Can Overnight Heart Rate After Alcohol Prove Next-Day Impairment? A Sleep-Lab Study

Short answer: No. Drinking alcohol before bed can increase sleeping heart rate and alter sleep architecture, but those physiological changes do not, by themselves, establish next-morning cognitive impairment, a particular blood-alcohol concentration, or the cause of an individual event. In a randomized crossover experiment involving 31 healthy adults, alcohol increased mean nocturnal heart rate from 56.4 to 65.0 beats per minute and reduced several objective measures of sleep quality. Yet the researchers did not find significant associations between those overnight changes and the modest mood or task-performance differences observed the next morning. That is a finding about this study and its statistical power—not proof that alcohol cannot cause next-day problems.

What did the controlled alcohol-and-sleep experiment ask?

Pabon, Greenlund, Carter, and de Wit reported a randomized, two-condition, overnight crossover study in Alcoholism: Clinical and Experimental Research in 2022. Their central question was more precise than whether alcohol affects sleep: Can a person’s acute response to alcohol, or the later changes in sleep and overnight cardiac activity, predict how that person feels or performs the next morning? The study included 31 men and women who completed alcohol and control-beverage nights separated by approximately one month.

This distinction matters in forensic interpretation. Finding that alcohol changes a sleep measure is a different proposition from showing that the measure predicts impaired attention, compromised driving, or negligence the following morning. The second proposition requires direct performance evidence and a valid predictive relationship. The authors investigated those relationships rather than assuming them.

Who was studied, and who was excluded?

Participants were healthy social drinkers aged 21–45 years; 15 were men and 16 were women. The sample mean age was 25.3 years (standard error 1.1). Eligibility required recent self-reported binge drinking. The protocol excluded participants with certain medical conditions or treatments, including diagnosed or treated obstructive sleep apnea, diabetes, daily tobacco use, blood-pressure medication, and moderate or severe alcohol-use disorder. A familiarization visit and sleep screening preceded the two experimental nights. These exclusions limit application to a medically complex person, an older adult, a person with sleep apnea, or someone who mixes alcohol with another drug.

How were the alcohol dose, breath results, sleep, and performance measured?

The assigned alcohol dose was 1.0 g/kg for men and 0.85 g/kg for women, divided across evening beverage periods beginning at 8:00 and 9:00 p.m. The control condition used fruit juice of equal volume with a small quantity of ethanol to help mask beverage identity; the paper describes 1 mL of ethanol in the control drink and ethanol mist on the cup. Accordingly, the control was not guaranteed to contain zero ethanol molecules. Subjects remained in the sleep laboratory, with lights out at 11:00 p.m. and waking at approximately 7:00 a.m.

Investigators sampled exhaled breath with an Intoximeter Alco-Sensor at multiple evening points and again at 7:00 a.m. They also obtained baseline breath and blood samples to confirm abstinence before the visit. However, the paper does not present a paired quantitative venous-blood ethanol series against which to validate a particular blood-alcohol result. The authors describe a mean peak “BAC” of approximately 0.1% (SD 0.02%) in their Results, while the related figure labels breath alcohol in mg/L. This reported convention must not be mistaken for an independently measured whole-blood concentration.

Sleep was assessed using laboratory polysomnography: EEG-based sleep staging, eye and muscle recordings, and cardiovascular monitoring that included pulse oximetry and two-lead ECG. It measured sleep-stage proportions, total sleep time, efficiency, and nighttime heart rate. Those are physiological and sleep-scoring observations, not assays of ethanol and not a clinical diagnosis of impaired driving.

At baseline, around the expected evening peak, and the next morning, participants completed selected tasks: a flanker attention task, pursuit-rotor motor control, digit span for working memory, and two-column arithmetic. Questionnaires separately assessed perceived sedation, arousal, mood, and urge to drink. A questionnaire rating is a subjective report; task accuracy or reaction time is an observed laboratory performance outcome. Neither automatically becomes a real-world injury or crash determination.

What changed overnight after alcohol?

