Soursop Safety: Annonacin, Neurotoxicity and Dose Limits

This is the most important page in this set, and it is the one that almost never appears alongside soursop marketing. Soursop leaf contains a family of compounds called annonaceous acetogenins — the most studied of which is annonacin — that shut down the first enzyme complex of the mitochondrial respiratory chain. In the Caribbean island of Guadeloupe, neurologists documented an unusually high local burden of an atypical parkinsonism resembling progressive supranuclear palsy, and linked it in case-control work to heavy lifelong consumption of Annona fruit and leaf infusions. Laboratory studies then showed that annonacin kills dopaminergic neurons, produces the matching brain lesions in rats, and induces the same kind of tau pathology found in the patients.

None of that makes soursop fruit dangerous to eat occasionally, and this page will not pretend otherwise. What it does mean is that drinking soursop leaf tea every day, indefinitely, is the exposure pattern the concern is about — and that concentrated graviola capsules and extracts are a larger version of the same exposure with an even thinner safety case. If you take one thing from this page, take that sentence.

It is also worth being straight about what is not established: the human epidemiology has not proven causation, it rests on a modest number of studies from a single island using retrospective dietary recall, and other exposures were present in that population. That uncertainty is discussed at length below rather than glossed over. Uncertainty about the epidemiology does not, however, erase a well-characterised toxicological mechanism — and a mechanism with no known safe chronic dose is a reason for caution, not a reason to wait for proof.

What Annonacin Actually Is
Complex I, and Why Neurons Are the Vulnerable Cell
The Guadeloupe Story, in Order
What the Laboratory Work Showed
What Is Genuinely Uncertain
Where the Acetogenins Are: Leaf, Seed, Bark, Pulp
Does an Infusion Actually Extract Them?
Why the Concern Is Cumulative, Not Acute
Dose and Duration: What Can and Cannot Be Said
Capsules Concentrate What Tea Infuses
Who Should Avoid It Entirely
Drug and Condition Interactions
What Symptoms Would Even Look Like
If You Have Been Drinking It Daily for Years
Key Research Papers
Connections

What Annonacin Actually Is

Annonaceous acetogenins are a chemical family found almost exclusively in the custard-apple family, Annonaceae. Structurally they are long fatty-acid-derived chains — typically 35 or 37 carbons — carrying one or more tetrahydrofuran rings in the middle and a small α,β-unsaturated lactone ring at one end. Over two hundred have been isolated from Annona muricata alone. Alongside annonacin, the commonly named ones include annomuricin, muricatocin, annomutacin, murisolin, bullatacin and squamocin.

Three physical properties matter more than the structure diagrams:

They are lipophilic. They dissolve in fat far better than in water, which means they distribute into fatty tissue, cross cell membranes readily, and — critically — cross the blood–brain barrier.
They are potent. Acetogenins inhibit their target enzyme at nanomolar to low-micromolar concentrations in cell systems. This is not a weak plant constituent that only matters in absurd quantities; it is one of the more potent naturally occurring inhibitors of its target that has been described.
They are chemically stable enough to survive food preparation. Boiling a leaf does not destroy them, and drying does not either. A dried-leaf teabag stored for a year still contains them.

It is worth noting where else this compound class shows up: powdered Annona seed has been used across the plant's entire range as an insecticide, a head-lice treatment and a fish poison. That is not a folk curiosity to be dismissed — it is the same chemistry doing the same thing to a smaller organism, and it is the most honest single fact about the acetogenins available to a general reader.

Complex I, and Why Neurons Are the Vulnerable Cell

Nearly all of your ATP is made in mitochondria by the electron transport chain, and the entry point to that chain is complex I — formally NADH:ubiquinone oxidoreductase. It accepts electrons from NADH generated by the Krebs cycle and passes them onward, pumping protons across the inner mitochondrial membrane in the process. That proton gradient is what ATP synthase later converts into chemical energy.

Acetogenins bind complex I and stop it. Downstream, three things happen at once: ATP production falls, electrons back up and leak to form reactive oxygen species, and the NADH/NAD⁺ balance shifts. A cell with spare metabolic capacity can compensate for a while by leaning on glycolysis. A cell without that flexibility runs down.

