a breathing-pattern metric: it measures how the person is breathing as much as — or sometimes more than — how the autonomic nervous system is behaving as a stable trait. This fragility for normative interpretation is the same reason modern guidelines insist on keeping posture, time of day, activity state, and breathing protocol consistent across sessions and clients before drawing any clinical conclusions from this index (Sammito et al., 2024; Task Force, 1996).

RMSSD: The Best Short-Term HRV Measure

This section covers the single most important metric for biofeedback practitioners working with short-term recordings. The RMSSD — the root mean square of successive differences between normal heartbeats — is calculated by first measuring the time difference between each pair of adjacent interbeat intervals in milliseconds, then squaring each difference, averaging the squared values, and finally taking the square root. The result captures rapid beat-to-beat variance in heart rate and provides a better estimate of vagal activity than SDNN (Shaffer, McCraty, & Zerr, 2014).

The RMSSD is conceptualized as vagally-mediated HRV (vmHRV), meaning it predominantly reflects parasympathetic influence on the heart (Jarczok et al., 2021). It is the best overall measure of short-term HRV because it is less affected by outliers and artifacts than SDNN (Gevirtz, 2020). At least a 5-minute sample is required, though researchers have proposed UST periods of 10 seconds (Salahuddin et al., 2007), 30 seconds (Baek et al., 2015), and 60 seconds (Shaffer, Meehan, & Zerr, 2020).

Clinical Significance and Cross-Domain Connections

The RMSSD is mathematically identical to the nonlinear metric SD1, which reflects short-term HRV on a Poincaré plot (Ciccone et al., 2017). This equivalence means that when you see SD1 in a nonlinear analysis, you are looking at the same information as RMSSD expressed in a different visual format.

Because RMSSD values are right-skewed across populations — most individuals cluster at lower values while a smaller number show very high variability — researchers and apps frequently apply a natural-log transformation, yielding lnRMSSD (the natural logarithm of RMSSD). This transformation normalizes the distribution and makes statistical comparisons more stable, which is why lnRMSSD appears in performance-monitoring literature alongside the raw ms value (Besson et al., 2025; Ciccone et al., 2017).

While the RMSSD correlates with HF power (Kleiger et al., 2005), the influence of respiration rate on this index remains uncertain (Penttilä et al., 2001; Schipke et al., 1999). Notably, the RMSSD is less affected by respiration than RSA across several tasks (Hill & Siebenbrock, 2009).

Normative data from a large healthy cohort using 5-minute recordings help contextualize values you see in clinical practice. Mean RMSSD was approximately 36.5 ± 20.1 ms in women and 34.0 ± 18.3 ms in men aged 25–49, falling to roughly 22.0 ± 13.2 ms in women and 20.5 ± 11.0 ms in men aged 50–74 (Voss et al., 2015).

This age-related decline is consistent across studies and underscores the importance of age-matched interpretation: a 65-year-old client with an RMSSD of 20 ms may be entirely within normal limits. Breathing protocols also shift these values substantially. A systematic review found mean 5-minute SDNN of approximately 49 ms and RMSSD of approximately 43 ms under free breathing, compared with SDNN of approximately 59 ms and RMSSD of approximately 55 ms under paced breathing — a quiet reminder that "normal values" are conditional on how the person is breathing (Nunan et al., 2010).

The RMSSD has important prognostic value beyond biofeedback. Lower RMSSD values are correlated with higher scores on a risk inventory for sudden unexplained death in epilepsy (DeGiorgio et al., 2010). A novel ratio combining short-term RMSSD with C-reactive protein (an inflammatory marker) predicted survival in both cancer patients and the general population (Jarczok et al., 2021). These findings underscore how a metric you use in daily biofeedback sessions connects to life-and-death clinical outcomes.

Recent reliability research adds an important measurement-context dimension. In a 2025 study evaluating RMSSD across environments and body positions, RMSSD showed good-to-excellent reliability in supine morning measurements, with intraclass correlation coefficients (ICCs) above 0.75 — a commonly used benchmark for acceptable clinical reliability.

