Passion fruit: the research is about the peel, not the pulp

The strongest evidence on this page was produced by a part of the plant most people throw away. In a network meta-analysis of 23 randomized trials in 2,455 people with hand, hip or knee osteoarthritis, purple passion fruit peel extract ranked first of 21 supplements for WOMAC pain, with a mean difference of -9.2 points (95% CI -16.0 to -2.3), on evidence the authors themselves graded low and on a small number of trials. A second review, covering 69 trials of 20 supplements, put the peel extract among 7 with a large short-term effect on pain, and found that no supplement held a clinically important effect at medium or long term. Its authors call the overall quality of evidence very low and point out that the biggest effects come from the supplements with the fewest studies. None of this was tested as fruit on a plate. No agency has set a safe upper amount for the peel extract or the peel flour, the only toxicity data are from rats, and the only human safety finding is an absence of reported harm in one 4-week trial. Passion fruit is also among the fruits whose allergens cross-react with latex. What the fruit itself reliably brings is fiber: 10.4 g per 3.5 oz (100 g), 37 percent of the Daily Value, from a single USDA legacy record with no Foundation samples behind it, so the table carries no measured spread. Vitamin C sits at 30 mg per 3.5 oz (100 g) there and at 23 mg in the UK table. That 100 g is not a portion. USDA counts one fruit as 18 g of edible flesh and a cup as 236 g, which is about 13 fruits.

The numbers, per 3.5 oz (100 g)

As people eat it, per 3.5 oz (100 g)

How much is actually in it

• Passion fruit does contain cyanogenic glycosides: in a survey cited by EFSA, cyanide was found in 2 of 5 passion fruit samples, at 4.7 to 6.6 mg per kg. — passion fruit sold in Australia and New Zealand, FSANZ 2014 survey cited by EFSA (n = 5 samples) Evidence E source
  Hydrogen cyanide detected in 2 of 5 samples at 4.7 to 6.6 mg/kg (limit of detection 4 mg/kg)
• Passion fruit is unusually fibrous. The USDA reference record gives 10.4 g of fiber per 3.5 oz (100 g), which is 37% of the Daily Value, while vitamin C differs by 23% between the USDA table (30 mg) and the UK CoFID table (23 mg). — — (n = 1 USDA record, 2 CoFID records) Evidence E source
  Fiber 10.4 g per 3.5 oz (100 g) (37% DV), vitamin C 30 mg (33% DV) in USDA against 23 mg in CoFID

What a real portion looks like

• One passion fruit counts as 18 g of edible flesh in the USDA survey portions, and 1 cup counts as 236 g, so a cup takes about 13 fruits. — — (n = 3 household measures) Evidence E source
  1 fruit = 18 g, 1 cup = 8.3 oz (236 g) of edible portion

What your body absorbs

• Blending passion fruit with mango lowered the glycemic index from 42.56 to 26.30 in adults with obesity (p = 0.044) compared with eating the same fruit whole. — 21 adults completed the passionfruit and mango arm, 10 of healthy weight and 11 with obesity, test meals matched at 25 g total sugar (n = 21) Evidence C source
  GI 26.30 ± 25.72 for the blended meal versus 42.56 ± 20.64 for whole fruit in adults with obesity, p = 0.044

