Do THC and CBD Oils Affect a Dog’s Heart Rhythm and Body Temperature Differently? A Six-Beagle Study

Short answer: Yes, in a small controlled crossover experiment, THC-containing cannabis oils lowered rectal temperature and altered selected electrocardiogram measurements in conscious dogs, while CBD-only oil did not produce the same temperature or ECG pattern. Six male Beagles each received THC-only, CBD-only, and THC-plus-CBD oils on separate experimental days. The researchers also measured cannabinoid concentrations in plasma by liquid chromatography–mass spectrometry (LC–MS). Their findings support a cautious, formulation-specific veterinary toxicology assessment—not a universal THC blood threshold, an automatic diagnosis of poisoning, or a retrospective determination of an individual animal’s impairment or cause of death.

The original experiment was reported by Rutjanavate, Dong, and Sawangkoon in the International Journal of Veterinary Science and Medicine, published online February 5, 2026: doi:10.4103/IJVSM.IJVSM_3_25. The distinctions among dose, sampled specimen, measured physiology, and observed clinical signs are particularly important when interpreting a pet’s medical record, a possible exposure, or a questioned laboratory result.

What Did the Experiment Actually Show?

Across their separate THC-containing treatment periods, the dogs experienced statistically significant cooling beginning at a cumulative THC dose of 2 mg/kg. Mean rectal temperature fell below 36°C in the researchers’ plotted data at cumulative 4 mg/kg and was still reduced at higher doses. They recorded shivering and vomiting after THC-only and combined THC/CBD preparations; ataxia and a urinary-incontinence episode occurred in the THC-only condition. The CBD-only condition did not produce a corresponding significant temperature reduction or those prominent clinical signs in this six-dog experiment.

The cardiovascular effects were not interchangeable. THC alone lengthened the ECG RR interval at particular cumulative doses, indicating a lower rate of measured cardiac beats during that resting recording. It also prolonged the uncorrected QT interval at 2, 4, and 8 mg/kg and showed a corrected-QT change at 2 mg/kg. Both THC-containing formulations shortened the PQ interval at selected higher doses. The combination oil did not reproduce every ECG finding observed with THC alone, despite containing THC. The investigators did not document an induced malignant arrhythmia or establish that any dog experienced a cardiac death.

The study’s largely negative findings matter too. Systolic and diastolic pressure, most respiratory measurements, and several heart-rate-variability indices did not change significantly at all comparisons. Two specific exceptions appear in the actual source table: mean arterial pressure increased in the THC-plus-CBD condition at cumulative 16 mg/kg, and minute ventilation fell in the THC-only condition at cumulative 2 mg/kg. Calling the entire blood-pressure or respiratory evaluation “no effect” would conceal these reported comparisons.

Who Were the Dogs, and Why Does the Crossover Design Matter?

The sample consisted of six healthy male Beagle dogs, 2–3 years old, averaging 12.6 ± 1.0 kg. They were acclimated to their housing and telemetry equipment before dosing. This was a randomized, three-period, three-treatment crossover: each dog received each oil type in a different sequence, with seven days between treatment periods. Six dogs are therefore six animals, not eighteen independently recruited subjects, even though each animal completed three treatment periods.

Repeated testing of the same dogs helps compare treatments while reducing between-animal variability, but it also creates within-animal correlation. The authors used two-way repeated-measures analysis of variance and Tukey post-hoc comparisons, with a stated statistical-significance threshold of p < 0.05. The design does not remove potential effects of measurement timing, cumulative dosing, previous physiological responses during the same day, or an unmeasured interaction between the components of the combination oil.

No females, puppies, geriatric dogs, dogs with heart disease, or other breeds were enrolled. This experiment is more informative about short-term physiological responses under a laboratory protocol than about everyday exposure across the canine population. The animals were conscious and largely resting during recordings, not running an obstacle course or performing a validated canine coordination test.

What Were the Three Formulations and the Actual Dose Schedule?