Dopaminergic neurons of the substantia nigra are close to a worst case. They are enormous by neuronal standards, with axons that branch into hundreds of thousands of terminals; they are unmyelinated over much of that length, so they pay a large energetic cost simply to maintain their resting membrane potential; they cycle a reactive neurotransmitter that generates oxidative stress as a by-product of its own metabolism; and they do not divide, so a lost neuron is lost permanently. Their baseline energy demand runs so close to capacity that they are the classic casualty of any metabolic insult.

The strongest reason to take the soursop question seriously is that this exact mechanism has already produced parkinsonism twice, unambiguously, by other routes:

MPTP. A contaminant in an illicitly synthesised opioid in California in the early 1980s. Its metabolite MPP⁺ inhibits complex I, and young people who injected it developed severe, permanent, levodopa-responsive parkinsonism within days to weeks. This is the single most decisive piece of evidence that a complex I inhibitor can cause human parkinsonism.
Rotenone. A plant-derived pesticide and another complex I inhibitor, which reproducibly produces nigral degeneration with α-synuclein-positive inclusions in rats and has been associated with Parkinson's disease risk in agricultural cohort studies.

Annonacin belongs to the same pharmacological category. In cell-culture comparisons it has been reported to be toxic to dopaminergic neurons at lower concentrations than rotenone requires, and unlike MPTP it does not need to be metabolically activated by glia first. It is also more lipophilic, which favours accumulation.

The Guadeloupe Story, in Order

The story is genuinely complicated, and both the alarmist and the dismissive retellings of it are wrong. Told in sequence, it holds together as an unresolved but serious signal.

1. The clinical observation

Through the 1990s, neurologists on the French Caribbean island of Guadeloupe noticed something odd about their movement-disorder clinic. A strikingly high proportion of parkinsonian patients did not have ordinary Parkinson's disease. Instead they presented with rigidity and slowness without the classic asymmetric rest tremor, early falls and postural instability, difficulty moving the eyes vertically, frontal cognitive change, and prominent problems with speech and swallowing. Clinically this is the picture of progressive supranuclear palsy.

The decisive detail was the treatment response. These patients did not improve on levodopa — the drug that reliably, sometimes dramatically, helps true Parkinson's disease. In the referral series described, atypical cases outnumbered classic Parkinson's disease, which inverts the pattern seen almost everywhere else in the world.

2. The case-control study

In 1999, Caparros-Lefebvre and Elbaz published a case-control study in The Lancet on behalf of the Caribbean Parkinsonism Study Group, asking whether consumption of tropical plants could explain the cluster. Patients with atypical parkinsonism reported substantially higher lifetime consumption of Annonaceae — soursop fruit, soursop leaf infusions and related Annona species — than controls did. The association was strongest for the heaviest and longest consumers.

3. The pathology

A case-control association based on dietary recall is a weak instrument on its own. What raised the stakes was the neuropathology. In 2002, Caparros-Lefebvre, Sergeant, Lees and colleagues characterised the Guadeloupean cases in Brain as a cluster of progressive supranuclear palsy-like tauopathy — that is, the brains showed abnormal accumulation of tau protein in a distribution resembling PSP. This mattered because it gave the syndrome a specific molecular signature that a toxin would need to be able to reproduce.

4. The follow-up phenotyping

In 2007, Lannuzel, Höglinger, Verhaeghe and colleagues published further work in Brain examining whether the two clinical phenotypes seen on the island — the PSP-like syndrome and a parkinsonism with dementia — shared a common risk factor. The framing of that paper is worth noting: by then the question had moved from "is there a cluster" to "what does the shared exposure explain".

5. The regulatory response

French food-safety authorities reviewed the evidence on annonaceae-based products and issued caution, advising against the consumption of concentrated preparations. That is a notable step: a food-safety agency does not restrict a traditional food lightly, and the position was taken on a mechanism-plus-association basis rather than on proof of causation. The relevant agency today is ANSES.

