READER RESOURCES: THE APOCALYPSE ALMANAC: Hidden cures in our dystopian age. Check out the “Cure Cancer in Your Kitchen” chapter. FULLSCRIPT SUPPLEMENTS: top quality and economical. AFFILIATE STORE: I offer competitively priced products HEREthat I have personally tested and used.
● Sinemet (carbidopa-levodopa) is the most effective medication for Parkinson’s symptoms. Within three to ten years, most patients taking it develop dyskinesia, the involuntary writhing that occurs when the drug is working.
● Dyskinesia appears in about 30% of patients by 3 years and in most long-term users, though 10% to 20% never develop it. Younger patients develop it sooner and more severely.
● Dyskinesia is not untreatable. Amantadine, deep brain stimulation (DBS), and smaller, more frequent doses can reduce it. Nothing erases it.
● Mucuna pruriens, the velvet bean, contains levodopa. Its seed is 4% to 6% levodopa by weight, the same molecule in Sinemet, grown in a pod rather than made in a plant.
● In head-to-head trials, the bean matches Sinemet in motor benefit and, in animal and short-term human studies, causes less dyskinesia. No trial has run long enough to prove the bean avoids the three-to-ten-year problem.
● Doses, brands, websites, and phone numbers are listed at the end.
Every Parkinson’s patient reaches the same fork. The tremor and the stiffness are winning, and there is one drug that reverses them like nothing else in medicine. It is cheap, it is old, and it works the first day you take it. The catch arrives later, and by then the drug has taken over your day and is hard to walk away from.
Levodopa (L-dopa) is the precursor the brain converts into dopamine, the signaling chemical Parkinson’s destroys. Sinemet pairs it with carbidopa, a helper that prevents the body from breaking down levodopa before it reaches the brain. Swallow it, and the tremor eases, the stiffness loosens, the frozen face moves again. No other Parkinson’s treatment does this much. Neurologists reach for it first because it is the one that works.
Levodopa reached patients in the late 1960s and remade the disease overnight. It turned frozen people back into moving ones, and nothing since has matched that first effect. Every newer Parkinson’s drug is measured against it, and every one falls short in raw power to relieve symptoms.
For the first few years, the effect is smooth. A dose lasts for hours, the benefit is even, and a patient looks almost well. Doctors call this the honeymoon. It ends.
Dyskinesia is the involuntary writhing, bobbing, and swaying that emerges after years of levodopa. It is not the tremor of Parkinson’s but the opposite problem, too much movement instead of too little. It arrives on the schedule the disease and the drug set together.
The rate is steady and grim. About 10% of treated patients develop dyskinesia each year, and prevalence rises by roughly 10% annually during the first seven years of treatment.1 By 3 years, nearly 30% have it. In advanced disease, the figure reaches 80%. Between 10% and 20% of patients never develop it, which means the writhing is common but not certain.2 Younger patients draw the short straw: in one analysis, 70% of those who started between 40 and 49 had dyskinesia by 5 years, compared with 42% of those who started between 50 and 59.2
The mechanism explains why the natural route might behave differently. Levodopa has a short half-life. A pill spikes the brain’s dopamine, and the level crashes before the next dose. Healthy dopamine neurons smooth these swings by storing and releasing dopamine steadily. In Parkinson’s, those neurons are dying, so the storage tank shrinks year by year, and the brain feels every peak and trough of every pill. This on-and-off pulsing rewires the movement circuits until they misfire as dyskinesia.2
The damage tracks the disease as closely as the drug. Patients in the early stages show no dyskinesia, while most in the late stages do. Higher starting doses, above 600 mg a day, bring it on sooner.1 Wearing-off accompanies it. As the years pass, each dose covers less of the gap until the next, and the dose that once lasted four hours now lasts two.
The two problems close in from both sides. The window in which a dose helps without wearing off or causing dyskinesia shrinks year by year, and dosing becomes a daily negotiation with a moving target.
Dyskinesia responds to treatment. Amantadine, an old antiviral repurposed for Parkinson’s, reduces dyskinesia scores by about a quarter and is the most effective drug for it.2 An extended-release form, Gocovri (amantadine), has specific approval for it. Deep brain stimulation, in which a surgeon places electrodes in the movement circuits, quiets dyskinesia in patients who have run out of drug options. Dividing the same daily levodopa into smaller, more frequent doses, or delivering it as a continuous gel through a pump, softens the peaks that drive it.2
None of these is a cure. Amantadine helps some patients and does nothing for others. The surgery is major, the pump is a commitment, and every option manages a problem the drug created rather than eliminating it. Untreatable overstates the case. Poorly treated is the accurate charge, and it is bad enough.