Standing measurements were generally less stable, and the absolute variability for RMSSD in the supine position was on the order of 6–8 ms. The practical takeaway: when you want to track RMSSD repeatedly over time, standardize posture and time of day before worrying about recording length (Besson et al., 2025).

The RMSSD in Consumer Technology

Many popular HRV apps — including Apple Health, Elite HRV, Fitbit, Optimal HRV, and the Oura Ring— use RMSSD or lnRMSSD (its natural logarithm) as their primary HRV metric. This means your clients are increasingly likely to arrive at sessions with wearable-generated RMSSD data in hand. Understanding how these consumer measurements compare to clinical-grade recordings is essential for guiding clients who ask whether their smartwatch data is "accurate."

HRV-CV: Measuring the Stability of Heart Rate Variability Over Time

This section covers a metric that shifts the focus from a single recording to the consistency of HRV across days — an approach increasingly central to sport science and consumer health monitoring. The coefficient of variation (CV) is a standard statistical measure that expresses a standard deviation as a percentage of its mean. When applied to HRV, the CV appears in two distinct forms. In the first form, called the CVRR, it normalizes the variability of RR intervals within a single recording by dividing SDNN by the mean RR interval and multiplying by 100. This produces a heart-rate-corrected index of autonomic function expressed as a percentage, making it easier to compare individuals who have different resting heart rates (Shaffer & Ginsberg, 2017). In the second and more recent form, called HRV-CV, the CV is calculated across a series of daily HRV scores — most commonly the weekly mean and standard deviation of lnRMSSD — to quantify how much a person's autonomic activity fluctuates from day to day (Plews et al., 2012).

CVRR: A Heart-Rate-Corrected Window Into Autonomic Function

The CVRR addresses a practical problem with raw SDNN: because longer RR intervals (slower heart rates) tend to produce larger absolute variability, two individuals with identical autonomic function but different resting heart rates can show different SDNN values. Dividing by the mean RR interval removes this heart-rate dependency and yields a dimensionless percentage (Shaffer & Ginsberg, 2017). Typical CVRR values in healthy young adults at supine rest range from approximately 4% to 7%, declining with age in both sexes. In patients with essential hypertension, CVRR is significantly lower than in age-matched normotensive controls, suggesting impaired parasympathetic function.

The CVRR has proven especially valuable in detecting cardiovascular autonomic neuropathy (CAN), a complication of diabetes that can cause resting tachycardia, orthostatic hypotension, and sudden cardiac death. In young patients with type 1 diabetes, CVRR during deep breathing showed sensitivity and specificity exceeding 96% for detecting CAN, outperforming most other single time-domain measures (Razanskaite-Virbickiene et al., 2017). Decreased CVRR at rest is also associated with higher resting heart rate, prolonged QTc interval, and greater orthostatic blood pressure decline in non-elderly patients with diabetes (Sugimoto et al., 2024). Beyond diabetes, the CVRR is reduced in neurological conditions including Parkinson’s disease and dementia with Lewy bodies, where it serves as a convenient, noninvasive marker of autonomic deterioration. In elderly long-term care residents, higher CVRR predicted survival more effectively than standard annual health examinations (Kurita et al., 2013).

HRV-CV: Day-to-Day Variability as a Training and Health Signal

The day-to-day HRV-CV emerged from sport science as researchers recognized that a single morning HRV reading can fluctuate for many reasons — poor sleep, alcohol, stress, or simply measurement noise — but the pattern of fluctuation across a week tells a more meaningful story. Plews and colleagues (2012) first demonstrated this concept in a landmark case comparison of two elite triathletes. One athlete, who became non-functionally over-reached, showed a progressive decline in both the 7-day rolling average of lnRMSSD and its coefficient of variation, while the healthy control athlete maintained stable values. This observation launched a productive line of research showing that the weekly CV of lnRMSSD captures autonomic perturbations that the weekly mean alone may miss. Subsequent team-sport studies reinforced the pattern: national-level swimmers displayed more stable 7-day HRV-CV than their conference-level teammates during standardized training (Flatt et al., 2022), and rugby and futsal players with lower HRV-CV during training blocks showed superior gains in aerobic capacity (Flatt & Howells, 2019; Nakamura et al., 2020). These early findings were compelling but limited by small, homogeneous athletic samples and a narrow focus on human performance outcomes.

