Bitter Orange

What is Bitter Orange?

Bitter orange is the fruit of Citrus aurantium L., a Rutaceae tree grown commercially in southern Spain, Sicily, the Caribbean and Arizona. In supplements it is sold for one molecule: p-synephrine, a protoalkaloid that made up 0.020 to 0.027 percent of fresh fruit pulp in the analyses compiled by the National Toxicology Program, and 6 to 10 percent of the concentrated immature-peel extracts sold to manufacturers. That is a concentration step of roughly 300-fold before anything reaches a capsule. The peel is simultaneously a food ingredient with GRAS status under 21 CFR 182.20 and 21 CFR 582.20, the base of Seville marmalade, and the flavoring in Grand Marnier, Cointreau and Curacao.

Common Names

• Bitter orange, sour orange, Seville orange, bigarade

• Citrus aurantium L., Citrus aurantium var. amara, Citrus bigaradia

• Zhi shi (immature dried fruit, Fructus Aurantii Immaturus) in traditional Chinese medicine

• Zhi qiao or zhi ke (the larger, later-harvest unripe fruit, Fructus Aurantii)

• Aurantii pericarpium (the pharmacopoeial name for the dried peel)

• Advantra Z, the trade name of the dominant standardized extract

Primary Active Compounds

• p-Synephrine (oxedrine), the marketed active, a phenylethanolamine with a hydroxyl group in the para position

• Octopamine, present in commercial extracts at 0.023 to 0.028 percent

• Tyramine, 0.055 to 0.056 percent in the same extracts

• N-methyltyramine and hordenine, both minor amines with weak adrenergic activity

• Naringin, neohesperidin and hesperidin, the bitter flavanone glycosides that give the fruit its taste

• Bergapten, bergamottin and 6',7'-dihydroxybergamottin, the furanocoumarins responsible for the drug interactions

• Limonene, which makes up roughly 78 to 95 percent of the cold-pressed peel oil and contains no alkaloids at all

Key Note

The compound sold in the bottle and the compound in the marmalade are the same molecule at wildly different doses. A 240 mL glass of Seville orange juice carries about 13.6 mg of p-synephrine, calculated from the 56.9 micrograms per mL measured in juice. A single capsule of a 95 percent standardized extract can deliver more than that in a serving one-tenth the mass, and a Dutch market survey found one product supplying 226 mg per day.

Everything commercially interesting about bitter orange traces back to a rule the FDA published in February 2004 about a different plant entirely.

Primary Functions & Benefits

The claims made for bitter orange supplements fall into three groups, and they have very different amounts of evidence behind them.

Weight loss and appetite suppression. This is the largest commercial claim and the weakest evidence. The 2022 systematic review and meta-analysis by Koncz and colleagues in Nutrients 14(19):4019 pooled 18 double-blind placebo-controlled trials covering 341 adults at doses of 6 to 214 mg per day. Pooled weight change was a mean difference of 0.6 kg across 94 subjects over 42 to 56 days, p = 0.85. That is not a small effect. That is no effect.

Thermogenesis and fat oxidation during exercise. Here the mechanistic data are real and reproducible. Gutierrez-Hellin and Del Coso, British Journal of Clinical Pharmacology 2016, gave 18 active adults 3 mg/kg p-synephrine, roughly 214 mg, and measured maximal fat oxidation during an incremental cycle test. Fat oxidation rose from 0.29 plus or minus 0.15 g/min on placebo to 0.40 plus or minus 0.18 g/min, p = 0.01. Five of the 18 subjects showed no change. Substrate shift during a single exercise bout is not the same finding as fat loss over weeks, and the meta-analysis above is what happened when people looked for the second thing.

Resistance-exercise performance. Ratamess and colleagues, Journal of the International Society of Sports Nutrition 2015, tested 100 mg p-synephrine alone and with 100 mg caffeine in 12 college-aged men and reported about 11 percent more total repetitions than a no-treatment condition. The study was funded by a grant from Nutratech, Inc., the company that sells Advantra Z, and one author disclosed consulting for the same company. That does not make the result wrong. It makes independent replication the thing to look for, and it has not arrived at any scale.

Traditional and culinary functions. Zhi shi has been used in Chinese formulas for centuries as a qi-moving and digestive agent, and the dried peel appears in Western pharmacopoeias as a bitter stomachic and carminative. Neither traditional use is a stimulant use, and neither has been tested against the endpoints the supplement industry sells.

Forms & Standardization

This is where the gap between the fruit and the product becomes visible.