What the Laboratory Work Showed

The epidemiology would be much easier to dismiss without the biology. The biology is where this signal earns its weight, and it arrived in four steps that each closed a gap in the argument.

Annonacin kills dopaminergic neurons, and does so by starving them. Lannuzel, Michel, Höglinger and colleagues reported in Neuroscience in 2003 that annonacin is toxic to mesencephalic dopaminergic neurons specifically through impairment of energy metabolism — and that supplying an alternative energy substrate reduced the toxicity, which is the experiment that pins the mechanism to ATP depletion rather than to some incidental effect. Evidence tier: preliminary, cell culture.
Annonacin produces the right lesions in a living animal. Champy, Höglinger, Féger and colleagues reported in the Journal of Neurochemistry in 2004 that annonacin infused systemically into rats caused nigral and striatal neurodegeneration — the anatomical pattern the hypothesis required. A compound that killed neurons in a dish but produced no lesion in an animal would have ended the story here. Evidence tier: preliminary, animal.
Annonacin produces the right kind of pathology, not just cell death. Escobar-Khondiker, Höllerhage and colleagues reported in the Journal of Neuroscience in 2007 that annonacin caused tau pathology in cultured neurons — redistribution of tau from axons into the cell body, the direction of change seen in tauopathies. This connected the compound to the specific molecular signature found in the Guadeloupean brains. Höllerhage, Matusch, Champy and colleagues extended the argument in Experimental Neurology in 2009, proposing natural lipophilic complex I inhibitors as candidate toxins for sporadic tau pathologies more generally. Evidence tier: preliminary, cell culture and animal.
The quantities involved are not absurd. Champy and colleagues published a quantification study in Movement Disorders in 2005 measuring acetogenin content in Annona muricata material and estimating what a real consumer ingests from a fruit or a cup of infusion. The estimate that a habitual consumer could accumulate, over years, into the range that produced neurodegeneration in the rat experiments is the crucial bridge between the epidemiology and the toxicology. It is also, honestly, the most contestable step in the whole chain, because it extrapolates across species and across routes of administration.

Read together, the laboratory work establishes something narrower but firmer than the newspaper version of the story: annonacin is a neurotoxin with a characterised mechanism, a demonstrated anatomical target, and a pathological signature that matches the human syndrome. Whether it caused the Guadeloupean cases is a separate question.

What Is Genuinely Uncertain

Overstating this would be as much of a failure as ignoring it. The honest summary is that the hypothesis is plausible, biologically supported and unproven. The specific weaknesses:

The epidemiology is small and geographically narrow. A handful of case-control studies from one island. No prospective cohort has tested the hypothesis, and it is hard to see how one ethically could.
Dietary recall is a weak instrument. Participants were asked about lifetime consumption of local foods, retrospectively, in a setting where the hypothesis eventually became locally known. Recall bias is a real concern and cannot be excluded after the fact.
Other exposures were present. Guadeloupe has a history of agricultural pesticide use — including organochlorines used in banana cultivation — and other traditional plant preparations are consumed alongside Annona. Disentangling one dietary exposure from a whole environment is genuinely difficult.
Genetics cannot be excluded. A relatively isolated population can carry risk variants that concentrate a rare phenotype without any environmental trigger.
The animal experiments used a different exposure route. The rat lesion studies infused purified annonacin systemically. Drinking an infusion involves oral absorption, first-pass hepatic metabolism, and a mixture of hundreds of compounds rather than one. Oral bioavailability of annonacin in humans is not characterised.
Hundreds of millions of people eat Annona fruit across Latin America, West Africa and Southeast Asia without a comparable cluster having been reported. That is a real counterargument. It is weakened, but not eliminated, by the observation that PSP-like syndromes are substantially under-diagnosed wherever specialist movement-disorder services are scarce, and that the implicated exposure was heavy daily medicinal use rather than ordinary fruit-eating.
Similar clusters elsewhere have other explanations. The parkinsonism–dementia complex of Guam has been attributed at various times to cycad consumption and to other factors, and the history of that hypothesis is a cautionary tale about confident environmental attribution.