The HRV-CV is calculated by dividing the standard deviation of daily lnRMSSD values over a defined period (typically 7 days) by the weekly mean and multiplying by 100. Typical values range from roughly 2% to 20%, with lower values indicating more stable autonomic recovery and higher values signaling greater day-to-day perturbation (Flatt & Esco, 2016; Grosicki et al., 2026). A decrease in the weekly CV of lnRMSSD during training is generally a favorable sign: in collegiate female soccer players, the change in lnRMSSD_CV over the first three weeks of a conditioning program correlated strongly (r = −0.74) with eventual improvement in intermittent running performance (Flatt & Esco, 2016). Conversely, athletes with the lowest CV during increased training loads showed the most favorable performance gains, suggesting that physiological resilience manifests as stable daily HRV (Plews et al., 2013).

Importantly, lower HRV-CV is not universally advantageous. The vagal tank theory proposes that cardiac vagal control comprises three domains — resting (tonic), reactivity (phasic), and recovery (phasic) — and that healthy autonomic regulation sometimes requires greater HRV fluctuation (Laborde et al., 2018). During exercise, larger vagal withdrawal supports improved cardiovascular performance, meaning that some day-to-day HRV variability reflects appropriate physiological engagement rather than maladaptive instability. Very low HRV-CV could also signal blunted autonomic reactivity in individuals with limited physiological responsiveness, such as older adults with diminished vagal reserve, or a lack of autonomic challenge in those leading highly sedentary lifestyles (Grosicki et al., 2026). Both high and low HRV-CV must therefore be interpreted in context, ideally alongside absolute HRV trends (Plews et al., 2012).

Consumer Applications and Emerging Population-Level Evidence

Several consumer platforms now incorporate HRV-CV or closely related trend metrics. HRV4Training displays the coefficient of variation alongside weekly baseline lnRMSSD, allowing athletes to identify periods of autonomic instability before the baseline itself changes. Elite HRV calculates a rolling 7-day CV of morning readiness scores. WHOOP has conducted large-scale research on sleep-derived HRV-CV, although the metric is not yet displayed directly in the app. The Oura Ring provides data that third-party platforms like Heads Up Health use to compute HRV-CV trends.

In the largest population study to date, Grosicki and colleagues (2026) analyzed approximately 2 million nocturnal HRV readings from over 21,000 wearable device users and established that at least five nights of data are required for reliable 7-day HRV-CV estimates (ICC ≥ 0.80). Reliability held across all day-of-week combinations as long as five days were included, although four-day samples that excluded weekend nights fell below acceptable thresholds — consistent with evidence that sleep and activity patterns shift systematically on weekends. Higher HRV-CV was associated with greater alcohol consumption, lower physical activity, shorter and less consistent sleep, older age, and higher body mass index. Notably, alcohol consumption and sleep patterns showed stronger standardized associations with HRV-CV than with HRV or resting heart rate, suggesting that day-to-day fluctuation captures behavioral influences on autonomic function that absolute HRV values may underrepresent.

The study also revealed distinct age trajectories by biological sex that persisted after adjusting for health behaviors, body mass index, and HRV. In males, HRV-CV remained stable through early adulthood but rose progressively after approximately age 40 — a pattern the authors linked to the typical midlife decline in testosterone. In females, HRV-CV followed a U-shaped trajectory, declining through midlife and increasing after roughly age 50 in a pattern broadly consistent with the menopausal transition. HRV-CV also increased with higher body mass index in both sexes, with stronger effects in males even after controlling for behavioral factors. These findings support HRV-CV as a behavior-sensitive digital biomarker with applications extending beyond athletic performance into population health monitoring and risk stratification.