What it does to your health

• Across 23 randomized trials in 2,455 people with osteoarthritis, passion fruit peel extract ranked first for WOMAC pain with a mean difference of -9.2 points (95% CI -16.0 to -2.3). — 2,455 participants with hand, hip or knee osteoarthritis in 23 randomized trials of 21 dietary supplements (n = 2,455) Evidence A source
  WOMAC pain MD -9.2 (95% CI -16.0 to -2.3), SUCRA 91%, and WOMAC function MD -41.0 (95% CI -66.0 to -16.0), SUCRA 99%
• In 33 adults with knee osteoarthritis, 150 mg a day of purple passion fruit peel extract for 60 days cut self-reported pain by 18.6% and the composite WOMAC score by 19.2%. — 33 patients with knee osteoarthritis, 150 mg daily of peel extract or placebo for 2 months (n = 33) Evidence B source
  Self-reported reductions of 18.6% in pain, 18% in stiffness, 19.6% in physical function and 19.2% in composite WOMAC, with no intervals reported in the abstract
• In a 4-week randomized placebo-controlled trial, 150 mg a day of purple passion fruit peel extract lowered the prevalence of wheeze, cough and shortness of breath in adults with asthma (P < .05). — adults with asthma in a 4-week trial of 150 mg/d peel extract or placebo, sample size not stated in the abstract Evidence B source
  Significant fall in wheeze, cough and shortness of breath (P < .05), with no change in FEV1 and no significant improvement on placebo
• In 43 people with type 2 diabetes, 30 g a day of yellow passion fruit peel flour for two months changed fasting glucose (P = 0.000) and glycated hemoglobin (P = 0.032) and lowered HOMA IR (P = 0.005). — 43 volunteers with type 2 diabetes, 28 women and 15 men, each serving as their own control for two months (n = 43) Evidence C source
  Fasting glucose P = 0.000, glycated hemoglobin P = 0.032, HOMA IR P = 0.005, effect sizes not given in the abstract
• In 82 Japanese women aged 30 to 59, a daily drink with 10 mg of piceatannol from passion fruit seeds for 8 weeks raised facial skin hydration compared with placebo. — 82 healthy Japanese women aged 30-59 years analyzed out of 86 enrolled, one drink a day for 8 weeks (n = 82) Evidence B source
  Significant rise in facial stratum corneum hydration against placebo, effect sizes not given in the abstract

Safety, limits and interactions

• The same NIH page rates passionflower as likely safe when used as a flavoring in food. — general population Evidence E source
  Likely safe as a food flavoring
• EFSA set an acute reference dose of 20 mcg of cyanide per kg of body weight and concluded in 2019 that it protects against acute effects of cyanide from cyanogenic glycosides whatever the food it comes from. — all age groups, EU risk assessment Evidence E source
  Acute reference dose 20 mcg cyanide per kg body weight, applicable whatever the food
• The only toxicity data found for passion fruit peel are from rats: a review reports no toxic effect of an unripe peel extract given by mouth at 550 mg per kg. We found no human upper limit for the extract or the peel flour. — rats given ethanol extract of unripe passion fruit peel by mouth Evidence D source
  No toxic effect at 550 mg/kg
• In a 2-month study of 30 g a day of passion fruit peel flour in 43 people with type 2 diabetes, those who dropped out did so because of the strong aftertaste and abdominal discomfort. — 43 people with type 2 diabetes (28 women, 15 men) (n = 43) Evidence C source
  Participants who left the study did so because of the flour's strong aftertaste and abdominal discomfort
• In the 4-week trial of 150 mg a day of passion fruit peel extract in asthma, no participant reported an adverse effect. — adults with asthma taking 150 mg/d of purple passion fruit peel extract for 4 weeks Evidence B source
  Zero adverse effects reported over 4 weeks

What people believe that the data does not support

• The NIH herbal medicine center says passionflower, the medicinal herb Passiflora incarnata, should not be used in pregnancy because it may cause uterine contractions. The warning concerns the herb, and the fruit is a different species. — pregnant women using passionflower herb (tea, extract) Evidence E source
  Avoid the herb in pregnancy; possible uterine contractions
• The NIH herbal medicine center warns against passionflower herb in pregnancy because it may cause uterine contractions, and rates it likely safe as a food flavoring. The warning is about a different species, not passion fruit. — pregnant women using passionflower herb (Passiflora incarnata) Evidence E source
  Avoid the herb in pregnancy; likely safe as a food flavoring
• In a norovirus outbreak that hit at least 240 diners in England, the dish of oyster with passion fruit jelly carried odds of illness 7.0 times higher (95% CI 1.1 to 45.2), and the investigation traced the source to the oyster supply. — at least 240 diners at one restaurant over 7 weeks in 2009, case-control study within the outbreak investigation (n = at least 240 affected diners) Evidence C source
  Odds ratio 7.0 (95% CI 1.1 to 45.2) for the oyster, passion fruit jelly and lavender dish