So how should a reader weigh it? The way you would weigh any hazard with a strong mechanism and inconclusive human data: by asking what you gain by accepting the risk. If soursop leaf tea were the only treatment for a serious illness, this would be a hard trade-off. It is not. Every use for which people drink it daily is either unsupported in humans (cancer), supported only in animals (glucose, blood pressure), or traditional (sleep, digestion) with gentler and better-studied alternatives available. The expected benefit is low, the mechanism of harm is well characterised, and the harm is irreversible. That asymmetry, not certainty about Guadeloupe, is the argument.

Where the Acetogenins Are: Leaf, Seed, Bark, Pulp

Every part of Annona muricata contains acetogenins, but not in the same amounts. The ordering is consistent across the phytochemical literature:

Seeds — highest. By a wide margin. This is why the powdered seed works as an insecticide and fish poison. Seeds should never be consumed internally in any amount, and fruit should never be blended whole.
Roots and bark — high. Used in some traditional decoctions; these are the preparations with the least justification of any.
Leaves — high enough to matter. Substantially more concentrated than pulp. The leaf is the part sold as tea and the part packed into capsules.
Fruit pulp — lowest. Present but dilute, and consumed inside a large volume of water, fibre and sugars.

This gradient is the single most practically useful fact on this page, because it maps directly onto how the plant is used. The part with the food tradition is the part with the least acetogenin; the part being sold as medicine is the part with more. Almost all of the popular framing has this backwards, treating "it's just the leaf, not the fruit" as reassuring.

One further complication: species substitution. Annona squamosa (sugar apple, sweetsop, srikaya), A. reticulata (bullock's heart), A. cherimola (cherimoya) and A. montana (mountain soursop) all contain acetogenins and are traded interchangeably in markets and, sometimes, in supplements. A. montana has been reported to contain higher levels than A. muricata. A label reading "custard apple" or simply "Annona" tells you almost nothing about which species is in the bag — but it does tell you that acetogenins are present.

Does an Infusion Actually Extract Them?

A reasonable objection: acetogenins are lipophilic, and tea is water. Shouldn't they stay in the leaf?

Partly, and this is a real reason a cup of tea is a smaller exposure than a capsule of powdered leaf. But "lipophilic" is not "insoluble", and several factors push the other way:

Traditional preparation is often a decoction, not a brief steep. Leaves simmered for ten to fifteen minutes, sometimes longer, in the Caribbean, West African and Indonesian preparations. Prolonged heat extracts substantially more than a two-minute dunk.
Leaf material is often torn or crushed, which ruptures cells and releases contents directly.
The quantification work was applied to infusions specifically. The 2005 Movement Disorders paper estimated intake from infusions, not only from fruit, because infusions were part of the implicated exposure. The researchers would not have modelled a route that extracted nothing.
Some people drink the leaves as well as the liquid, or blend leaf powder into drinks, which bypasses the extraction question entirely.
Alcohol-based tinctures extract them very efficiently. Ethanol is a good solvent for lipophilic compounds, so a tincture is relatively enriched in exactly the fraction of concern.

The practical conclusion is a hierarchy rather than a yes/no: fruit pulp < brief leaf steep < long leaf decoction < leaf powder capsule < tincture or standardised extract. Position on that ladder, multiplied by years of use, is what determines cumulative exposure.

Why the Concern Is Cumulative, Not Acute

Nobody has ever reported drinking a cup of soursop leaf tea and developing a movement disorder that afternoon. If you have had soursop tea a few times, or eaten the fruit on holiday, this page is not describing your situation.

Chronic neurotoxicity works differently from poisoning, in three ways that make it much harder for an individual to detect:

Lipophilic compounds accumulate. A compound that partitions into fat and crosses the blood–brain barrier does not fully clear between daily doses. Repeated intake builds a tissue burden.
Neuronal loss is silent until it is large. The nigrostriatal system has substantial functional reserve; in Parkinson's disease, motor symptoms typically appear only after a majority of nigral dopaminergic neurons are already gone. Damage accrues invisibly for a long time.
The damage does not reverse. These neurons do not regenerate. Stopping the exposure stops further injury; it does not recover what was lost.