Strengths and Limitations

The CVRR’s principal strength is its heart-rate correction, which allows fairer comparisons across individuals with different resting rates and makes it particularly sensitive to early autonomic neuropathy. Its limitation is that no single normative threshold has been universally adopted; reference values vary by age, posture, and breathing protocol, and the metric is not included in the 1996 Task Force standards. The day-to-day HRV-CV adds a longitudinal dimension that captures recovery dynamics invisible to any single-session metric. However, it requires consistent daily measurement under standardized conditions — same time of day, same posture, same recording duration — because protocol variation inflates the CV for reasons unrelated to autonomic function.

Both forms of the CV share a mathematical sensitivity: when the mean is small, even modest fluctuations produce disproportionately large CV values, which can exaggerate instability in individuals with low baseline HRV. Practitioners should also remember that HRV-CV is a trend metric — it tells you how stable your client’s autonomic profile is, not why it is unstable. A rising HRV-CV in an athlete may reflect training overload, poor sleep, alcohol use, illness onset, or simply a change in measurement routine. As with all HRV indices, context is everything.

HRV Triangular Index (HRVi): A Geometric Approach

This section introduces two geometric measures that take a different approach to quantifying HRV — using histogram shapes rather than statistical calculations. The HRV triangular index (HRVi) is a geometric measure based on 24-hour recordings. It is calculated by dividing the integral (total area) of the RR interval histogram's density by its maximum height, essentially measuring how spread out the distribution of interbeat intervals is (Task Force, 1996). A broader distribution produces a higher HTI, reflecting greater overall variability. HRVi is a dimensionless index tied to histogram peak and total counts.

Both the PNS and SNS contribute to the HRV triangular index (Billman et al., 1982; Schwartz et al., 1988). A 5-minute epoch may be sufficient to represent this metric (Jovic & Bogunovic, 2011), and a 120-second UST period has been proposed (Shaffer, Meehan, & Zerr, 2020). However, the original standards explicitly state that geometric methods require approximately 20 minutes as a minimum, with 24 hours preferred for reliable histogram formation — because the NN interval distribution needs sufficient data points before its shape becomes stable and interpretable (Task Force, 1996).

The canonical normative 24-hour value for healthy adults is approximately 37 ± 15 (Task Force, 1996). A particularly valuable clinical application is that HRVi and RMSSD can jointly distinguish between normal heart rhythms and arrhythmias: when HRVi ≤ 20.42 and RMSSD ≤ 0.068, the rhythm is normal, but when HRVi exceeds 20.42, the rhythm is arrhythmic (Jovic & Bogunovic, 2011).

The HRVi independently predicts cardiovascular mortality in patients diagnosed with atrial fibrillation (Hämmerle et al., 2020). For practitioners working with cardiac populations in hospital or VA settings, this makes the HRVi a useful screening tool that complements SDNN-based risk stratification.

The Triangular Interpolation of the NN Interval Histogram (TINN)

The Triangular Interpolation of the NN Interval Histogram (TINN) is the baseline width of a histogram displaying NN intervals. To visualize this metric, imagine a histogram where the X-axis represents interbeat interval length in milliseconds and the Y-axis represents the count of intervals at each length (Yilmaz et al., 2018). The TINN measures how wide this distribution is at its base — a wider base indicates more variability across the recording. The width is the distance between the left and right base intercepts of the fitted triangle. At least a 5-minute sample is required (Shaffer & Ginsberg, 2017).

Summary Tables

The following tables consolidate the key properties of time-domain indices for quick reference. Table 1 summarizes each metric's definition and clinical significance. Table 2 shows the minimum conventional and UST recording periods recommended for each metric (Shaffer, Meehan, & Zerr, 2020). Table 3 displays an example of Kubios time-domain calculations before and after artifact correction, illustrating how these values appear in commonly used HRV analysis software.

