Does a Higher Breath Alcohol Peak Predict an Arrhythmia? A 193-Person Holter Study

Answer: Not reliably from this study. In a 193-person prospective observational binge-drinking cohort, a higher measured peak breath-alcohol reading was associated with a faster heart rate and lower electrocardiogram-derived heart-rate variability (HRV). However, the subgroup analysis did not show that clinically relevant arrhythmias became significantly more frequent across the higher-drinking or higher-peak-alcohol groups. A breath result, a physiological response, and a diagnosed abnormal heart rhythm are different findings. None alone proves an individual’s impairment, the origin of a later cardiac event, or legal causation.

The relevant primary research is von Falkenhausen and colleagues, “The level of acute alcohol exposure during binge drinking associates with the extent of cardiac response,” published online in 2025 in Clinical Research in Cardiology (2026 volume), doi:10.1007/s00392-025-02722-4. It addresses graded alcohol exposure and actual Holter-detected rhythm outcomes, not whether pulse or HRV can be used as a standalone breath or blood alcohol test.

What did the investigators actually ask?

The Munich-BREW II research program followed adults during supervised binge-drinking episodes and subsequently monitored their electrocardiograms. This particular paper examined whether the amount of alcohol consumed or the highest recorded breath-alcohol value was more closely related to the size of the observed cardiac response. Its investigators analyzed groups ranked from lower to higher alcohol intake or peak breath-alcohol result.

That distinction matters in forensic evaluation. A recorded quantity of alcoholic beverages is an exposure-history estimate, influenced by drink contents and what was actually consumed. A breath instrument supplies a measurement of ethanol in exhaled breath at a testing time. Heart rate and HRV are physiological responses calculated from ECG recordings. Clinically relevant arrhythmia is a separate rhythm endpoint—not interchangeable with a faster pulse or altered HRV.

The paper did not validate a diagnostic algorithm for a particular civil or criminal case, an alcohol-induced cardiac arrest, an individual driver’s intoxication, or a universally safe or unsafe peak blood-alcohol threshold.

How were drinking, alcohol concentration and ECG activity recorded?

The researchers prospectively enrolled adults intending to drink between October 2016 and July 2017. Participants consumed alcoholic beverages at supervised public or private gatherings. Study staff documented beverage types and quantities; they measured alcohol in breath at the start and hourly through the drinking period. A wireless three-lead Holter ECG patch continuously recorded cardiac electrical activity for approximately 48 hours from the start.

Investigators identified the hour during which each person’s peak recorded breath-alcohol result occurred, then calculated average heart rate and two HRV measures within that hour. They divided the cohort into quartiles separately according to reported ethanol intake in grams and maximum recorded breath-alcohol concentration. Their reported “BAC” terminology and gram-per-kilogram displays derive from breath testing, not a separately documented venous whole-blood gas-chromatography specimen for each timepoint. A reported blood-equivalent display should not be substituted for a directly analyzed blood sample. The site’s breath alcohol measurement guide explains that analytical distinction.

This was a single-arm observational cohort, not a randomized dose-assignment experiment. Participants selected different types and quantities of beverages, and they differed in characteristics such as sex, body mass and habitual drinking. Quartile comparisons identify associations within the observed sample; they do not isolate an ethanol-only causal pathway from every other contemporaneous influence on cardiac measures.

Who was included in the 193-person analysis?

There were 193 participants, with a mean age of 29.9 ± 10.6 years; 70 (36%) were women. The study excluded people with known atrial fibrillation, implanted cardiac rhythm devices, and pregnant or breastfeeding individuals. Participants were generally young and without a substantial recognized cardiovascular-risk profile. These inclusion characteristics constrain extrapolation to an older person, someone with structural heart disease, or a patient taking medications affecting cardiac conduction.

Men consumed more reported drinks in the subgroup description, yet their maximum recorded breath-derived alcohol values were not significantly different from those of women: the reported medians were 1.46 versus 1.42 g/kg in the study’s BAC-equivalent presentation (p = 0.13). This comparison illustrates why a drink count and an observed peak physiological exposure should not be assumed numerically interchangeable. It does not establish a sex-independent exposure rule applicable to every drinking occasion.

What changed at higher levels of drinking or peak breath alcohol?