Put together: a daily habit produces no feedback signal at all until the reserve is exhausted, and by then the decision cannot be revisited. This is the profile of harm that is hardest for individual experience to detect, and it is exactly why an in-principle mechanism deserves weight even without proof from the human data.

Dose and Duration: What Can and Cannot Be Said

There is no established safe dose of soursop leaf, and there is no established therapeutic dose either. Both gaps have the same cause: no adequate human trials exist, so there is nothing from which to derive either figure. Anyone quoting a precise safe daily limit for acetogenin intake is quoting something that has not been established.

What can be said, honestly:

Traditional practice, as description not recommendation, is roughly three to five leaves steeped in 200–250 ml of water, taken occasionally — and in most traditional contexts, for a defined illness episode rather than as a lifelong daily beverage.
Commercial capsule labels commonly suggest 500–1,500 mg of leaf powder daily. These numbers come from manufacturers, not from trials, and they are not safety-derived.
Duration matters more than the size of any single dose. This is the most useful practical framing available. An occasional cup is a different proposition from a cup every night for three years.
Concentrated extracts and "standardised acetogenin" products have the weakest case of all and are best avoided entirely. Standardising a product for its neurotoxic constituent is a strange thing to sell and a stranger thing to buy.
The fruit is a separate question. Eating ripe soursop as part of a varied diet is a reasonable thing to do. The concern scales with concentration and with years, and fruit is at the bottom of both scales.

If you want a single practical rule from this page: do not take soursop leaf, in any form, daily and open-endedly. That is the pattern that matches the implicated exposure, and it is also the pattern that supplement marketing recommends.

Capsules Concentrate What Tea Infuses

It is worth spelling out why the supplement aisle is the higher-risk end of this topic, because "it's natural, it's just the plant in a capsule" sounds reassuring and is not.

A capsule delivers the whole leaf, not an extract of it. With tea, the acetogenins that stay bound in the leaf matter are thrown away with the spent leaves. Swallowing leaf powder discards nothing.
Solvent extracts deliberately enrich the lipophilic fraction. Ethanol, methanol and hexane extractions concentrate acetogenins relative to the water-soluble constituents. This is the opposite of what an infusion does.
Compliance is higher and dosing is easier to escalate. Brewing tea takes effort; taking two capsules with breakfast does not. Products commonly suggest multiple capsules per day, and readers who believe they are treating something serious tend to take more rather than less.
Content is unverified. Botanical supplements are not required to state acetogenin content, and essentially none do. There is no way to know from the label what a given product delivers, and no way to compare two of them.
Species and part are frequently unstated. "Graviola 1000 mg" does not tell you whether you are getting leaf, bark, seed or a mixture, or which Annona species supplied it.

Who Should Avoid It Entirely

These are not graded cautions. Each of these is a reason to avoid soursop leaf tea and all graviola supplements outright.

Any existing parkinsonian or movement disorder. Parkinson's disease, progressive supranuclear palsy, multiple system atrophy, corticobasal syndrome, essential tremor or any undiagnosed movement problem. Adding a dopaminergic neurotoxin to a failing dopaminergic system is the clearest contraindication on this page.
A family history of parkinsonism or other neurodegenerative disease. Genetic susceptibility plus environmental complex I inhibition is the standard model for how these diseases are thought to arise; there is no reason to volunteer for the second half of it.
Any known or suspected mitochondrial disorder. Inhibiting complex I in someone whose respiratory chain is already impaired is a straightforward hazard.
Pregnancy and breastfeeding. No human safety data; lipophilic compounds cross the placenta and enter milk; some Annona preparations carry traditional uterine-stimulant reputations. Avoid tea and supplements both.
Infants and children. The developing nervous system is the worst possible place to test a mitochondrial toxin.
Anyone with a neurodegenerative diagnosis of any kind, including Alzheimer's disease and other tauopathies, given that annonacin induces tau pathology in laboratory models.
Soursop seeds, internally, by anyone. Highest acetogenin content in the plant; traditional use is as a poison.