HRV Myths: Avoiding Common Misinterpretations

This section addresses three widespread misconceptions that can lead to clinical errors. Even experienced practitioners sometimes misapply HRV norms, and these mistakes can distort clinical decision-making. Understanding these pitfalls will protect you from overstating or understating your clients' autonomic health.

Myth 1: Short-Term and 24-Hour Measurements Are Interchangeable

You cannot interpret short-term metrics using 24-hour norms because they are obtained under fundamentally different conditions. A 24-hour recording captures circadian rhythms, sleep-wake transitions, physical activity, and emotional responses throughout the day. Briefer recording periods miss these slower sources of variability and therefore generally produce lower HRV values. Applying 24-hour cutoffs to a 5-minute recording will inevitably overestimate your client's cardiac risk.

Myth 2: Slow-Paced Breathing Values Can Be Compared with Resting Norms

Resting norms are established with participants breathing at typical rates, usually 12 to 14 breaths per minute. Slow-paced breathing at 6 breaths per minute — less than half the typical rate — amplifies RSA and inflates time-domain values well beyond resting baselines. Comparing slow-paced breathing data to resting norms would dramatically overestimate a client's baseline autonomic flexibility.

The magnitude of this inflation is not trivial: a systematic review of normal values found mean 5-minute RMSSD of approximately 43 ms under free breathing versus approximately 55 ms under paced breathing, and mean SDNN of approximately 49 ms versus approximately 59 ms, respectively (Nunan et al., 2010). In practical terms, a client who achieves an RMSSD of 55 ms during paced breathing is not demonstrating an above-average resting autonomic profile; they are demonstrating a normal response to a known HRV-amplifying procedure. Always match the norm to the breathing context.

Myth 3: UST and Short-Term Measurements Are Interchangeable

UST measurements are more vulnerable to corruption by artifact because they are based on fewer data points. Currently, there is no consensus on an acceptable UST measurement length for each metric. As Shaffer, Meehan, and Zerr (2020) put it, "UST measurements are proxies of proxies. They seek to replace short-term values, which, in turn, attempt to estimate reference standard long-term metrics." When your clients bring wearable data based on 30- or 60-second recordings, this context is essential for setting realistic expectations.

Test Your Understanding

A clinician calculates an SDNN value of 60 milliseconds from a 15-minute resting baseline and is concerned that their client may have an elevated heart attack risk. What have they overlooked?

They mistakenly applied cutoffs based on 24-hour recordings to a brief recording. Twenty-four-hour and brief recording values are not interchangeable because short monitoring periods exclude long-term sources of HRV like circadian rhythms. The 50/100 ms thresholds only apply to data collected over a full 24-hour period.

Assignment

Now that you have completed this module, review the missed beat and extra beat graphics and see whether you can identify abnormally long and short IBIs. Based on this module, how might you improve your artifacting?

Additionally, identify the index that is the "gold standard" for predicting the risk of morbidity and mortality when based on 24-hour recording. Which index should be most easily understood by your clients? Why?

Glossary

References

Andreassi, J. L. (2000). Psychophysiology: Human behavior and physiological response. Lawrence Erlbaum Associates.

Aubert, A. E., Seps, B., & Beckers, F. (2003). Heart rate variability in athletes. Sports Medicine, 33(12), 889–919. https://doi.org/10.2165/00007256-200333120-00003

Berntson, G. G., Quigley, K. S., & Lozano, D. (2007). Cardiovascular psychophysiology. In J. T. Cacioppo, L. G. Tassinary, & G. G. Berntson (Eds.), Handbook of psychophysiology (3rd ed.). Cambridge University Press.