Whole dried peel. Traditional Western herbal dosing is 4 to 6 g of dried peel as a tea or infusion, 2 to 3 g as tincture, 1 to 2 g as extract. At the 0.352 to 0.380 percent synephrine measured in dried fruit, 6 g of peel supplies roughly 21 to 23 mg of p-synephrine, and that is a ceiling rather than a typical figure, since much of the alkaloid does not extract quantitatively into water.

Standard commercial extracts. Dried extracts measured by NTP contained 2.85 to 3.08 percent synephrine. Immature-fruit peel extracts ran 6 to 10 percent. The industry default is a 4 to 6 percent standardization, and Advantra Z, the extract used in most of the sponsored clinical work, is the reference product at 6 percent, 10 percent, 30 percent, 50 percent and 95 percent grades.

High-percentage extracts. Once an extract is standardized to 50 or 95 percent p-synephrine, it is no longer a botanical extract in any meaningful sense. It is a purified amine with plant residue attached. The FDA's own 2020 analysis noted commercial claims as high as 98 percent. There is no plausible extraction path from a fruit containing 0.02 percent synephrine to a 98 percent powder that does not involve either aggressive purification or the addition of synthesized material.

Synthetic p-synephrine. The molecule is manufactured by hydrogenating omega-methylamino-4-hydroxyacetophenone over platinum or palladium. It is identical to the natural compound in structure but not in stereochemistry: fresh fruit pulp measured 92.4 percent l-enantiomer, and dried plant extracts were essentially 100 percent l-form, while synthetic material is racemic. Chiral analysis is the only reliable way to tell a botanically sourced extract from a spiked one, and almost no certificate of analysis reports it.

What the trials actually used. Kaats 2013 in Food and Chemical Toxicology used 98 mg/day of p-synephrine from Advantra Z for 60 days in 75 adults, the longest and highest-dose safety study on record, and it was funded by Nutratech. Bui 2006 used 900 mg of a 6 percent extract, about 54 mg synephrine. Ratamess 2015 used 100 mg. Gutierrez-Hellin 2016 used about 214 mg. Typical retail weight-loss products supply 10 to 40 mg per serving from 100 to 200 mg of extract, which means the consumer dose sits below most of the efficacy trials and the efficacy trials still failed to show weight loss.

Combination formats. Nearly every commercial product is a blend. Bitter orange appears with caffeine anhydrous, green tea extract, yohimbine, theobromine, and in the worst cases with 1,3-dimethylamylamine. Bitter orange alone in a single-ingredient capsule is a specialty item, not the market.

Food Sources

Bitter orange is a real food, and the food dose and the supplement dose are separated by an order of magnitude.

• Seville orange juice measured 56.9 plus or minus 0.52 micrograms per mL, which works out to roughly 13.6 mg of p-synephrine in a 240 mL glass.

• A typical sweet orange (Citrus sinensis) carries about 6 mg of p-synephrine, and a range of common citrus juices supply about 5 mg per 8 ounce glass.

• Mandarin juice (Citrus reticulata) runs 73 to 158 mg/L with a mean of 93 mg/L, meaning an 8 ounce serving can reach roughly 22 mg, higher than a Seville glass and higher than several supplement servings.

• Seville marmalade is made under an FDA recommendation of at least 25 lb of fruit to each 75 lb of sweetener, so a spoonful contains fruit solids in a heavily diluted matrix.

• Frozen concentrated orange juice may legally contain up to 5 percent bitter orange as a flavoring adjunct.

• Grand Marnier, Cointreau, Curacao and orange flower water are made from the peel oil, which is roughly 78 to 95 percent limonene and contains no alkaloids at all. There is no synephrine in the liqueur.

Does food intake matter? For the drug-interaction question, yes, and this is the part people get backwards. A single glass of Seville orange juice raised felodipine AUC by 76 percent and Cmax by 61 percent versus water in Malhotra and colleagues, Clinical Pharmacology and Therapeutics 2001. That effect comes from the furanocoumarins, not the synephrine, and it is a food-level exposure. For the stimulant question, no. Bui 2006 found measurable blood pressure changes from 54 mg of extract synephrine, while orange juice supplying 13 to 14 mg of synephrine produced no change in blood pressure or pulse in the comparison work. The fruit is not the problem. The concentrate is.

Who Should Take Bitter Orange

There is no population with a demonstrated need for supplemental p-synephrine, and the meta-analytic evidence for the primary marketed claim is null. The short list of people for whom it is a defensible experiment is short.

• Healthy adults with no cardiovascular history, no hypertension and no stimulant sensitivity, who understand they are buying a substrate-shift effect during exercise rather than a weight-loss effect, and who are willing to take it without caffeine.