Drug and Condition Interactions

Levodopa, dopamine agonists and other antiparkinsonian drugs. Theoretical antagonism at the level of the neuron the drug is trying to support. Avoid the combination.
Metformin. Worth singling out. Metformin's own principal action involves mild inhibition of mitochondrial complex I; stacking a second, more potent complex I inhibitor on top of it is a poor idea on mechanism alone, quite apart from additive glucose lowering. See the blood sugar article.
Insulin and sulfonylureas. Possible additive glucose lowering based on animal data; monitor if combined, and never substitute.
Antihypertensives — ACE inhibitors, ARBs, calcium-channel blockers, diuretics, beta-blockers. Possible additive blood-pressure lowering; dizziness on standing is the symptom to watch.
Chemotherapy and radiotherapy. Tell your oncology team. Beyond the general principle, two specific reasons: chemotherapy-induced peripheral neuropathy is already a common and dose-limiting problem, and adding a neurotoxin to it is a genuinely bad idea; and several agents act on or are affected by mitochondrial function.
Sedatives, benzodiazepines, sleep medication and alcohol. Possible additive sedation, given the traditional sleep use.
Liver or kidney impairment. Human clearance of acetogenins is not characterised. Reduced clearance of a compound that accumulates is exactly the wrong combination.
Nausea and vomiting are the commonly reported acute effects of ordinary use, especially on an empty stomach. Movement-disorder effects would not be expected to appear acutely.

What Symptoms Would Even Look Like

An honest answer is that there is no early warning sign specific to acetogenin exposure, and nobody should be self-monitoring for one as a substitute for not taking the tea. But because readers will reasonably ask, the atypical parkinsonism described in Guadeloupe presented with features that differ from classic Parkinson's disease in recognisable ways:

Falls early in the illness, often backwards, rather than after years of disease.
Difficulty with vertical eye movement — particularly looking down, which shows up as tripping on steps or trouble finding food on a plate.
Rigidity and slowness that are symmetrical from the start, and generally without the asymmetric rest tremor typical of Parkinson's disease.
Speech and swallowing difficulty appearing relatively early.
Frontal cognitive and behavioural change — apathy, reduced initiative, impulsivity.
Little or no benefit from levodopa, which is often the observation that prompts re-evaluation of a Parkinson's diagnosis.

These features are not specific to soursop — they describe atypical parkinsonism from any cause, and most people who have them have never touched an Annona. If any of them apply to you, the action is a neurological assessment, not a search of the internet. Mention your dietary and supplement history when you go, because it is the kind of detail nobody thinks to ask about.

If You Have Been Drinking It Daily for Years

This part is written for the reader who has arrived here after finding soursop leaf tea recommended somewhere else, and has been drinking it nightly for a long time. Plainly, without alarmism:

Stop the daily habit. There is no taper needed and no withdrawal to manage; acetogenins are not physically dependence-forming. Stopping halts further exposure, which is the only variable still under your control.
Do not panic, and do not assume damage. The human evidence is an association from a specific population with heavy lifelong intake, not a demonstrated dose–response curve you can locate yourself on. Most long-term consumers of Annona foods worldwide do not develop a movement disorder.
Mention it to your doctor, especially if you have any neurological symptoms. It belongs in your history in the same way a long-term supplement or an occupational exposure does. There is no test for cumulative acetogenin exposure, so the history is the only record of it.
Do not switch to capsules "to be safer". Capsules are the higher-exposure form, not the lower one.
Keep eating the fruit if you enjoy it. That was never the concern.
If you were drinking it for something specific — sleep, blood pressure, glucose, a cancer diagnosis — replace it with something that has evidence behind it rather than simply stopping and hoping. The other three articles on this hub cover each of those cases.

Key Research Papers

Per this site's citation policy, each reference links to a PubMed topic search rather than a numeric identifier, so a transcription error cannot silently point at the wrong paper. Titles, journals and years are stated in the text.