Who this works differently for

• In RAST-inhibition tests on sera of latex-allergic patients, passion fruit was among the fruits whose allergens cross-reacted with latex. — 136 patients with clinically relevant immediate-type latex hypersensitivity (n = 136) Evidence C source
  IgE cross-reacting between latex and passion fruit, among 12 fruits
• The FDA warns that pregnant women risk serious illness from untreated, unpasteurised juices, which covers fresh-squeezed passion fruit juice sold without treatment. — children, older adults, pregnant women and people with weakened immune systems Evidence E source
  Choose pasteurized or otherwise treated juice
• In children in Australia and New Zealand, passion fruit accounted for about 6% and under 5% of long-term dietary cyanide exposure, far behind linseed bread, linseed and cassava chips. — children (Australia 2 to 16 years, New Zealand 5 to 14 years), FSANZ chronic exposure estimate cited by EFSA Evidence E source
  Passion fruit contributed 6% (Australia) and under 5% (New Zealand) of chronic cyanide exposure in children
• In the peel flour study, glucose-lowering drug doses were not changed during the study, and people with liver or kidney dysfunction were excluded, so the flour was not tested in kidney disease. — people with type 2 diabetes on glyburide, metformin and/or insulin; liver and kidney dysfunction excluded (n = 43) Evidence C source
  Flour taken on top of unchanged glucose-lowering drugs; people with liver or kidney dysfunction were excluded
• Among 136 patients with documented immediate-type latex allergy, 69.1% carried fruit-specific IgE, and passion fruit was one of the fruits whose allergens cross-react with latex. — 136 patients with well-documented immediate-type hypersensitivity to latex proteins (n = 136) Evidence C source
  Fruit-specific IgE in 69.1% of sera, 42.5% reported symptoms after eating these fruits, IgE found in only 32.1% of those with symptoms

What is still unknown

• The only pregnancy study of passion fruit found is in rats: pregnant rats were given a polyphenol-rich passion fruit extract in the last week of gestation and first week of lactation. — pregnant rats and their pups with induced neonatal hypoxia-ischemia Evidence D source
  Outcomes in pups after induced hypoxia-ischemia; not quoted on this card
• A 2020 review of passion fruit research concludes that clinical trials on its efficacy and safety are very scarce. — all published research on Passiflora edulis up to 2020 Evidence D source
  Clinical trials on efficacy and safety very scarce
• The only pregnancy study of passion fruit found is in rats given a polyphenol-rich extract in late pregnancy and early lactation. There are no data in pregnant women or in children for the peel extract or flour. — pregnant rats and their pups with induced neonatal hypoxia-ischemia Evidence D source
  Only pregnancy exposure study found; it is in rats and tests an extract, not the fruit
• Of 20 supplements tested in 69 randomized trials, passion fruit peel extract was one of 7 with a large short-term effect on osteoarthritis pain, and no supplement kept a clinically important effect at long term. — 69 randomized trials of oral supplements against placebo in hand, hip or knee osteoarthritis (n = 69 trials of 20 supplements) Evidence A source
  Large short-term pain effect, effect size above 0.80, and nothing clinically important at medium or long term

