EVIDENCE LIBRARY
Do CBD and THC Nanoemulsions Damage Kidneys? Rat Histology Is Not Kidney Function
Rat kidney histology changed after 21 days of some CBD/THC nanoemulsions. Why tissue markers do not prove kidney failure or human causation.
Short answer: A 2026 experiment reported microscopic kidney-tissue changes in male rats given certain cannabidiol (CBD) or CBD-plus-delta-9-THC nanoemulsions for 21 days, particularly in the groups labeled 5 mg/kg. It did not measure kidney filtration, serum creatinine, blood urea, urine biomarkers, or human clinical outcomes. The results therefore support a limited finding about rat kidney morphology and tissue-marker staining under the tested conditions—not a demonstrated case of human kidney failure, a general declaration that cannabis products are nephrotoxic, or proof that a person experienced renal impairment.
The distinction matters in toxicology: an observed tissue change, a biochemical marker, an organ-function test, and a clinical diagnosis are different endpoints. They cannot be treated as interchangeable simply because each may be described as evidence relating to “toxicity.”
What did the researchers actually test?
Guedes Pinto and colleagues published an exploratory laboratory study in Pathophysiology in 2026. They randomized 40 male Wistar rats, eight per group, to a vehicle-only control or one of four cannabinoid nanoemulsion regimens. Animals received the assigned preparation by oral gavage once daily for 21 consecutive days. The formulations were either CBD alone or a combination labeled CBD + Δ9-THC, at group labels of 2.5 or 5 mg/kg. The control preparation used the nanoemulsion vehicle without cannabinoids.
A nanoemulsion is a formulation in which small oil droplets disperse lipophilic compounds in another phase. The delivery system may alter a compound’s exposure pattern. Consequently, this was not a trial of smoked cannabis, unformulated CBD oil, ordinary food-derived CBD, or clinical use in people. The formulation and administration route are part of the exposure definition.
Investigators examined kidney sections using standard microscopic histopathology and immunohistochemical staining. Histopathology was graded with a semiquantitative score. Additional markers were Ki-67, cleaved caspase-3, and glutathione S-transferase pi (GST-P). None of those procedures directly measured the kidney’s filtering capacity.
What happened in the five rat groups?
The published histological score table provides the clearest view of the pattern. Scores ran from 0 to 3; the publication reported statistically significant control-group comparisons for the two 5 mg/kg groups, not the two 2.5 mg/kg groups. The counts below are numbers of animals assigned a particular microscopic score, not percentages of renal function lost.
| Exposure group | Published renal histology scores (n = 8) | Interpretation within this experiment |
|---|---|---|
| Vehicle control | Eight scored 0 | Reference group; no cannabinoid in the vehicle. |
| CBD nanoemulsion, 2.5 mg/kg | Four scored 0; four scored 1 | Descriptive score-1 observations; authors reported no statistically significant difference from control. |
| CBD nanoemulsion, 5 mg/kg | Eight scored 3 | Higher microscopic scores; statistically different from control as reported. |
| CBD + THC nanoemulsion, 2.5 mg/kg | Eight scored 1 | Scores were not all zero; nevertheless authors reported no statistically significant control contrast. |
| CBD + THC nanoemulsion, 5 mg/kg | One scored 2; seven scored 3 | Higher microscopic scores; statistically different from control as reported. |
Reading the table correctly: A score describes observed tissue appearance, not a blood drug level, creatinine result, measured glomerular filtration rate, or clinical diagnosis. Because the low-dose combination group had score-1 observations even though its comparison was reported nonsignificant, “no statistically significant difference” should not be shortened to “nothing happened.” The exact grading and statistical procedures matter.
The authors describe higher-dose findings including tubular and glomerular changes, hemorrhagic areas, inflammatory infiltration, and focal necrosis. These are microscopic descriptions from an animal model. They should not be rephrased as a quantified rate of human acute kidney injury.
What do Ki-67, cleaved caspase-3, and GST-P measure?
The study’s immunohistochemistry measured the expression or distribution of selected tissue markers. Each marker provides a different window into a biological process, and none is synonymous with organ failure.
| Tissue marker | What staining may indicate | What it does not establish by itself |
|---|---|---|
| Ki-67 | Cells engaged in the cell cycle; increased staining may reflect changes in turnover or a response to tissue injury. | Malignancy, irreversible damage, or diminished kidney filtration. |
| Cleaved caspase-3 | Activation of an apoptosis-related pathway in stained tissue. | The amount of total organ dysfunction or the final clinical outcome. |
| GST-P | An enzyme-associated tissue response that may be consistent with cellular stress or altered detoxification activity. | A precancerous lesion, measured toxin burden, or diagnosed nephropathy. |
The authors reported increased Ki-67 and cleaved caspase-3 staining in the two 5 mg/kg groups relative to controls. GST-P staining increased significantly in the CBD-alone 5 mg/kg group, but not consistently across the other tested regimens. The cleaved caspase-3 score table reports five analyzed animals per group, whereas the main randomized groups contained eight rats each; that distinction limits assumptions about identical denominators across all endpoints.
One method can show an association between exposure and a tissue-marker response without establishing the complete chain of mechanism, sustained injury, clinical dysfunction, and causation in a human patient.
Did the researchers measure kidney function?
No. The paper explicitly states that renal functional parameters such as serum creatinine, urea, and urinary biomarkers were not assessed. Functional indicators are not interchangeable with tissue histology, and one should not invent missing measurements when describing a paper. The investigators also did not establish human renal outcomes or a time course of recovery after exposure ceased.