Besson, C., Baggish, A. L., Monteventi, P., Schmitt, L., Stucky, F., & Gremeaux, V. (2025). Assessing the clinical reliability of short-term heart rate variability: Insights from controlled dual-environment and dual-position measurements. Scientific Reports, 15, Article 5611. https://doi.org/10.1038/s41598-025-89892-3

Billman, G. E., Schwartz, P. J., & Stone, H. L. (1982). Baroreceptor reflex control of heart rate: A predictor of sudden cardiac death. Circulation, 66(4), 874-880. https://doi.org/10.1161/01.cir.66.4.874

Burr, R. L., Motzer, S. A., Chen, W., Cowan, M. J., Shulman, R. J., & Heitkemper, M. M. (2006). Heart rate variability and 24-hour minimum heart rate. Biological Research for Nursing, 7(4), 256-267. https://doi.org/10.1177/1099800405285268

Ciccone, A. B., Siedlik, J. A., Wecht, J. M., Deckert, J. A., Nguyen, N. D., & Weir, J. P. (2017). Reminder: RMSSD and SD1 are identical heart rate variability metrics. Muscle & Nerve, 56(4), 674-678. https://doi.org/10.1002/mus.25573

Combatalade, D. (2010). Basics of heart rate variability applied to psychophysiology. Thought Technology.

DeGiorgio, C. M., Miller, P., Meymandi, S., Chin, A., Epps, J., Gordon, S., Gornbein, J., & Harper, R. M. (2010). RMSSD, a measure of vagus-mediated heart rate variability, is associated with risk factors for SUDEP: The SUDEP-7 Inventory. Epilepsy & Behavior, 19(1), 78-81. https://doi.org/10.1016/j.yebeh.2010.06.011

Flatt, A. A., & Esco, M. R. (2015). Smartphone-derived heart-rate variability and training load in a women's soccer team. International Journal of Sports Physiology and Performance, 10(8), 994–1000. https://doi.org/10.1123/ijspp.2014-0556

Flatt, A. A., & Esco, M. R. (2016). Evaluating individual training adaptation with smartphone-derived heart rate variability in a collegiate female soccer team. Journal of Strength and Conditioning Research, 30(2), 378–385. https://doi.org/10.1519/JSC.0000000000001095

Flatt, A. A., Hornikel, B., Nakamura, F. Y., & Esco, M. R. (2022). Effect of competitive status and experience on heart rate variability profiles in collegiate sprint-swimmers. Journal of Strength and Conditioning Research, 36(10), 2898–2904. https://doi.org/10.1519/JSC.0000000000003992

Flatt, A. A., & Howells, D. (2019). Effects of varying training load on heart rate variability and running performance among an Olympic rugby sevens team. Journal of Science and Medicine in Sport, 22(2), 222–226. https://doi.org/10.1016/j.jsams.2018.07.014

Gevirtz, R. N. (2017). Cardio-respiratory psychophysiology: Gateway to mind-body medicine.

Grosicki, G. J., Carter, J. R., Laursen, P. B., Plews, D. J., Altini, M., Galpin, A. J., Fielding, F., von Hippel, W., Chapman, C., Jasinski, S. R., Beattie, U. K., & Holmes, K. E. (2026). Heart rate variability coefficient of variation during sleep as a digital biomarker that reflects behavior and varies by age and sex. American Journal of Physiology–Heart and Circulatory Physiology, 330(1), H187–H199. https://doi.org/10.1152/ajpheart.00738.2025

Hämmerle, P., Eick, C., Blum, S., Schlageter, V., Bauer, A., Rizas, K. D., Eken, C., Coslovsky, M., Aeschbacher, S., Krisai, P., Meyre, P., Vesin, J.-M., Rodondi, N., Moutzouri, E., Beer, J., Moschovitis, G., Kobza, R., Di Valentino, M., Corino, V. D. A., Laureanti, R., . . . Swiss-AF Study Investigators (2020). Heart rate variability triangular index as a predictor of cardiovascular mortality in patients with atrial fibrillation. Journal of the American Heart Association, 9(15), Article e016075. https://doi.org/10.1161/JAHA.120.016075