The researchers used a THC oil labeled 81 mg/mL, a CBD oil labeled 100 mg/mL, and a combination oil labeled 27 mg/mL THC plus 25 mg/mL CBD. For the combination, the nominal dose was based on its THC content; it consequently supplied approximately 0.93 mg CBD per 1 mg THC. The study was not a comparison between identical total cannabinoid masses or identical CBD exposures.

Each test day involved five administrations, 1, 1, 2, 4, and 8 mg/kg, approximately two hours apart through the buccal/oromucosal route. These were increments, not five standalone study groups. They produced cumulative nominal-dose checkpoints of 1, 2, 4, 8, and 16 mg/kg. The combination-oil checkpoints refer to cumulative THC-matched dose, with additional CBD present. A result at the 16 mg/kg checkpoint occurred after all earlier administrations and is not equivalent to giving a single isolated 16 mg/kg bolus to a previously unexposed dog.

Baseline measurements preceded the first administration by roughly thirty minutes. Subsequent physiological and blood-sampling observations were taken about 1.5 hours after each new dose. As a result, later observations reflect both the newly administered increment and earlier doses that may still have been absorbed, distributed, or eliminated. The trial should not be translated into a clinical dosing instruction or a claim that a specific marketed product is safe at these experimental doses.

Was THC Measured in Whole Blood, Serum, or Plasma?

Plasma was the quantitative toxicology matrix. The investigators collected blood and analyzed separated plasma for parent THC and parent CBD using an LC–MS method with isotope-labeled internal standards. The published table reports ng/mL concentrations as medians and observed ranges across six dogs per treatment. Those measurements are not whole-blood THC determinations, blood-alcohol measurements, oral-fluid screens, or modeled “blood THC equivalents.”

A specimen designation is an analytical fact, not a formality. Plasma and whole blood do not necessarily yield identical cannabinoid numbers; an apparent difference cannot be collapsed into a person- or animal-specific impairment conversion. Likewise, an administered mg/kg value describes the nominal oil delivered, while an LC–MS ng/mL result describes a measured parent-cannabinoid concentration in the sampled plasma at one time point. Neither alone establishes how much drug reached a target organ, how much was metabolized, or why a clinical sign occurred.

What Did the Plasma Data Show at Selected Cumulative Doses?

This evidence table preserves the measured matrix and units and summarizes three of the five experimental checkpoints. Every entry is median plasma concentration in ng/mL (observed six-dog range). The two combination-oil rows concern different analytes measured in the same treatment period, not twelve separate animals.

These figures show substantial animal-to-animal dispersion at the same nominal checkpoint. For instance, the 16 mg/kg combination-oil plasma THC range extended from 36.26 to 430.15 ng/mL, while its median was 230.62 ng/mL. Describing that median as a threshold for toxicity would misrepresent the experiment: it did not establish a diagnostic decision boundary, sensitivity, specificity, or independent validation population. It also did not show that CBD caused the higher median via a demonstrated metabolic interaction; such a mechanism remains a hypothesis, not a measured conclusion.

For additional background on why analyte and specimen names must remain explicit, see biological matrices in forensic toxicology and specimen collection principles. Both topics are relevant when evaluating how a canine toxicology laboratory labeled and handled a sample.

How Were Temperature and Observable Clinical Signs Evaluated?

Rectal temperature was measured directly with a digital thermometer. Video monitoring and observation supplied a separate record of clinical signs. Hypothermia was a physiological measurement; shivering, vomiting, ataxia, and urinary incontinence were observed events. A lowered thermometer reading is not itself a behavioral impairment score, and an observed abnormal gait is not a quantitative plasma THC assay.

The authors reported a significant fall in rectal temperature from the 2 mg/kg cumulative THC checkpoint for both THC-containing formulations. Shivering was seen in approximately half of the six dogs by that level and in all six by 16 mg/kg under THC-containing conditions. Vomiting appeared in a subset at 4 and 8 mg/kg. Ataxia was observed in about half of the dogs in the THC-only condition at 4 and 8 mg/kg, and one dog experienced urinary incontinence at 8 mg/kg THC only. These events were reported to resolve within twenty-four hours without specific treatment.