At the time of maximal recorded breath alcohol, the participant groups reporting greater alcohol intake had higher median heart rates and lower HRV. Groups reaching higher peak breath-alcohol values showed the same direction of association. The two HRV indices—standard deviation of normal beat-to-beat intervals (SDNN) and root mean square of successive differences (RMSSD)—are derived from timing intervals in an ECG trace. They are not direct ethanol concentrations, estimates of a particular person’s blood alcohol, or independently validated impairment scores.

These p-values compare the observed groups under the paper’s analytical methods. They do not quantify the magnitude of the effect for a randomly selected defendant, patient, decedent, or witness, and they do not give an individual’s likelihood of having an arrhythmia. The paper’s findings are strongest as evidence that peak exposure and the cardiac autonomic indices are associated in this study setting.

Did higher peak alcohol predict clinically relevant arrhythmias?

No statistically significant subgroup difference was demonstrated. The authors reported that clinically relevant arrhythmias occurred in 5.2% of the full cohort, consistent with the preceding parent study. In their new analyses, the incidence of those arrhythmias did not differ significantly across quartiles of total alcohol intake or peak breath-derived alcohol concentration.

This does not prove ethanol could not precipitate an arrhythmia. A non-significant result may reflect a small number of events, limited subgroup power, variability in susceptibility, or the actual absence of a detectable trend under these circumstances. But it equally does not validate the opposing claim that every rise in heart rate or fall in HRV during drinking establishes a clinically significant arrhythmia or predicts that an event must happen.

The 5.2% figure is a descriptive occurrence within this selected monitored cohort. It is not an individual predictive probability and should not be transformed into an alcohol-attributable fraction or a threshold at a given breath reading. The subgroup paper did not report validated sensitivity, specificity, likelihood ratios, or a personal risk prediction tool for atrial fibrillation or other rhythm disorders.

Why is a faster heart rate not the same as an arrhythmia?

Heart rate counts beats per unit time. HRV summarizes differences among successive beat intervals, with calculation and interpretation dependent on the data window and ECG quality. A clinically relevant arrhythmia involves a rhythm finding that meets defined electrical and clinical criteria. Although these may relate to one another, they are not interchangeable endpoints.

For example, someone with a faster but otherwise organized sinus rhythm does not thereby have documented atrial fibrillation. A lower SDNN or RMSSD does not demonstrate an ectopic beat, sustained tachyarrhythmia, a cardiac event, or the cause of loss of consciousness. Conversely, the absence of a large HRV change does not guarantee that no rhythm disturbance occurred. Actual ECG strips, timing, diagnostic criteria and the individual’s medical history matter more for a particular cardiac allegation than a quartile-level p-value.

Do the 24-hour findings establish recovery or sobriety?

The researchers compared resting heart rate and HRV about 24 hours after the documented peak breath-alcohol interval and did not find subgroup differences at that reference point. This is a statement about measured resting indices across their groups; it is not a forensic elimination curve, a validated time-to-zero alcohol calculation, or proof that no individual had residual effects or later arrhythmia.

The 48-hour Holter recording interval is also distinct from 48 hours of direct ethanol measurement. Study breath alcohol was recorded hourly during the drinking period, whereas ECG recordings continued much longer. No continuous breath or whole-blood ethanol curve was acquired for all those hours. It would be erroneous to interpret a later Holter finding as if an alcohol concentration had necessarily been measured at that same moment.

For why a concentration–time relationship cannot be inferred from an isolated dose, visit the site’s pharmacokinetics guide. That explanatory resource does not replace the individual’s original breath-test record or the actual Holter timeline.

Can a breath-derived alcohol value determine a precise venous BAC?

No. The paper used breath testing to track alcohol exposure and reported concentrations in a BAC-style unit. A valid forensic whole-blood concentration is a different measurement requiring an identified specimen and laboratory method. Any breath-to-blood inference also raises questions about device operation, breath sampling, reporting convention, time relative to consumption, and physiological partitioning. None of those issues is resolved by observing that people grouped at higher breath values had faster heart rates.

Similarly, the monitored cardiac endpoints are not alternative methods of measuring ethanol. A normal-looking ECG does not exclude recent drinking; a high pulse cannot identify an alcohol dose; and even a technically correct breath measurement does not prove that the person was unsafe to operate a vehicle or that alcohol caused a later death. Analytical exposure, biological response, task performance and causation remain separate evidentiary questions.