Table 2 of the original paper provides condition means and their standard errors. In this specific protocol, alcohol raised nighttime heart rate and changed the distribution and amount of sleep. The comparison is within a controlled crossover design; it is not an alcohol-detection threshold or a value that can be plugged into a blood-alcohol calculator.

Expressed as contrasts between reported means, the alcohol night had an 8.6-beat/min higher sleeping heart rate, about 15.3 fewer minutes of recorded sleep, 3.2 percentage points lower efficiency, and 3.5 percentage points less REM sleep. These arithmetic contrasts describe group means; they do not quantify an individual person’s expected change, and the paper does not validate any such forensic prediction. A statistically significant mean heart-rate change is also not a documented arrhythmia or proof of a pathological cardiac event.

Were participants actually impaired during the evening?

For selected laboratory tasks, yes: the paper reports significant group-level acute effects during the drinking session. Alcohol worsened flanker congruent response times and error counts, pursuit-rotor time-on-target, and two-column addition accuracy relative to their control comparisons (reported drug-by-time p values below 0.05). Acute changes in reported sedation, low-arousal negative feelings, high-arousal positive feelings, and drinking urge were also observed.

These findings should be stated with the correct temporal and evidentiary boundaries. The evening results occurred under the assigned experimental alcohol conditions and were measured on laboratory tasks. They do not establish whether a particular person at an unknown time was unable to drive safely, violated a legal standard, or caused an incident. Conversely, the lack of a significant result on one task cannot be taken as proof of intact performance on all other tasks.

Did the higher sleeping heart rate predict next-morning impairment?

Not in this 31-participant experiment. The investigators calculated within-person alcohol-minus-control changes in nocturnal heart rate and sleep measures and examined how those differences related to next-morning changes in mood and performance. They report that none of the overnight alcohol effects on sleep or heart rate was significantly associated with the next-morning mood and behavior changes they measured. This is a failure to demonstrate the proposed relationship in the observed sample—not a definitive demonstration that the true association is zero in every population or drinking circumstance.

There were still modest morning-after findings. Participants reported more low-arousal negative feelings and sedation, and less urge to drink. Digit-span performance showed a small statistically significant improvement in the alcohol condition (drug-by-time p = 0.044). That isolated result must not be reframed as proof that alcohol benefits memory generally; it arose amid multiple outcomes in a small laboratory experiment and does not identify a reliable effect in an individual.

What about perceived sedation and stage N2 sleep?

The researchers detected one positive association between the acute increase in self-rated evening sedation and the increase in proportion of N2 sleep (reported r = 0.5, p < 0.05). The association involved evening subjective sensation and later sleep-stage composition. It did not demonstrate that either variable caused the other. More importantly for a morning-after claim, it did not supply an established conversion from N2 percentage or sleeping heart rate to an individual’s next-day impairment level.

What do the different evidence types actually establish?

Why is this different from a cardiac ECG or blood-alcohol validation study?

A normal or elevated pulse has many potential causes, and a two-lead ECG during sleep is not a forensic ethanol assay. By contrast, blood-alcohol analysis directly measures ethanol in a properly characterized biological specimen using an appropriate validated analytical method. Breath testing addresses exhaled ethanol under instrument-specific sampling assumptions. The two measurement systems should not be conflated.

This experiment focuses on overnight sleep and whether those changes predicted next-day study outcomes. It is not a study validating autonomic ECG features to classify acute BAC, not a field-sobriety validation trial, and not a direct comparison of breath and whole-blood assays. For an explanation of that analytical distinction, see the site’s breath alcohol measurement guide. General timing and interpretation questions are separately covered in pharmacokinetics in forensic toxicology.

What scientific limitations matter in an individual case?

Selection: the sample consisted of 31 relatively healthy young social drinkers. Diagnosed or treated sleep apnea and several other health conditions were exclusions, not tested modifiers. The findings cannot be extrapolated to patients with those conditions without further evidence. Setting: alcohol was given in two scheduled evening periods, with standardized laboratory routines, meals, monitoring, and sleep opportunities. That differs from extended social drinking, concurrent drug exposure, irregular sleep, pain, illness, or other stressors.