Hill, L. K., & Siebenbrock, A. (2009). Are all measures created equal? Heart rate variability and respiration – biomed 2009. Biomedical Sciences Instrumentation, 45, 71-76. https://pubmed.ncbi.nlm.nih.gov/19369742

Jarczok, M. N., Koenig, J., & Thayer, J. F. (2021). Lower values of a novel index of vagal-neuroimmunomodulation are associated to higher all-cause mortality in two large general population samples with 18 year follow up. Scientific Reports, 11, 2554. https://doi.org/10.1038/s41598-021-82168-6

Jovic, A., & Bogunovic, N. (2011). Electrocardiogram analysis using a combination of statistical, geometric, and nonlinear heart rate variability features. Artificial Intelligence in Medicine, 51, 175-186. https://doi.org/10.1016/j.artmed.2010.09.005

Kleiger, R. E., Miller, J. P., Bigger, J. T., & Moss, A. J. (1987). Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. American Journal of Cardiology, 59, 256-262. https://doi.org/10.1016/0002-9149(87)90795-8

Kurita, A., Takase, B., Kodani, E., Iwahara, S., Kusama, Y., & Atarashi, H. (2013). Prognostic value of heart rate variability in comparison with annual health examinations in very elderly subjects. Journal of Nippon Medical School, 80(6), 420–425. https://doi.org/10.1272/jnms.80.420

Laborde, S., Mosley, E., & Mertgen, A. (2018). Vagal tank theory: The three Rs of cardiac vagal control functioning – resting, reactivity, and recovery. Frontiers in Neuroscience, 12, Article 458. https://doi.org/10.3389/fnins.2018.00458

Lehrer, P. M. (2007). Biofeedback training to increase heart rate variability. In P. M. Lehrer, R. M. Woolfolk, & W. E. Sime (Eds.), Principles and practice of stress management (3rd ed.). The Guilford Press.

Lehrer, P. M., Vaschillo, E., & Vaschillo, B. (2000). Resonant frequency biofeedback training to increase cardiac variability: Rationale and manual for training. Applied Psychophysiology and Biofeedback, 25(3), 177-191. https://doi.org/10.1023/a:1009554825745

Moss, D. (2022). HRV biofeedback bootcamp. Association for Applied Psychophysiology and Biofeedback.

Nakamura, F. Y., Antunes, P., Nunes, C., Costa, J. A., Esco, M. R., & Travassos, B. (2020). Heart rate variability changes from traditional vs. ultra–short-term recordings in relation to preseason training load and performance in futsal players. Journal of Strength and Conditioning Research, 34(10), 2974–2981. https://doi.org/10.1519/JSC.0000000000002910

Nunan, D., Sandercock, G. R. H., & Brodie, D. A. (2010). A quantitative systematic review of normal values for short-term heart rate variability in healthy adults. Pacing and Clinical Electrophysiology, 33(11), 1407-1417. https://doi.org/10.1111/j.1540-8159.2010.02841.x

Penttilä, J., Helminen, A., Jartti, T., Kuusela, T., Huikuri, H. V., Tulppo, M. P., Coffeng, R., & Scheinin, H. (2001). Time domain, geometrical and frequency domain analysis of cardiac vagal outflow: Effects of various respiratory patterns. Clinical Physiology, 21, 365-376. https://doi.org/10.1046/j.1365-2281.2001.00337.x

Plews, D. J., Laursen, P. B., Kilding, A. E., & Buchheit, M. (2012). Heart rate variability in elite triathletes, is variation in variability the key to effective training? A case comparison. European Journal of Applied Physiology, 112(11), 3729–3741. https://doi.org/10.1007/s00421-012-2354-4

Plews, D. J., Laursen, P. B., Stanley, J., Kilding, A. E., & Buchheit, M. (2013). Training adaptation and heart rate variability in elite endurance athletes: Opening the door to effective monitoring. Sports Medicine, 43(9), 773–781. https://doi.org/10.1007/s40279-013-0071-8