The bottom line

The human trials all test a product. In 33 adults with knee osteoarthritis, 150 mg a day of purple passion fruit peel extract for 60 days brought self-reported drops of 18.6 percent in pain and 19.2 percent in the composite WOMAC score, with no confidence intervals in the abstract and 33 people in the trial. A 4-week trial of the same 150 mg dose in adults with asthma lowered the prevalence of wheeze, cough and shortness of breath (P < .05) while FEV1 did not move. The abstract never says how many people were randomized, and no trial on this page tested that extract against or alongside an asthma treatment. In 43 people with type 2 diabetes, 30 g a day of yellow passion fruit peel flour for two months shifted fasting glucose (P = 0.000) and glycated hemoglobin (P = 0.032) and lowered HOMA IR (P = 0.005). The abstract reports a difference rather than a direction for the first two. The study had no placebo group and no randomization, each volunteer served as their own control, and the abstract gives no effect sizes. The skin study used neither peel nor pulp. It gave 82 Japanese women aged 30 to 59 a daily drink with 10 mg of piceatannol from passion fruit seeds for 8 weeks and measured a rise in facial skin hydration against placebo, again with no effect sizes reported. Blending is the one thing tested on the fruit: mixed with mango and matched at 25 g of sugar, the blended meal gave a glycemic index of 26.30 against 42.56 for the same fruit eaten whole in adults with obesity (p = 0.044), in 21 people, with standard deviations almost as wide as the means and passion fruit never fed on its own. On caution the page is thin, and pretending otherwise would be dishonest. For the peel products there is no ceiling. No human study of the extract ran longer than 8 weeks, and the only toxicity figure is from rats: a review reports no toxic effect of an unripe peel extract given by mouth at 550 mg per kg, which does not translate into a safe human dose. The one human safety finding for the extract is a null: no participant in that 4-week asthma trial reported an adverse effect, which is an absence of reported harm in one small short study rather than a safety margin. In the peel flour study, those who dropped out did so because of the strong aftertaste and abdominal discomfort. Its volunteers stayed on glucose-lowering drugs at unchanged doses, the paper does not report low blood sugar episodes, and people with liver or kidney problems were excluded, so the flour was never tested in kidney disease or as a change to diabetes treatment. The fruit itself contains cyanogenic glycosides. In a survey cited by EFSA, cyanide was found in 2 of 5 passion fruit samples at 4.7 to 6.6 mg per kg, without saying whether pulp, seeds or whole fruit were tested. EFSA set an acute limit of 20 mcg of cyanide per kg of body weight. At the highest level found, a 132 lb (60 kg) adult would need about 6.3 oz (180 g) of fruit in one sitting to reach it, our arithmetic rather than an EFSA statement, and because the limit is per kg a child reaches it with proportionally less. In children in Australia and New Zealand, passion fruit made up about 6 percent and under 5 percent of long-term dietary cyanide exposure, far behind linseed bread, linseed and cassava chips. The best-known warning is cross-reaction. Among 136 patients with documented immediate-type latex allergy, 69.1 percent carried fruit-specific IgE, and passion fruit was among the fruits whose allergens cross-reacted with latex in the laboratory, although that is serology rather than a food challenge and the authors say such tests predict food allergy poorly. The only pregnancy study of passion fruit we found is in rats given an extract, and there are no data in pregnant women or in children for the peel extract or flour. The pregnancy warning on the NIH passionflower page concerns a different species, the medicinal herb, not the fruit. We found no interaction study with any drug. If a search points you at an English norovirus outbreak that hit at least 240 diners, read it carefully. The oyster, passion fruit jelly and lavender dish carried odds of illness 7.0 times higher (95% CI 1.1 to 45.2), an interval spanning a factor of forty, and odds rather than risk, and the investigators traced the source to the oyster supply.