What would be necessary in a disputed cardiac or driving case?

First, establish the exact evidentiary proposition: was an individual drinking, was ethanol present at a particular time, was a rhythm disturbance documented, was the person functionally impaired, or did a specific exposure cause a cardiac event? The study cannot answer all those questions through a single variable.

For a case-specific chronology, obtain original instrument records and the reporting basis for breath values, contemporaneous clinical or forensic specimens if any, beverage history and known timing, full Holter or ECG traces with timestamps, diagnostic rhythm interpretation, symptoms and medical records. Review competing physiological or environmental factors—including activity, medication and preexisting disease—without assuming their contribution. Then identify which inferences rely on actual measurements and which are reconstructed or hypothetical.

The authors’ dose-versus-peak comparison is useful because actual peak exposure was more closely associated with their chosen cardiac autonomic indices than raw beverage quantity alone. It is not a justification for substituting ECG findings for toxicology, deriving a BAC from pulse, or diagnosing one person’s arrhythmia from a population curve. The site’s scientific evidence library provides related method-specific interpretation resources.

What are the important limitations of this study?

This was an observational subgroup analysis from a single-center, selected binge-drinking cohort. The design involved no randomized alcohol doses and did not isolate all potential co-variables affecting heart rate or HRV. Participants were mainly young and relatively healthy; those with a history of atrial fibrillation were excluded. Results are indexed to each participant’s maximum recorded hourly breath value, which need not equal an unobserved true peak or a venous whole-blood measurement. Clinically significant arrhythmias were comparatively uncommon, leaving uncertain power to identify differences across the quartiles.

The authors found no validated clinical cutoff, no device-level test-performance statistics for prediction of a rhythm disorder, and no assessment of driving, falls, crash mechanics, or legal impairment. A lack of observed significant arrhythmia differences is not proof of absence of risk. Conversely, evidence of a group-level autonomic response does not by itself establish that alcohol precipitated an individual’s medical event.

Frequently asked questions

Can a high peak breath alcohol reading prove atrial fibrillation?

No. The study found dose-related heart-rate and HRV patterns but did not establish significantly different clinically relevant arrhythmia rates across breath-exposure quartiles. Atrial fibrillation requires appropriate rhythm evidence.

Does low heart-rate variability mean a person was drunk or unsafe to drive?

No. SDNN and RMSSD are ECG-derived timing variables affected by many circumstances. This study did not validate them as classifiers of BAC, legal intoxication or actual driving impairment.

Did 5.2% of people develop an alcohol-caused arrhythmia?

The researchers reported clinically relevant arrhythmias in 5.2% of the observed cohort. That is an event frequency, not an experimentally demonstrated alcohol-attributable percentage or a prediction for any individual.

Was alcohol measured directly in venous blood?

Not as the exposure series reported for this subgroup analysis. The authors obtained hourly breath values and used the term BAC and associated display units. Those values should not be relabeled independent laboratory whole-blood ethanol determinations.

Does consuming more alcohol always produce a higher heart rate?

No such universal rule was established. The reported findings compared grouped observations; individuals can have different concentration histories and physiological responses at a given reported drink count.

Did the absence of group differences at 24 hours prove alcohol was eliminated?

No. The finding concerned resting ECG measures at a reference time, not a full 24-hour concentration measurement or verified zero ethanol in every participant.

Can the cohort establish why a particular person collapsed or drowned?

No. A causation analysis requires a reliable clinical timeline, direct rhythm evidence, relevant toxicology, individual medical context and competing explanations. This study supplied neither a case-specific diagnostic algorithm nor a test of drowning or collapse mechanisms.

Primary scientific study

von Falkenhausen AS, Krewitz C, Winter R, Kern A, Brunner D, Brunner S, Sinner MF. The level of acute alcohol exposure during binge drinking associates with the extent of cardiac response. Clinical Research in Cardiology. Published online August 19, 2025; volume 115 (2026), pages 1337–1342. https://doi.org/10.1007/s00392-025-02722-4. Interpretation is confined to the original prospective cohort and subgroup analyses; it is not an individual toxicology or cardiac-causation validation study.