Those observations establish what happened during this protocol, not a general rule that all canine marijuana exposures produce the same constellation. Dogs with a different body size, breed, comorbidity, route, formulation, or co-exposure were not tested. A suspected real-world case also requires the history, examination, clinical time course, and exclusion of competing causes rather than merely matching one symptom to a study figure.

What Did the Electrocardiogram Show About Rate and Conduction?

Electrocardiography was acquired using external jacket telemetry while the dogs rested. The authors mainly evaluated lead II traces, averaged selected intervals over consecutive beats, and also calculated a QT interval corrected by Van de Water’s formula. An RR interval is the time between consecutive cardiac beats: a longer interval ordinarily corresponds to a slower instantaneous rate. It is not an “RR test” for intoxication.

THC-only oil was associated with significant RR prolongation relative to baseline at cumulative 4 and 8 mg/kg in the Results description. The mixed oil showed a significant RR increase at 2 mg/kg. PQ intervals shortened with THC-only and THC-plus-CBD at 4, 8, and 16 mg/kg compared with baseline, with reported differences relative to CBD. Uncorrected QT intervals lengthened with THC alone at 2, 4, and 8 mg/kg; the corrected-QT comparison was significant at 2 mg/kg. Neither the mixed formulation nor CBD alone reproduced the same THC-only QT pattern.

The original figure distinguishes within-condition baseline comparisons from between-oil comparisons. Those are different statistical questions. A result that differs from baseline in one condition but not another does not, by itself, demonstrate that the two treatment effects significantly differ from each other unless that contrast was tested and supported. Nor does an ECG interval difference prove that a clinically consequential arrhythmia occurred.

Did the Study Prove THC Causes Fatal Arrhythmias in Dogs?

No. QT prolongation can be clinically important because ventricular repolarization abnormalities may relate to arrhythmic risk, but this experiment did not demonstrate torsades de pointes, fatal arrhythmia, cardiac arrest, or a cause-of-death pathway. The authors also assessed the T-wave peak-to-end interval and its ratio to QT; those measures did not show significant treatment changes in this sample.

Temperature provides a plausible competing explanation for part of the ECG response. Hypothermia can affect conduction and rate, and both cooling and longer QT measures appeared in THC-exposed periods. The study did not isolate mediation experimentally, so attributing the QT change solely to a direct cannabinoid effect on myocardial ion channels—or solely to thermoregulation—would exceed what was measured. A forensic assessment should distinguish an observed electrical interval, a proposed physiological mechanism, and an actual cardiac injury or death.

How Did CBD-Only Oil Affect Heart-Rate Variability?

Heart-rate variability (HRV) describes changes in successive cardiac intervals rather than a single pulse value. Researchers evaluated time-domain indices including SDNN and RMSSD and frequency-domain indices such as total power, high-frequency power, low-frequency power, and the LF/HF ratio. A higher HRV index is not a standalone clinical outcome or proof that an intervention is universally beneficial.

At cumulative 8 and 16 mg/kg, CBD-only oil had higher SDNN values than the THC-plus-CBD condition; RMSSD was higher with CBD than with the combination at 8 mg/kg and higher than THC alone at 16 mg/kg. Total HRV power was also higher with CBD than with the combination at 16 mg/kg. The paper reported no significant LF or LF/HF ratio differences among the tested comparisons. This mixed pattern precludes the overbroad claim that every autonomic index improved, or that CBD fully neutralized THC-related cardiac effects.

For example, at cumulative 16 mg/kg, mean ± SEM SDNN was 351.0 ± 13.7 ms under CBD alone and 200.1 ± 26.5 ms under the THC-plus-CBD oil. Those are six-dog repeated-condition means with standard errors, not a clinical reference interval that can classify an unfamiliar animal. HRV may vary with temperature, respiration, activity, stress, instrumentation, and baseline physiology. The investigators did not validate an individual forensic classifier from HRV.