• Endurance and resistance athletes not subject to NCAA rules, who have read that maximal fat oxidation increased about 38 percent in one 18-person crossover and who accept that this has never been shown to change body composition.

• Culinary and traditional users of the whole peel, zhi shi or zhi qiao, at traditional doses under a practitioner, where the exposure is a fraction of a supplement serving.

Anyone taking it for weight loss is buying an effect that 18 pooled trials and 341 subjects failed to find.

Who Should AVOID or Use Caution

Contraindications

• Anyone with hypertension, coronary artery disease, arrhythmia, prior myocardial infarction or stroke. The published case series is cardiovascular almost end to end.

• Anyone taking a monoamine oxidase inhibitor. Bitter orange extracts contain tyramine at 0.055 to 0.056 percent and octopamine at 0.023 to 0.028 percent, both MAO substrates, and p-synephrine itself is a substrate for MAO-A and MAO-B with a Km of 250 micromolar. This is a hypertensive-crisis interaction, not a theoretical one.

• Pregnancy and lactation. Rats given 55 or 110 mg/kg on gestation days 7 to 16 showed reduced uterine implants and viable fetuses, retarded cranial and thoracic ossification, and cases of brain hypoplasia and unilateral microphthalmia. A separate rat study reported no developmental toxicity to 100 mg/kg. With that conflict unresolved and no human data, this is not the ingredient to gamble on.

• Narrow-angle glaucoma, tachyarrhythmia and hyperthyroidism, all standard adrenergic-agonist exclusions.

• NCAA athletes. The NCAA bans synephrine from Citrus aurantium, zhi shi and bitter orange, and also bans octopamine.

Use Caution

• Anyone taking a CYP3A4 substrate. The furanocoumarin issue is separate from the stimulant issue and applies to the juice as well as the extract.

• Anyone combining it with caffeine, which in practice is nearly everyone taking a commercial pre-workout or fat-burner.

• Military service members. Octopamine is on the Department of Defense Prohibited Dietary Supplement Ingredients list, and Operation Supplement Safety classifies synephrine-containing products as high-risk, citing at least two case reports involving service members.

• Anyone using a product also containing yohimbine, DMAA or an unfamiliar amine name. The 2020 FDA survey found undeclared synthetic stimulants in more than 10 percent of bitter orange products tested.

Critical Safety Point

The industry's defense is that p-synephrine alone has never been shown to cause a cardiovascular event, and on the narrow evidence that is arguable. The 2023 review by van der Heyden and colleagues in Cardiovascular Toxicology 23(1):1 to 9 examined 30 published case reports covering 35 patients from eight countries. Of the 20 patients with a circulatory-system diagnosis, 16, or 80 percent, had taken a product combining synephrine with caffeine. The defense is technically defensible and practically empty, because the market does not sell synephrine alone.

Recommended Dosages

There is no established therapeutic dose, because there is no established therapeutic effect. What exists are regulatory ceilings and trial doses, and they do not agree with each other.

Regulatory ceilings

• Germany's BfR recommended that synephrine in sports and weight-loss products not exceed 6.7 mg/day.

• France's ANSES, in its opinion of May 5, 2014, set 20 mg/day as the reference not to be exceeded, matching plausible dietary intake, and recommended against combining it with caffeine or taking it during physical activity.

• Belgium enforces a 20 mg/day limit.

• Health Canada's January 2010 notice capped total synephrine plus octopamine at 30 mg/day and prohibited caffeine in the same product without human safety data. In November 2011 it relaxed this to 50 mg/day of p-synephrine alone, or 40 mg/day combined with a maximum of 320 mg caffeine.

• The United States has no numeric limit at all.

Typical retail dose. 100 to 200 mg of extract supplying 10 to 40 mg of p-synephrine per serving, often two or three servings a day.

Trial doses. 54 mg single dose (Bui 2006), 98 mg/day for 60 days (Kaats 2013), 100 mg (Ratamess 2015), about 214 mg single dose at 3 mg/kg (Gutierrez-Hellin 2016). The Koncz meta-analysis spanned 6 to 214 mg/day.

What is actually on the market. The Dutch RIVM assessment, report 2017-0069, analyzed 56 positive food supplement samples: 39 supplied 1.6 to 65 mg/day, 19 exceeded the Belgian 20 mg/day limit, and one product measured 81,000 mg/kg, corresponding to 226 mg/day. That single product exceeds the German ceiling by a factor of 34.