In a case involving alleged renal injury, a careful analysis would distinguish at least four questions: what exposure or product was documented, what biological matrix or tissue was tested, what renal functional and clinical endpoints were actually measured, and whether the chronology and alternative explanations support causation. A rat histology experiment may generate a scientific hypothesis, but it cannot substitute for medical records or patient-specific testing.
For the related distinction between drug concentrations and their clinical interpretation, see therapeutic drug monitoring in toxicology. Concentration-based extrapolation also requires the exposure and sample-timing considerations explained in pharmacokinetics in forensic toxicology.
Does “no significant difference” at 2.5 mg/kg prove a safe dose?
No. Failure to detect statistical significance in this small, short-duration experimental design is not a demonstration of universal safety. Eight animals per group afford limited precision for uncommon effects. The endpoint was kidney tissue after 21 days, not long-term human use. The study did not establish a human no-observed-adverse-effect level or compare all possible administration routes, doses, formulations, and populations.
Conversely, a significant change in a histology score at the larger experimental regimen does not create a general human toxicity threshold. The relevant unit is the study’s oral rat dosing label, not a validated human clinical cutoff. Even a statistically real difference can leave substantial uncertainty about biological mechanism, clinical significance, and external validity.
Can this study separate CBD’s renal effect from THC’s?
Only incompletely. The study compared CBD-alone groups with CBD + THC combinations but had no THC-only group. It therefore cannot isolate a THC-specific effect from the combined formulation. Nor can it demonstrate that the observed combination effect was caused by THC rather than differences in cannabinoid exposure or interactions. The cannabinoid-containing nanoemulsions were also evaluated against a vehicle formulation, not against every possible commercial formulation.
The authors studied male rats only. Sex-related differences, species differences, metabolism, human conditions, concomitant medications, and different formulations could alter the relationship between administered dose and tissue response. Nothing in the paper quantified an individual user’s blood CBD or THC concentration, psychomotor impairment, or workplace/driving fitness.
Measurement, inference, and causation: a forensic reading framework
A strong scientific opinion identifies the rung of evidence on which a claim rests. This one-paper example makes the hierarchy visible:
- Direct observation: Under the stated 21-day rat experiment, investigators observed and graded microscopic renal structures and tissue-marker staining.
- Study-level association: The 5 mg/kg treatment groups differed from the vehicle controls in reported histological scores and selected immunohistochemical measures.
- Biological interpretation: Such observations are consistent with cellular stress or injury-related responses in this model, although marker specificity and pathway evidence limit mechanistic certainty.
- Unmeasured clinical endpoint: Actual renal function, acute kidney injury diagnosis, reversibility, and human patient outcomes were not measured in the study.
- Unsupported forensic leap: Neither a positive cannabinoid test nor this animal histology study proves a particular person’s kidney disease, impaired performance, or legal causation.
The appropriate conclusion is narrower than either “CBD and THC are proven safe for kidneys” or “CBD and THC cause kidney failure.” Both formulations outrun the endpoint measurements. For further context on matching a conclusion to specimen and method, the forensic toxicology evidence library organizes related guides by analytical question.
Questions and answers
Did the 2026 study find kidney injury?
It reported abnormal renal tissue morphology and selected immunostaining changes in high-dose experimental rat groups. It did not establish the severity or incidence of human clinical kidney injury.
Were blood creatinine, urea, or estimated GFR measured?
No. The authors explicitly acknowledged that they did not collect functional renal markers. No functional values should be inferred from their tissue scores.
Does Ki-67 positivity prove cancer or a permanent lesion?
No. Ki-67 indexes cell-cycle activity and is not a stand-alone diagnosis of malignancy, permanence, or renal functional loss.
What does a significant cleaved caspase-3 result prove?
In this experiment it supports a difference in an apoptosis-related tissue staining endpoint, not the magnitude of organ failure or a proven clinical prognosis.
Why should the low-dose results not be called proof of safety?
Because absence of a reported significant control comparison in a small, 21-day study does not eliminate smaller effects, uncommon harms, delayed outcomes, or human risks. The low-dose groups also had some nonzero histology scores.
Can the experiment tell whether THC alone caused the changes?
No. A THC-only treatment arm was not included, and the data cannot fully separate THC-specific activity from effects of the combined nanoemulsion formulation.
Can a positive THC blood or urine test establish kidney damage?
No. Drug detection or concentration, tissue response, renal dysfunction, and individual causation are different propositions. Supporting clinical evidence would have to be evaluated separately.
Does this research establish a human impairment cutoff?
No. The researchers did not measure driving, workplace performance, psychomotor testing, or human cannabinoid concentrations. A kidney histology endpoint cannot be converted into an impairment threshold.
Primary scientific source
Guedes Pinto T, Rosario BdA, de Aquino PEA, et al. “Subchronic Exposure to Low-Dose Cannabidiol (CBD) and Δ9-Tetrahydrocannabinol (Δ9-THC) Nanoemulsions Suggests Low Renal Toxicity: A Preliminary Study.” Pathophysiology. 2026;33(3):65. https://doi.org/10.3390/pathophysiology33030065.
Scope of review: This is an educational interpretation of the identified study, not a clinical treatment recommendation or an opinion about any particular patient, product, test result, case, or cause of disease. The original experimental methods, specimen scope, and study limitations control the interpretation.