Measurements: the paper’s analysis relied on a breath testing schedule, rated sensations, polysomnography, and selected performance tests. It did not measure actual automobile driving, workplace injury, or legal causation. The label “BAC” in its Results should not override the stated breath sampling method. Inference: small-sample non-significant associations do not establish equivalence, clinical safety, or lack of impairment; statistically significant group effects likewise cannot determine the functional condition of an identified individual. The source analyzed multiple physiological, mood, and task outcomes, which calls for particular caution with isolated findings.

For a specific event, relevant evidence would include the individual timeline, alcohol and co-exposure records, analytical specimen and collection details, contemporaneous observations, any directly assessed performance, sleep history, medical conditions, and competing explanations. None of those case-level facts can be reconstructed from this study’s group averages.

How should the study be used in forensic testimony?

The strongest defensible description is narrow: under a controlled high-dose, overnight crossover protocol, alcohol increased sleeping heart rate and disrupted several sleep measurements, yet those changes were not significantly linked to the limited next-morning outcomes observed. It would overstate the evidence to call nocturnal tachycardia an alcohol “fingerprint,” to use sleep-stage percentages as a BAC conversion table, or to claim that either proves an accident was alcohol-caused.

The reverse overstatement is also unsound. The study cannot rule out morning-after impairment in another person, especially one outside its inclusion criteria. Scientists must distinguish an exposure-related physiological response from a validated diagnostic classifier; they must also distinguish correlation from causation and a group observation from an individual conclusion. For related resources on analytical limitations and evidence categories, consult the forensic toxicology evidence library.

Frequently asked questions

Can alcohol increase heart rate while a person sleeps?

Yes. In this laboratory crossover sample, mean nocturnal heart rate was 65.0 beats/min after alcohol compared with 56.4 after the control beverage (p < 0.001). That is an observed group effect, not a specific test for ethanol in an individual.

Does an elevated morning or nighttime pulse prove a person was drunk?

No. Heart rate is a nonspecific physiological variable. The study did not validate a pulse cutoff for alcohol exposure, a blood-alcohol concentration, or clinical/legal impairment.

Did alcohol reduce REM sleep?

In this study, mean REM percentage decreased from 20.0% in the control condition to 16.5% after alcohol (p < 0.005). A group-level sleep-stage contrast cannot identify the cause of a particular person’s sleep pattern or prove next-day impairment.

Did worse sleep predict poor morning-after performance?

Researchers did not detect significant associations between alcohol-related changes in overnight sleep or heart rate and the next-morning mood or performance changes they measured. The negative association results are limited by sample size, selected endpoints, and study conditions.

Does the paper prove that hangovers cannot impair driving?

No. There was no on-road or driving-simulator test in this experiment, and the authors did detect some next-morning subjective changes. Lack of predictive sleep/heart-rate associations is not proof that all next-day performance is normal.

Was the reported peak alcohol value a directly measured blood sample?

The paper reports peak “BAC” in its narrative, but its serial alcohol monitoring is described as breath testing with an Alco-Sensor; no paired direct quantitative venous-blood ethanol results are provided to validate the narrative percentage as a whole-blood measurement.

Does this study apply to a person with obstructive sleep apnea?

Not directly. Diagnosed or treated obstructive sleep apnea was an exclusion criterion. An individual with sleep apnea or other relevant health conditions would require separate clinically and analytically grounded evaluation.

Primary scientific source

Pabon E, Greenlund IM, Carter JR, de Wit H. Effects of alcohol on sleep and nocturnal heart rate: Relationships to intoxication and morning-after effects. Alcoholism: Clinical and Experimental Research. 2022;46(10):1875–1887. DOI: 10.1111/acer.14921. This discussion addresses that original study’s protocol, reported results, and limitations; it is educational scientific analysis, not a case-specific clinical or legal conclusion.