Razanskaite-Virbickiene, D., Danyte, E., Mockeviciene, G., Dobrovolskiene, R., Verkauskiene, R., & Zalinkevicius, R. (2017). Can coefficient of variation of time-domain analysis be valuable for detecting cardiovascular autonomic neuropathy in young patients with type 1 diabetes: A case control study. BMC Cardiovascular Disorders, 17(1), Article 34. https://doi.org/10.1186/s12872-016-0467-0

Sammito, S., Thielmann, B., Klussmann, A., Deußen, A., Braumann, K.-M., & Böckelmann, I. (2024). Guideline for the application of heart rate and heart rate variability in occupational medicine and occupational health science. Journal of Occupational Medicine and Toxicology, 19, Article 15. https://doi.org/10.1186/s12995-024-00414-9

Schipke, J. D., Arnold, G., & Pelzer, M. (1999). Effect of respiration rate on short-term heart rate variability. Journal of Clinical and Basic Cardiology, 2, 92-95.

Schwartz, P. J., Vanoli, E., Stramba-Badiale, M., De Ferrari, G. M., Billman, G. E., & Foreman, R. D. (1988). Autonomic mechanisms and sudden death. New insights from analysis of baroreceptor reflexes in conscious dogs with and without a myocardial infarction. Circulation, 78(4), 969-979. https://doi.org/10.1161/01.cir.78.4.969

Shaffer, F., & Ginsberg, J. P. (2017). An overview of heart rate variability metrics and norms. Frontiers in Public Health, 5, Article 258. https://doi.org/10.3389/fpubh.2017.00258

Shaffer, F., McCraty, R., & Zerr, C. L. (2014). A healthy heart is not a metronome: An integrative review of the heart's anatomy and heart rate variability. Frontiers in Psychology, 5, Article 1040. https://doi.org/10.3389/fpsyg.2014.01040

Shaffer, F., Meehan, Z. M., & Zerr, C. L. (2020). A critical review of ultra-short-term heart rate variability norms research. Frontiers in Neuroscience, 14, Article 594880. https://doi.org/10.3389/fnins.2020.594880

Sugimoto, K., Miyaoka, H., Sozu, T., Watanabe, Y., Tamura, A., Yamazaki, T., Ohta, S., Suzuki, S., Shimbo, T., & Hoshino, T. (2024). Associations of age-adjusted coefficient of variation of R-R intervals with autonomic and peripheral nerve function in non-elderly persons with diabetes. Journal of Diabetes Investigation, 15(2), 186–196. https://doi.org/10.1111/jdi.14094

Tarvainen, M. P., & Niskanen, J.-P. (2017). Kubios HRV version 3.0 user's guide. University of Eastern Finland.

Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology (1996). Heart rate variability: Standards of measurement, physiological interpretation, and clinical use. Circulation, 93, 1043-1065. https://pubmed.ncbi.nlm.nih.gov/8598068

Umetani, K., Singer, D. H., McCraty, R., & Atkinson, M. (1998). Twenty-four hour time domain heart rate variability and heart rate: Relations to age and gender over nine decades. Journal of the American College of Cardiology, 31(2), 593-601. https://doi.org/10.1016/s0735-1097(97)00554-8

Voss, A., Schroeder, R., Heitmann, A., Peters, A., & Perz, S. (2015). Short-term heart rate variability—Influence of gender and age in healthy subjects. PLOS ONE, 10(3), Article e0118308. https://doi.org/10.1371/journal.pone.0118308

Zerr, C., Kane, A., Vodopest, T., Allen, J., Fluty, E., Gregory, J., . . . Shaffer, F. (2014). Heart rate variability norms for healthy undergraduates [Abstract]. Applied Psychophysiology and Biofeedback, 39(3), 300. https://doi.org/10.1007/s10484-014-9254-9