Yoho Comment: The textbooks undersell dyskinesia. A man rigid and slow keeps his place at the table. A man thrown around by his own nervous system cannot. Neurologists score it as the lesser evil because their scales reward movement over stillness, and dyskinesia is movement. That is a scoring artifact dressed as mercy. Anyone weighing Sinemet should price this in now, while there is still room to choose, before year ten arrives and the choice is gone.
Mucuna pruriens, the velvet bean, has been used in Ayurvedic medicine for centuries. Its seed contains 4% to 6% levodopa by weight, sometimes as high as 7%.6 That is the same levodopa found in Sinemet. A patient who takes the bean is taking levodopa, grown in a pod rather than made in a plant. This single fact changes the whole comparison, because most of what is said for and against the two is about delivery rather than the molecule.
The trials are small but consistent. In a 2004 double-blind study of 8 patients, published in the Journal of Neurology, Neurosurgery, and Psychiatry, a single 15- or 30-gram dose of Mucuna powder worked faster than standard levodopa, lasted longer, and produced less dyskinesia.3 In 2017, a double-blind crossover study of 18 advanced patients in the journal Neurology compared high-dose Mucuna with the usual levodopa and found equal or better motor benefit with fewer dyskinesias and no worse tolerance.4
The longest and newest test matters most for the dyskinesia question. In 2026, researchers conducted a 12-month trial of 32 untreated patients in Ghana, published in the Journal of Parkinson’s Disease. One group took Mucuna, about 37 grams a day; the other took standard levodopa, about 410 mg a day. Motor improvement was the same in both groups, with no statistically significant difference. Motor complications, the category that includes dyskinesia, rose by a small and equal amount in each group over the year.5
That trial is the definitive test of the central claim, and it splits the difference. The bean did not eliminate dyskinesia. It matched levodopa in benefit and early motor complications, but caused more nausea and diarrhea: 56% of the bean group had side effects and 12.5% quit, compared with 37.5% and none in the levodopa group.5
The evidence that Mucuna avoids dyskinesia is strong in animals and thin in people. In monkeys and rats given the bean for up to a year, dyskinesia was mild or absent, whereas animals on synthetic levodopa developed it. A patent on Mucuna preparations states that the treatment does not produce dyskinesia.6 Those results point somewhere, and they align with a mechanism: whole-seed Mucuna is absorbed more slowly than a bare pill, so the peaks are lower, and the seed includes other compounds a purified tablet lacks.
The human proof is missing. Every human trial showing less dyskinesia was either a single dose or a few months long. Dyskinesia takes three to ten years to develop. No one has run Mucuna against Sinemet in people for that long, so the belief that the bean escapes late writhing remains unproven. It is a reasonable inference from the animal data and the mechanism, but it sits a tier below the animal findings and well beneath the human numbers. The closest available 12-month trial showed the same early motor complications in both arms.
What would settle the question is a trial no one has funded. Take several hundred patients, put half on Sinemet and half on standardized Mucuna, follow them for ten years, and score dyskinesia with raters blind to the arm. That study would cost money and would help sell a product no company owns. Until it exists, the bean rests on strong animal data, a clean short-term human record, and a plausible mechanism. That is a fair place to stand, as long as no one dresses it up as proof.
One correction to a common belief. The bean is not weaker than the pill. Milligram for milligram of levodopa delivered, Mucuna matches Sinemet and, on some measures, beats it because natural L-dopa reaches the brain slightly more efficiently in animal studies.6 What limits the bean is dosing rather than strength. The levodopa content of a seed varies from batch to batch, from 0.90% to 6.4% across sources, so a steady dose is harder to hit, and the large powder volumes upset the stomach.7
Yoho Comment: A bean you grow cannot be patented, and a molecule that cannot be patented gets no trials, no sales force, and no journal spread. That is why a plant used for 3,000 years remains in the supplement aisle. The synthetic version, the same molecule with a factory behind it, is the standard of care. The science gap is no accident. Pharma makes the business decision, and the patient pays in dyskinesia.
The fava bean, or broad bean (Vicia faba), is the other natural source of levodopa. The bean, its pods, and, above all, the young sprouts contain it. Small studies have shown that a plate of favas lengthened the time to response in patients with response swings.8 The dose from food is low and varies with the plant, the soil, and the cooking. Favas work better as a supplement to a known dose than as a stand-alone treatment. Anyone with the enzyme deficiency called G6PD should avoid them, since favas trigger a dangerous breakdown of red blood cells in those patients.
Sinemet comes as carbidopa-levodopa tablets, most often 25/100, meaning 25 mg carbidopa and 100 mg levodopa. A common starting dose is one-half to one tablet three times a day, with a neurologist setting the schedule. From there, it is titrated upward to control symptoms, often to 3 to 6 tablets a day across the waking hours. The immediate-release form is taken every 3 to 4 hours because that is how long one dose lasts. Controlled-release and extended forms extend the dosing interval.