Caparros-Lefebvre and Elbaz, for the Caribbean Parkinsonism Study Group, Possible relation of atypical parkinsonism in the French West Indies with consumption of tropical plants: a case-control study, The Lancet, 1999. The original epidemiological signal. Case-control. PubMed topic search
Caparros-Lefebvre, Sergeant, Lees and colleagues, Guadeloupean parkinsonism: a cluster of progressive supranuclear palsy-like tauopathy, Brain, 2002. The neuropathological characterisation. PubMed topic search
Lannuzel, Michel, Caparros-Lefebvre and colleagues on the toxicity of Annonaceae for dopaminergic neurons and its potential role in atypical parkinsonism in Guadeloupe, Movement Disorders, early 2000s. Cell culture. PubMed topic search
Lannuzel, Michel, Höglinger and colleagues, The mitochondrial complex I inhibitor annonacin is toxic to mesencephalic dopaminergic neurons by impairment of energy metabolism, Neuroscience, 2003. The mechanism paper. Cell culture. PubMed topic search
Champy, Höglinger, Féger and colleagues, Annonacin, a lipophilic inhibitor of mitochondrial complex I, induces nigral and striatal neurodegeneration in rats, Journal of Neurochemistry, 2004. The animal lesion study. Animal. PubMed topic search
Champy and colleagues, Quantification of acetogenins in Annona muricata linked to atypical parkinsonism in Guadeloupe, Movement Disorders, 2005. The intake-estimate study that bridges epidemiology and toxicology. PubMed topic search
Caparros-Lefebvre and Steele on atypical parkinsonism in Guadeloupe compared with the parkinsonism–dementia complex of Guam and environmental toxic hypotheses, mid-2000s. Essential for understanding the counterarguments. PubMed topic search
Escobar-Khondiker, Höllerhage, Muriel and colleagues, Annonacin, a natural mitochondrial complex I inhibitor, causes tau pathology in cultured neurons, Journal of Neuroscience, 2007. Links the compound to the specific pathology. Cell culture. PubMed topic search
Lannuzel, Höglinger, Verhaeghe and colleagues, Atypical parkinsonism in Guadeloupe: a common risk factor for two closely related phenotypes?, Brain, 2007. PubMed topic search
Höllerhage, Matusch, Champy and colleagues, Natural lipophilic inhibitors of mitochondrial complex I are candidate toxins for sporadic neurodegenerative tau pathologies, Experimental Neurology, 2009. PubMed topic search
Reports of atypical parkinsonism and Annonaceae consumption from outside Guadeloupe, including New Caledonia, which are how far the signal has and has not replicated. PubMed topic search
The rotenone and MPTP literature, which establishes that complex I inhibition can cause parkinsonism in animals and in humans respectively — the precedent that gives the annonacin hypothesis its plausibility. PubMed topic search
Coria-Téllez, Montalvo-Gónzalez, Yahia and Obledo-Vázquez, a comprehensive review of Annona muricata traditional uses, phytochemicals, pharmacology, mechanisms and toxicity, published in Arabian Journal of Chemistry, 2018 — useful for the toxicity section specifically. PubMed topic search
Moghadamtousi, Fadaeinasab, Nikzad and colleagues, Annona muricata (Annonaceae): a review of its traditional uses, isolated acetogenins and biological activities, International Journal of Molecular Sciences, 2015. PubMed topic search

Live PubMed Searches

Connections

Safety and disclaimer. This page is health information, not medical advice. Annonacin and related acetogenins in soursop leaf are mitochondrial complex I inhibitors and are neurotoxic to dopaminergic neurons in laboratory models; heavy long-term consumption of Annona products has been associated with atypical parkinsonism in epidemiological work from Guadeloupe. That association is not proven and it is not dismissible. No safe long-term dose has been established. Do not take soursop leaf tea daily on an open-ended basis, and avoid concentrated graviola extracts and capsules. Avoid entirely in pregnancy, breastfeeding, childhood, any parkinsonian or movement disorder, a family history of neurodegenerative disease, and any known mitochondrial condition. Never consume the seeds. If you have neurological symptoms, seek a proper assessment and tell the clinician what you have been taking — do not self-diagnose from a web page.

Back to Table of Contents