In the largest analysis of American overdose deaths in which kratom was detected, almost every person who died had other drugs in their body as well. That is the best-established fact about kratom and other substances, and both sides of the kratom argument misquote it.
This page sets out what that dataset says, what the single clinical study on the topic measured, and how much of the question has never been studied at all. The last category is the largest.
What the CDC dataset shows
CDC runs a system called SUDORS, which collects detailed records on unintentional overdose deaths from death certificates, medical examiner reports and postmortem toxicology. Olsen and colleagues (2019) reported what it held on kratom.
Across 27,338 overdose deaths recorded between July 2016 and December 2017, 152 decedents tested positive for kratom. That is 0.56 percent of the total. A medical examiner or coroner named kratom as a cause of death in 91 of those 152 cases.
Of those 91, seven had no other substance detected on postmortem toxicology. CDC attaches its own qualification to that number: the presence of additional substances cannot be ruled out, because testing scope varied.
What was found alongside kratom, as a listed cause of death among kratom-positive decedents:
- Fentanyl or a fentanyl analogue, in 65.1 percent
- Heroin, in 32.9 percent
- Benzodiazepines, in 22.4 percent
- Prescription opioids, in 19.7 percent
- Cocaine, in 18.4 percent
Roughly 80 percent of these decedents had a documented history of substance misuse.
What that does not prove, in either direction
This dataset is quoted constantly and accurately almost never.
It does not show that kratom killed 152 people. Two thirds of the kratom-positive decedents had fentanyl listed as a cause of death. Fentanyl kills people who have taken no kratom at all, and a substance appearing in a toxicology screen is not a substance that caused the death.
It also does not show that kratom has never killed anyone. Seven decedents had nothing else detected. That is a small number attached to a real qualification, but it is not zero, and treating it as zero is the mirror-image error.
And it is not a study. It is an administrative collection of coroner and medical examiner findings. CDC states the limitations itself: detection of kratom is method-dependent, so the count may be an underestimate by an amount that cannot be determined, and postmortem toxicology protocols were not documented and varied between and within states. Different jurisdictions were looking for different things with different sensitivity.
What the dataset supports is narrower than either camp wants: when kratom turns up in fatal overdoses, it is nearly always in company, and the company is usually opioids.
Two different things people mean by “interaction”
Worth separating, because they carry different risks and different evidence.
A pharmacodynamic interaction is two substances acting on the body in the same direction at once. Two drugs that each slow breathing will slow it more together. This does not require either drug to change how the other is processed. It is simply addition, and it is the mechanism behind most opioid overdose deaths involving more than one substance.
A pharmacokinetic interaction is one substance changing how the body handles another: absorption, breakdown or clearance. If kratom slows the enzyme that clears a medicine, that medicine builds up to a higher level than the prescriber intended, even though the dose on the packet never changed.
The mortality data above is consistent with the first kind. The clinical research that exists addresses the second.
The one clinical study
Tanna and colleagues (2023) ran the first clinical test of kratom’s pharmacokinetic interaction potential in humans. Twelve healthy adults, six men and six women, took a single low dose of kratom tea made from two grams of a well-characterised leaf powder. Fifteen minutes later they took two probe drugs: midazolam, a standard marker for the enzyme CYP3A, and dextromethorphan, a standard marker for the enzyme CYP2D6. Both enzymes handle a large share of prescription medicines, including many opioids and benzodiazepines.
The results were mixed, and the mixture is the interesting part.
Kratom had no measurable effect on dextromethorphan. The CYP2D6 pathway looked untouched.
Kratom did raise midazolam exposure. Total exposure rose about 39 percent and peak concentration about 50 percent. Midazolam’s half-life did not change, which pointed the authors at the gut rather than the liver: kratom appeared to be inhibiting CYP3A in the intestinal wall as the drug was absorbed, rather than slowing its clearance afterwards.
Now the part that gets left out. Laboratory work on isolated tissue had predicted that mitragynine would raise midazolam exposure by roughly 5.7-fold. The measured figure in actual people was 1.39-fold. The test-tube prediction overstated the real effect by a wide margin, which is a useful reminder about how much weight in-vitro findings can carry on their own.
The limits of the study are substantial and the authors are direct about them. Twelve people. A single dose. Two grams, which is at the low end of what people report using. Healthy volunteers, not people on long-term medication. Two probe drugs standing in for entire enzyme families. It says nothing about repeated use, nothing about larger amounts, and nothing about what happens when kratom meets an opioid or a benzodiazepine.
The authors’ own conclusion was that co-consuming kratom with certain drugs extensively metabolised by CYP3A may precipitate serious interactions. That is a statement about plausibility, not a measured risk.
Case reports, and what they can carry
Below the clinical study sits a scatter of individual reports. The EUDA drug profile records adverse interactions involving kratom taken with carisoprodol (a muscle relaxant), modafinil (a wakefulness-promoting drug), propylhexedrine (a stimulant found in nasal inhalers) and Datura stramonium (a plant with strong anticholinergic effects). It also records a fatal American case involving a blend sold as a herbal drug that contained kratom together with fentanyl, diphenhydramine, caffeine and morphine.
A case report describes one person. It shows that something can happen. It cannot show how often, and it is biased toward the dramatic, because unremarkable outcomes do not get written up. Do not read this list as a complete inventory of dangerous combinations. Read it as proof that the category is not hypothetical.
The interaction you did not choose
One route runs through the product rather than the person taking it.
The Swedish “Krypton” deaths are the clearest example: nine fatal intoxications in under a year from a preparation sold as kratom that had the opioid O-desmethyltramadol added to it, with no indication on the label. Everyone who took it experienced a drug combination they had not agreed to. Concentrated and synthetic 7-hydroxymitragynine products raise a newer version of the same problem, and the FDA’s position is that these are materially different from leaf powder.
Both are covered in more detail in our guide to what testing has found in kratom products. The point for this page is that “what did you take” and “what was in it” are separate questions, and only the first one is under a reader’s control.
What nobody has studied
There is no controlled human study of kratom taken with opioids, benzodiazepines or alcohol.
There is unlikely ever to be one. You cannot ethically give volunteers a combination you suspect may suppress their breathing in order to measure how far it does. That is not a gap in the literature that better funding will close. It is a permanent limit.
The consequence is uncomfortable and should be said plainly: on the combination most likely to kill someone, the best available evidence is coroners’ reports. Everything stronger is either about single drugs, about healthy volunteers taking small amounts, or about cells in a dish.
The honest position
Reasonably supported. Kratom appears in fatal overdoses overwhelmingly in combination with other substances, most often opioids. Kratom inhibits intestinal CYP3A enough to measurably raise the blood levels of at least one drug processed by that enzyme. Individual case reports document harmful interactions with several specific substances.
Not established. Whether kratom meaningfully increases the danger of opioids or benzodiazepines, and by how much. Whether the CYP3A effect matters at the doses people actually take. Whether kratom alone, in the absence of anything else, presents a serious overdose risk. The seven decedents with nothing else detected are not enough to answer that, and neither is their small number enough to dismiss it.
This page is not medical advice and cannot be. If you take prescribed medication, the person who can actually answer the question for your situation is your doctor or your pharmacist, and they can only answer it if you tell them what you are taking. That conversation is awkward. It is also the only version of this question that comes with a real answer attached.
If we have missed relevant research, please tell us at editorial@wikikratom.com.