PPG vs ECG Accuracy Comparison: Heart Rate, HRV, and Arrhythmia Detection Data

Arrhythmia Detection: PPG vs ECG

Arrhythmia detection is where the comparison shifts most sharply against PPG. ECG directly observes electrical waveform morphology, which is essential for classifying most arrhythmias. PPG sees only the mechanical output of those electrical events. However, one arrhythmia stands out as well-suited to PPG detection: atrial fibrillation.

Atrial Fibrillation Detection

AFib causes a chaotic, irregularly irregular pulse rhythm. This irregularity is visible in PPG-derived pulse interval sequences, even without seeing the underlying P-wave absence or fibrillatory baseline. Multiple algorithms have been validated for PPG-based AFib detection.

Sensitivity in controlled settings is high (98.0% in Tison et al.'s cardioversion cohort), but the same algorithm reached only 67.7% sensitivity and 67.6% specificity in ambulatory users. In the Apple Heart Study, the positive predictive value of an irregular pulse notification was 0.84 for AF on a simultaneous ECG patch, reflecting the challenge of population screening. A positive PPG screen still requires ECG confirmation before clinical action.

For a thorough breakdown of PPG-based AFib screening performance and limitations, see our PPG atrial fibrillation screening guide.

Other Arrhythmias

For arrhythmias beyond AFib, PPG performance drops substantially and no validated screening applications exist for routine wearable use.

For any rhythm disorder beyond AFib, ECG remains mandatory for diagnosis. PPG can flag that something may be abnormal (unusual rate or rhythm irregularity), but it cannot characterize what that abnormality is.

Respiratory Rate Accuracy: PPG vs ECG

Both PPG and ECG can be used to estimate respiratory rate through respiratory-induced amplitude modulation, frequency modulation, and baseline wander. Neither is a direct measurement of breathing, but both can extract respiratory rate without requiring dedicated respiratory hardware.

At rest, the best ECG- and PPG-derived algorithms approach the agreement of impedance pneumography, the clinical standard, though Charlton et al. found that algorithms generally performed better on ECG than on PPG. Both deteriorate during rapid breathing or in the presence of motion. Neither replaces dedicated spirometry or chest expansion sensors for clinical respiratory monitoring.

The reference standard for respiratory rate varies by study (thoracic impedance, nasal thermistor, capnography), which makes direct numeric comparison across studies less clean than for heart rate.

Summary Table: PPG vs ECG Across Measurement Domains

Key Papers and Validation Sources

The accuracy data above draws from published validation literature. Two foundational papers are worth reading directly:

Schäfer A, Vagedes J. "How accurate is pulse rate variability as an estimate of heart rate variability? A review on studies comparing photoplethysmographic technology with an electrocardiograph." International Journal of Cardiology, 2013. https://doi.org/10.1016/j.ijcard.2012.03.119

Tison GH, et al. "Passive Detection of Atrial Fibrillation Using a Commercially Available Smartwatch." JAMA Cardiology, 3(5), 409-416, 2018. https://doi.org/10.1001/jamacardio.2018.0136

For the broader comparison of PPG and ECG as signal types, including waveform anatomy and timing relationships, see our PPG vs ECG comprehensive guide.

FAQ

Is PPG as accurate as ECG for measuring heart rate?

At rest, yes in practical terms. Modern wrist PPG devices achieve mean absolute errors of 1-3 BPM against simultaneous ECG, which is within the IEEE standard of 5 BPM for wearable heart rate monitors. During exercise, the gap grows and becomes device- and activity-dependent because of motion artifacts, which matters more if you are trying to stay within narrow training zones.

Can PPG be used for HRV analysis instead of ECG?

For wellness HRV at rest (RMSSD, SDNN from 5-minute recordings), PPG shows strong correlation with ECG references in healthy subjects, with a median RR-to-PP interval correlation of 0.97 in a fingertip PPG study (Selvaraj et al., 2008) and agreement above 0.9 for Apple Watch RR intervals (Hernando et al., 2018). For clinical HRV in patients with arrhythmias or during activity, PPG is not reliable enough. The pulse arrival time variability adds noise that does not represent cardiac autonomic variation.

What is the sensitivity of PPG for detecting atrial fibrillation?

In controlled validation, a smartwatch PPG algorithm detected AF with 98.0% sensitivity and 90.2% specificity against 12-lead ECG, but only 67.7% sensitivity and 67.6% specificity in ambulatory users (Tison et al., 2018). The Apple Heart Study found a 0.84 positive predictive value for an irregular pulse notification in a population screening context, meaning about 16% of such alerts were not confirmed as AF on a simultaneous ECG patch. Any positive PPG-based AFib alert requires ECG confirmation.

Why can PPG detect atrial fibrillation but not other arrhythmias?

AFib produces a characteristic irregularly irregular pulse rhythm that is detectable in the PPG waveform's inter-beat interval sequence without needing to see the electrical waveform. Other arrhythmias (flutter, SVT, heart block, bundle branch block) are diagnosed based on P-wave shape, QRS morphology, or specific interval measurements that require the ECG electrical signal. PPG simply does not carry that information.

How does PPG accuracy for HRV change with exercise?

HRV accuracy from PPG degrades significantly during exercise. Frequency-domain metrics like LF/HF ratio show the steepest decline. Time-domain metrics like RMSSD are more resilient but still show increased error during movement. Most researchers recommend a 5-10 minute quiet recovery period before PPG HRV measurements intended for comparison with resting ECG references.

Does PPG measure the same intervals as ECG for HRV?

No. ECG measures R-R intervals between consecutive QRS complexes. PPG measures pulse-to-pulse intervals (PPI) between consecutive pulse peaks. The difference is the pulse arrival time (PAT), which is typically 200-400 ms. Because PAT is roughly constant beat-to-beat at rest, the variability in PPI is close to the variability in RR. But PAT fluctuates with blood pressure changes, sympathetic tone, and movement, adding a source of HRV-like variance to PPG that has no cardiac autonomic origin.

What measurement can ECG do that PPG cannot?

ECG can measure ST segment changes (ischemia), P-wave morphology (atrial abnormalities, AFib classification, flutter), QRS duration and morphology (bundle branch blocks, ventricular hypertrophy), QT interval (drug safety, channelopathies), and precise AV conduction timing (heart block). None of these require an optical signal. They are all about the electrical sequence within the heart, which PPG does not observe at all.