Duration

The longest controlled exposure on record is 60 days at 98 mg/day. Nothing longer has been done. The Koncz analysis found that 8 weeks of use raised systolic pressure by 6.37 mmHg (p = 0.02) and diastolic by 4.33 mmHg (p = 0.03), which is the opposite of the pattern you want from a compound with no chronic safety data. If a person is going to take it, the case for cycling short and stopping is stronger than for most stimulants, not weaker.

Timing & Administration

p-Synephrine reaches peak serum concentration between one and two hours after an oral dose, with a biological half-life of about two hours. Roughly 66 percent is recovered in urine as a conjugated metabolite and only about 2.5 percent unchanged, so first-pass handling is extensive and interindividual variation is wide.

• Exercise-performance studies dose 60 minutes before activity. Gutierrez-Hellin used exactly this interval.

• Taking it late in the day is the wrong call even though the CNS penetration is low compared with ephedrine, because the peripheral cardiovascular effect in Bui 2006 was still measurable at hour 5.

• Food slows absorption but there is no controlled fed-versus-fasted pharmacokinetic study to quantify it.

• Do not take it with a CYP3A4 substrate drug in the same window, and do not wash it down with Seville orange juice or grapefruit juice.

• The combination with caffeine is the one every product makes and the one every regulator that has looked at it has flagged.

Timeline of Effects

Within 1 to 2 hours. Peak plasma concentration. In Bui 2006, 15 healthy young adults given a single 900 mg dose of 6 percent extract showed systolic pressure significantly above placebo from hour 1 through hour 5, with a peak difference of 7.3 plus or minus 4.6 mmHg. Heart rate rose from hour 2 through hour 5, peak difference 4.2 plus or minus 4.5 beats per minute.

Within 1 hour, during exercise. The substrate shift. Maximal fat oxidation moved from 0.29 to 0.40 g/min in the 2016 crossover, and carbohydrate oxidation fell correspondingly at low to moderate intensity.

6 to 12 weeks. This is the window the weight-loss trials used, and it is the window in which nothing happened. The pooled 42 to 56 day weight difference was 0.6 kg with p = 0.85.

8 weeks. The point at which the meta-analysis detected a blood pressure signal rather than a weight signal.

60 days. The end of the longest controlled exposure ever run.

What counts as a fair trial. If a person is testing it for exercise substrate use, a single session answers the question. If a person is testing it for weight loss, the honest answer is that 18 trials with 341 subjects already ran that test and found nothing, and an individual n-of-1 over 8 weeks is not going to detect what a meta-analysis could not. Stop at 8 weeks.

Benefits of Taking Bitter Orange

Documented in controlled human trials

• Increased fat oxidation during submaximal exercise at 3 mg/kg, reproduced in trained cyclists and in active women in follow-up work by the same group.

• Increased repetition volume during resistance exercise at 100 mg, in a single industry-funded 12-person study.

• Increased resting metabolic rate, reported across several of the studies collected in the 2012 review by Stohs, Preuss and Shara in International Journal of Medical Sciences 9(7):527 to 538, which covered 23 studies and about 360 subjects at 10 to 80 mg/day.

Not documented

• Weight loss. Mean difference 0.6 kg, p = 0.85.

• Fat mass, fat-free mass or body-fat percentage change. All non-significant in the pooled analysis.

• Appetite suppression as an isolated effect. The trials that reported it used multi-ingredient products.

Traditional and food uses that are genuinely useful

• The peel is a bittering and aromatic agent with a long culinary record, and the essential oil is a fragrance and flavor material regulated by IFRA at a maximum 1.25 percent in leave-on products for sun-exposed skin because of phototoxicity.

• Zhi shi is a standard component of Chinese digestive and qi-regulating formulas with centuries of use, evaluated on endpoints that have nothing to do with thermogenesis.

The compound does measurably shift fuel use. It does not measurably change what people weigh.

Potential Negatives & Side Effects

Reported in controlled trials

• Systolic blood pressure increase, peak 7.3 mmHg at hours 1 to 5 after a single 54 mg dose.

• Diastolic increase, peak 2.6 mmHg, hours 4 to 5.

• Heart rate increase, peak 4.2 bpm, hours 2 to 5.

• After 8 weeks, systolic plus 6.37 mmHg and diastolic plus 4.33 mmHg in the pooled analysis.

• Anxiety, sensitivity to light and a sensation the NTP summary recorded as "warmed blood," reported with Xenadrine RFA-1.

Reported in the case literature

The van der Heyden 2023 review found 30 case reports covering 35 patients. Among them: chest pain in 11, syncope in 6, dizziness in 6, palpitations in 4. Diagnoses included ischemic heart disease in 10, cardiac arrhythmia in 4, cerebrovascular disease in 2. Five patients were left disabled or remained on medication at last follow-up. Where a dose was reported, in 11 patients, exposure ranged from 12 to 100 mg.