Sources

1. hf2-pfru-1 · EFSA scientific opinion
   Table 16 Total cyanide contents in food commodities containing cyanogenic glycosides a (FSANZ, 2014 ) ... Passion fruit 2/5 4.7–6.6 4
   EFSA CONTAM Panel, Evaluation of the health risks related to the presence of cyanogenic glycosides in foods other than raw apricot kernels, EFSA Journal 2019;17(4):5662 · checked 2026-10-08
2. pfru-r5-11 · USDA FoodData Central SR Legacy and CoFID 2021, dataset
   Passion-fruit, (granadilla), purple, raw (FDC 169108, SR Legacy, Fruits and Fruit Juices): fiber 10.4 g per 100 g = 37% DV, vitamin C 30 mg = 33% DV, riboflavin 0.13 mg = 10% DV. No Foundation Foods sample exists for this fruit, so USDA carries no measured spread. CoFID 2021 Passion fruit, flesh and pips [14-178]: vitamin C 23 mg, riboflavin 0.12 mg.
   USDA FoodData Central, SR Legacy, FDC 169108, and CoFID 2021 (Public Health England, OGL v3.0) · checked 2026-09-24
3. pfru-r5-12 · USDA FNDDS, dataset
   Passion fruit, raw (FNDDS, FDC 2709248, WWEIA Other fruits and fruit salads): 1 fruit = 18 g, 1 cup = 236 g. Portion nutrient values are recalculated from the 100 g record, not measured per portion. USDA Table of Nutrient Retention Factors Release 6 (2007) holds no method and no group for this fruit, and fiber is not among its 26 nutrients.
   USDA FNDDS, FDC 2709248, WWEIA category Other fruits and fruit salads · checked 2026-09-24
4. pfru-r5-07 · comparative clinical trial, crossover meals
   Nutrient extraction was shown to significantly lower GI, compared to eating fruit whole, in subjects with obesity (raspberry/mango: 25.43 ± 18.20 vs. 44.85 ± 20.18, p = 0.034 and passion fruit/mango (26.30 ± 25.72 vs. 42.56 ± 20.64, p = 0.044).
   Nutrients, 2020 · checked 2026-09-24
5. pfru-r5-01 · network meta-analysis of randomized controlled trials
   In the WOMAC pain score, the SUCRA of passion fruit peel extract was 91% (mean difference [MD]: -9.2; 95% confidence interval [CI]: [-16.0, -2.3]), followed by methylsulfonylmethane (89%), undenatured type II collagen (87%), collagen (84%), and Lanconone (82%).
   J Integr Med, 2025 · checked 2026-09-24
6. pfru-r5-03 · randomized controlled trial
   At 60 days, reductions of 18.6%, 18%, 19.6%, and 19.2% in pain, stiffness, physical function, and composite WOMAC score, respectively, were self-reported in the PFP group.
   Nutr Res, 2010 · checked 2026-09-24
7. pfru-r5-04 · randomized controlled trial
   The prevalence of wheeze, cough, as well as shortness of breath was reduced significantly in group treated with PFP extract (P < .05), whereas the placebo caused no significant improvement.
   Nutr Res, 2008 · checked 2026-09-24
8. pfru-r5-08 · controlled clinical trial, non-randomized
   There was a significant difference in the fasting blood glucose values (P = 0.000) and glycated hemoglobin (P = 0.032) after supplementation. It was also seen a reduction in HOMA IR (P = 0.005) in the supplemented group, however it was not observed changes in insulin values for females.
   Nutr J, 2012 · checked 2026-09-24
9. pfru-r5-09 · randomized controlled trial
   The piceatannol group exhibited a statistically significant increase in facial stratum corneum hydration compared to the placebo group.
   Front Nutr, 2026 · checked 2026-09-24
10. ev-pfpg-02 · agency position
    Passionflower is likely safe when it's used as a flavoring in food.
    NIH National Center for Complementary and Integrative Health, Passionflower, page updated April 2025 · checked 2026-10-07
11. hf2-pfru-2 · EFSA scientific opinion
    The CONTAM Panel concludes that the ARfD of 20 μg CN/kg bw should be protective for acute effects of CN from CNGs, regardless of the dietary source.
    EFSA CONTAM Panel, Evaluation of the health risks related to the presence of cyanogenic glycosides in foods other than raw apricot kernels, EFSA Journal 2019;17(4):5662 · checked 2026-10-08
12. hf2-pfru-4 · narrative review
    In vivo acute and subacute toxicity studies indicate that oral administration of the ethanol extract of unripen fruit peel of P. edulis at the dose of 550 mg/kg had no toxic effect on the rats.
    He X et al., Passiflora edulis: An Insight Into Current Researches on Phytochemistry and Pharmacology, Frontiers in Pharmacology 2020;11:617 · checked 2026-10-08
13. hf2-pfru-6 · uncontrolled clinical trial
    People who did not remain until the end abandoned the study due to the strong aftertaste of the flour and abdominal discomfort.
    Nutr J, 2012 · checked 2026-10-08
14. pfru-r5-05 · randomized controlled trial