Were Blood Pressure and Breathing Unaffected?

The correct answer is mostly, but not completely. In the study’s pressure table, the THC-plus-CBD treatment showed mean blood pressure rising from 105.3 ± 5.2 mmHg at baseline to 123.5 ± 6.1 mmHg at cumulative 16 mg/kg, with the source marking that comparison significant versus baseline. The other displayed pressure comparisons lacked equivalent statistical marks. Thus “no significant blood-pressure effect” is an oversimplification.

Likewise, respiratory rate and tidal volume did not show significant differences in the published table, but minute volume under THC-only oil fell from 365.3 ± 47.4 to 208.5 ± 29.8 mL/kg/min at cumulative 2 mg/kg, a marked significant comparison with baseline. Other minute-volume comparisons were not marked significant. The measurements were obtained by respiratory inductive plethysmography and do not amount to direct arterial blood-gas assessment or a diagnosis of respiratory failure.

When a scientific abstract summarizes a domain as having little or no significant effect, the primary tables may contain important exceptions. Those exceptions deserve explicit discussion, particularly in a legal or veterinary chart review in which a selective description of “normal vital signs” could be materially misleading.

Evidence Summary: Which Observations Were Significant, and Which Were Not?

Does Adding CBD Demonstrably Protect a Dog From THC?

The absence of the same QT or RR findings in the THC-plus-CBD arm raises a scientifically interesting hypothesis about cannabinoid interactions. It does not establish protection in the clinical or legal sense. The combination was not simply THC with a matched inert additive: it had its own oil concentration and nearly equal nominal CBD co-dose. The data were obtained in just six dogs, all males, across escalating exposures. In addition, combination oil did not prevent the significant temperature reduction.

A mechanistic claim about CYP inhibition, displacement from protein binding sites, direct receptor antagonism, or a lower risk of fatal arrhythmia requires measurements and experimental contrasts beyond those provided here. The measured median plasma THC values were sometimes higher in the combination condition. That observation is consistent with several possible explanations but is not proof of which pharmacokinetic interaction, if any, caused the difference.

CBD-only oil was comparatively well tolerated in this study, yet that should not be generalized into an unrestricted safety guarantee. The experiment was short, relatively small, and restricted to healthy adult male Beagles. A clinician considering product exposure in a different animal must not translate these cumulative experimental checkpoints into a home-use or treatment recommendation.

How Should These Findings Be Used in Veterinary Forensic Toxicology?

Forensic interpretation starts with the actual claim. Is the question whether cannabis was present, whether a particular product was ingested, whether an observed sign is compatible with exposure, whether the exposure was sufficient to cause clinical illness, or whether the exposure caused an individual death? Each question requires different evidence. A laboratory plasma result may identify and quantify a cannabinoid at a sampling time; it does not automatically establish the clinical significance of that value.

A careful case review should recover the product name and independently verified cannabinoid content, the amount actually delivered, the route, the exposure time, the animal’s characteristics, subsequent care, the observed signs and their chronology, and the full laboratory method and matrix. Where a laboratory reports whole blood, urine, serum, or plasma, retain that exact label rather than borrowing plasma concentrations from this research as a direct reference range. More on time-dependent exposure interpretation appears in pharmacokinetics and toxicological interpretation.

Clinical hypothermia, abnormal gait, or ECG findings may support compatibility with cannabinoid effects, but alternative explanations must be evaluated. A medical record may contain anesthesia, sedation, other medications, illness, trauma, heat loss, or unrelated neurologic disease. This experimental paper did not test those differential diagnoses or establish the sensitivity and specificity of any single sign for a forensic exposure classification.

It is also a category error to apply a canine physiological experiment directly to adult human driving impairment, witness memory reliability, a human serum-to-whole-blood conversion, or an intoxication standard. Dogs and humans differ in cannabinoid pharmacology and in how physiological outcomes are measured. The authors themselves discuss differences in cardiac responses between species. Any inference to a human legal question needs appropriate human evidence and independent validation.

What Are the Major Limitations and Uncertainties?