Dose mucuna by its levodopa content; that is the number to track. The Ayurvedic product Zandopa (HP-200) comes in a 7.5-gram sachet containing 250 mg of levodopa. Stir into water, never milk, and drink at once.6 To match a Sinemet regimen of 25/250 throughout the day requires roughly 30 grams of the powder, about four scoops. To match a single 250 mg levodopa dose, you need about 1000 mg of Mucuna levodopa, or four sachets. Standardized capsules run leaner: a 15% extract delivers about 120 mg of levodopa per two capsules. Start low, at one dose, and titrate up over weeks, as you would with the pill, every 3 to 4 hours while awake.
One practical rule separates the bean from the pill. Mucuna is levodopa without carbidopa, so more of each dose is wasted before it reaches the brain, and doses are larger. Some patients add a small amount of carbidopa or a low-dose Sinemet to stretch the bean, which cuts nausea and lowers the volume. This is the most dangerous move in the plan. Piling the bean on top of a full levodopa prescription pushes the total too high and brings on the exact dyskinesia you were trying to avoid. It belongs under a doctor’s eye, with the levodopa from every source added up.
Food fights the drug, and this is true for the pill and the bean alike. Dietary protein breaks into amino acids that compete with levodopa for the same doorways, in the gut and at the blood-brain barrier. A dose taken with a steak does less. Take levodopa, in either form, 20 to 30 minutes before a meal or 1 to 2 hours after.9 Patients with bad swings sometimes push most of the day’s protein to dinner, keeping breakfast and lunch light, so the daytime doses land clean.9
Yoho Comment: Test what you buy, or find a seller who tests it. The velvet bean is medicine at these doses, and a powder labeled at a strength it does not have is a dosing error waiting to happen. This is the one place the natural route is more hazardous than the pill. A Sinemet tablet delivers the dose on the label, every time. A scoop of powder is a guess unless someone measured it.
1. “Levodopa-induced dyskinesia: mechanisms and management,” British Columbia Medical Journal. Annual incidence near 10%, 10% to 20% of patients never affected, amantadine and surgery as management. bcmj.org
2. “Levodopa-Induced Dyskinesia in Parkinson’s Disease: Pathogenesis and Emerging Treatment Strategies,” Cells, 2022. 30% by 3 years, up to 80% in advanced disease, young-onset risk, pulsatile mechanism, amantadine, deep brain stimulation, continuous infusion. mdpi.com
3. Katzenschlager R, et al. “Mucuna pruriens in Parkinson’s disease: a double blind clinical and pharmacological study,” Journal of Neurology, Neurosurgery, and Psychiatry, 2004. Faster onset, longer benefit, and less dyskinesia than levodopa in 8 patients. pubmed.ncbi.nlm.nih.gov
4. Cilia R, et al. “Mucuna pruriens in Parkinson disease: A double-blind, randomized, controlled, crossover study,” Neurology, 2017. Equal or better motor benefit with fewer dyskinesias in 18 advanced patients. neurology.org
5. Cilia R, et al. “Mucuna pruriens in untreated Parkinson’s disease in sub-Saharan Africa: A 12-month, multicenter, randomized, controlled trial,” Journal of Parkinson’s Disease, 2026. Equal motor benefit and equal early motor complications versus levodopa, with more gastrointestinal side effects. journals.sagepub.com
6. “Mucuna and Parkinson’s Disease: Treatment with Natural Levodopa.” Seed levodopa content, Zandopa dosing, animal dyskinesia data, the carbidopa question, and safety. intechopen.com
7. “Plan B: Itchy velvet beans, Mucuna pruriens,” The Science of Parkinson’s. Levodopa variability from 0.90% to 6.4%, preparation losses, and cautions. scienceofparkinsons.com
8. Rijntjes M. “Knowing Your Beans in Parkinson’s Disease,” Parkinson’s Disease, 2019. Vicia faba as a natural source of levodopa. hindawi.com
9. “To restrict or not to restrict? Practical considerations for optimizing dietary protein interactions on levodopa absorption in Parkinson’s disease,” npj Parkinson’s Disease, 2023. Timing of 20 to 30 minutes before meals, amino-acid competition, and protein redistribution. nature.com
10. “The Science of Mucuna Pruriens for Treating Parkinson’s,” American Parkinson Disease Association. Mainstream caution on product standardization and self-use. apdaparkinson.org
Thanks to my readers for all their help and appreciation. Sharing is love.
Below the following paywall, I tell you where to get these medications and who to call. You need a paid subscription for this information unless you are financially destitute, in which case you may game the system with a free one-week subscription and take a look.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.