    No adverse effect was reported by any of study participants.
    Nutr Res, 2008 · checked 2026-09-24
15. ev-pfpg-01 · agency position
    Passionflower should not be used during pregnancy as it may induce uterine contractions and possibly lead to other problems.
    NIH National Center for Complementary and Integrative Health, Passionflower, page updated April 2025 · checked 2026-10-07
16. hf2-pfru-8 · agency position
    Passionflower should not be used during pregnancy as it may induce uterine contractions and possibly lead to other problems. ... Passionflower is likely safe when it's used as a flavoring in food.
    NIH National Center for Complementary and Integrative Health, Passionflower, page updated April 2025 · checked 2026-10-08
17. pfru-r5-10 · case-control study
    The case-control study demonstrated increased risk of illness in those who ate from a special 'tasting menu' and in particular an oyster, passion fruit jelly and lavender dish (odds ratio 7·0, 95% confidence interval 1·1-45·2).
    Epidemiol Infect, 2012 · checked 2026-09-24
18. ev-pfpg-03 · case-control study
    Cross-reacting IgE antibodies recognizing latex and fruit allergens (papaya, avocado, banana, chestnut, passion fruit, fig, melon, mango, kiwi, pineapple, peach, and tomato) were demonstrated by RAST-inhibition tests.
    Allergy, 1997 · checked 2026-10-07
19. ev-pfpg-04 · agency position
    While most people's immune systems can usually fight off the effects of foodborne illness, children, older adults, pregnant women, and people with weakened immune systems (such as transplant patients and individuals with HIV/AIDS, cancer, or diabetes) risk serious illnesses or even death from drinking untreated juices.
    US FDA, What You Need to Know about Juice Safety, content current as of 03/05/2024 · checked 2026-10-07
20. hf2-pfru-3 · EFSA scientific opinion
    In the non‐adult groups (in Australia 2–16 years, in New Zealand 5–14 years), linseed‐containing bread contributed 37% and 24% to chronic exposure in Australia and New Zeeland, respectively. Corresponding values were 26% and 32% for linseed, < 5% and 7% for cooked cassava, 15% and 22% for cassava chips, 6% and < 5% for each apple juice and passion fruit.
    EFSA CONTAM Panel, Evaluation of the health risks related to the presence of cyanogenic glycosides in foods other than raw apricot kernels, EFSA Journal 2019;17(4):5662 · checked 2026-10-08
21. hf2-pfru-7 · uncontrolled clinical trial
    The study excluded patients considered unfit during the interview and/or physical examination, which showed changes in laboratory tests such as: liver and renal dysfunction, cardiac severe changes, alcoholism and those who changed diet, medical treatment or physical activity during the study. ... The drugs dosage was not changed during the study.
    Nutr J, 2012 · checked 2026-10-08
22. pfru-r5-06 · case-control study
    Fruit-specific IgE antibodies were detected in 69.1% of serum samples. Cross-reacting IgE antibodies recognizing latex and fruit allergens (papaya, avocado, banana, chestnut, passion fruit, fig, melon, mango, kiwi, pineapple, peach, and tomato) were demonstrated by RAST-inhibition tests.
    Allergy, 1997 · checked 2026-09-24
23. ev-pfpg-05 · animal study
    Pregnant rats received oral supplementation with a passion fruit extract of defined polyphenolic composition during the last week of gestation and the first week of lactation.
    Biomedicine & Pharmacotherapy, 2026 · checked 2026-10-07
24. hf2-pfru-5 · narrative review
    More importantly, clinical trial on P. edulis efficacy and safety are very scarce to support claims of efficacy.
    He X et al., Passiflora edulis: An Insight Into Current Researches on Phytochemistry and Pharmacology, Frontiers in Pharmacology 2020;11:617 · checked 2026-10-08
25. hf2-pfru-9 · animal study
    Pregnant rats received oral supplementation with a passion fruit extract of defined polyphenolic composition during the last week of gestation and the first week of lactation.
    Biomedicine & Pharmacotherapy, 2026 · checked 2026-10-08
26. pfru-r5-02 · meta-analysis of randomized controlled trials
    Of 20 supplements investigated in 69 eligible studies, 7 (collagen hydrolysate, passion fruit peel extract, Curcuma longa extract, Boswellia serrata extract, curcumin, pycnogenol and L-carnitine) demonstrated large (effect size >0.80) and clinically important effects for pain reduction at short term...No supplements were identified with clinically important effects on pain reduction at long term.
    Br J Sports Med, 2018 · checked 2026-09-24