The principal numerical limitation is n = 6 healthy male dogs, followed across five escalating checkpoints for each of three oil types. More sampling moments do not increase the number of biologically independent animals. With many physiological endpoints and pairwise comparisons, a few significant contrasts should not be mistaken for a validated predictive model. The published numerical tables mainly provide means ± SEM or medians and ranges; they do not supply case-specific prediction intervals, diagnostic cutoffs, or false-positive and false-negative rates.

The source did not quantify cannabinoid metabolites or independently measure drug distribution into brain tissue and cardiac tissue. Plasma parent THC and CBD concentrations were measured, but causal tissue exposure and toxicity pathways remained unverified. The protocol included potentially high cumulative dosing over one day rather than isolated low exposures under ordinary household conditions, and the seven-day washout separated study periods rather than doses within a period.

Additional uncertainty arises from the route of oromucosal delivery, individual variability, controlled housing, and resting telemetry. The study did not establish veterinary emergency-department prevalence, rate of clinically dangerous arrhythmias, causes of death, chronic effects, or safety in medically vulnerable dogs. The directly observed findings should therefore be treated as controlled experimental evidence of possible effects, not a case-level reconstruction method.

Frequently Asked Questions

1. Can THC-containing oil lower a dog’s body temperature?

In these six conscious male Beagles, both THC-only and THC-plus-CBD oils caused a statistically significant decrease in rectal temperature beginning at a cumulative THC dose of 2 mg/kg. The finding shows a possible experimental response under this schedule; it is not a validated threshold for all canine exposures.

2. Does a high plasma THC concentration prove a dog was impaired?

No. The researchers quantified parent THC in separated plasma, but did not establish an individual impairment classifier. The same nominal dose produced wide concentration ranges across the dogs. A quantitative result must be interpreted with exposure history and contemporaneous clinical observations.

3. Does CBD cancel out THC’s heart effects?

Not as an established general rule. Some THC-only ECG responses were less evident in the combined-oil condition, but the combination still produced hypothermia and other signs. This small study did not prove a protective clinical mechanism or demonstrate fewer serious arrhythmias.

4. Were dangerous arrhythmias actually observed?

The authors described interval changes, including QT prolongation with THC alone, but did not report induced malignant arrhythmia or arrhythmic death. An ECG interval is a measured electrical feature, not evidence that an individual animal suffered a fatal rhythm disorder.

5. Is a 16 mg/kg checkpoint one single administered dose?

No. It was a cumulative checkpoint reached through separate administrations of 1, 1, 2, 4, and 8 mg/kg at roughly two-hour intervals. It must not be treated as a single-dose trial or a recommended clinical regimen.

6. Did breathing and blood pressure remain completely unchanged?

No. Most pressure and respiratory comparisons lacked significant differences, but the source reported increased mean blood pressure under THC plus CBD at cumulative 16 mg/kg and lower minute ventilation under THC alone at cumulative 2 mg/kg, each versus baseline. These limited findings did not demonstrate respiratory failure or persistent hypertension.

7. Can these six dogs establish the cause of a pet’s illness or death?

No. The study can inform whether certain signs are compatible with acute exposure, but individual causation requires verified exposure, full medical and necropsy evidence when relevant, specimen-specific analysis, chronology, differential diagnoses, and appropriately supported expert interpretation.

Primary Source and Interpretation Boundary

The sole original scientific study analyzed here is Rutjanavate N, Dong VNK, and Sawangkoon S, “Cumulative Effects of Cannabis Oils on Body Temperature, Electrocardiography, Heart Rate Variability, Blood Pressure, and Respiratory Function in Telemetered Dogs,” International Journal of Veterinary Science and Medicine (online February 5, 2026), DOI 10.4103/IJVSM.IJVSM_3_25. Its findings concern controlled acute canine exposure, directly measured plasma cannabinoids, observed clinical signs, and physiological recordings. They do not themselves resolve disputed individual exposure history, intoxication severity, treatment appropriateness, human impairment, or fatal causation.