Individual reports worth naming, because they are checkable:

• Nasir and colleagues, Mayo Clinic Proceedings 2004: exercise-induced syncope with QT prolongation on ephedra-free Xenadrine.

• Nykamp and colleagues, Annals of Pharmacotherapy 2004: acute lateral-wall myocardial infarction in a 55-year-old woman taking a product declaring 300 mg bitter orange.

• Bouchard and colleagues, Mayo Clinic Proceedings 2005: ischemic stroke in a man after roughly a week of daily use.

• Gange and colleagues, Mayo Clinic Proceedings 2006: variant angina.

• Sultan and colleagues, 2006: ischemic colitis in a 26-year-old woman.

• Holmes and Tavee, Military Medicine 2008: cerebral vasospasm and stroke on ephedra-free Xenadrine.

• Smedema and Muller, 2008: coronary spasm and thrombosis in a bodybuilder using a product containing synephrine, octopamine, tyramine and caffeine together.

• Stephensen and Sarlay, Military Medicine 174(12):1313, 2009: ventricular fibrillation.

• Thomas and colleagues, Texas Heart Institute Journal 2009: ST-elevation myocardial infarction in a 24-year-old man.

The confound, stated plainly

Eighty percent of the circulatory cases involved caffeine in the same product. Case reports cannot apportion blame between two stimulants taken simultaneously, and the industry is right that they do not prove p-synephrine caused these events. What the industry does not say is that no commercially meaningful product isolates p-synephrine, so the exposure the case series describes is the exposure consumers actually have.

Deficiency Symptoms

Bitter orange is not an essential nutrient and p-synephrine is not a required dietary component, so no deficiency state exists and none has ever been described. Trace p-synephrine does circulate endogenously in humans, at a mean of about 58 pg/mL in normotensive subjects, but that is a byproduct of catecholamine-adjacent metabolism rather than evidence of a dietary requirement. Three healthy men placed on a plant-free diet for three days excreted no detectable synephrine in urine and showed no consequences of any kind.

What bitter orange addresses

Nothing that a deficiency causes. It is a stimulant offered against body weight and exercise performance, and the pooled trial evidence supports the second in a narrow mechanistic sense and not the first at all. There is no physiological deficit it corrects.

Bottom line

If a company implies your body is short of synephrine, that is a marketing claim with no biology under it. The correct question is whether adding an adrenergic amine to a healthy system produces a benefit worth its blood pressure cost, and on current evidence the answer for weight loss is no.

Toxicity Symptoms

At high intake

Acute human toxicology is thin because deliberate overdose is rare, but the signals that exist are consistent. Children given an overdose of synthetic synephrine with N-methyltyramine developed rapid blood pressure elevation, nausea, vomiting, irritability and tachycardia, with symptoms described as short-lived. Rats given standardized Citrus aurantium extracts at 2.5 to 20 mg/kg developed ventricular arrhythmias with widening of the QRS complex, and mortality reached 30 percent for a 4 percent extract and 50 percent for a 6 percent extract at the high dose. Murine LD50 values are 270 mg/kg intravenous and 1,000 mg/kg intraperitoneal for synephrine, with subcutaneous LDLo of 1,500 mg/kg in both mouse and rat.

Signs to reduce or stop

• Chest pain or pressure of any kind, at any dose. This was the single most common presenting symptom across the 35 case-report patients.

• Palpitations, an irregular pulse or a resting heart rate persistently above baseline.

• Syncope or near-syncope, particularly during or just after exercise, which is the pattern in the QT-prolongation and ventricular-fibrillation reports.

• A measured rise in blood pressure. A 6 mmHg systolic increase is easy to miss without a cuff and is exactly what the 8-week pooled data predict.

• Severe headache, visual disturbance or unilateral weakness. Retinal artery occlusion and stroke are both in the published case series.

General note

The realistic hazard here is not a toxic overdose from a single bottle. It is a sustained adrenergic load, stacked with caffeine, taken by someone doing high-intensity exercise, in a product whose actual synephrine content the FDA could not match to the label in 18 of 23 cases. The exposure is uncontrolled at the point of purchase, which makes the dose-response reasoning that would normally protect a consumer unavailable to them.

How Bitter Orange Works

p-Synephrine is a phenylethanolamine. It shares the aromatic ring, the beta-hydroxyl and the N-methyl amine with norepinephrine, ephedrine and phenylephrine, and the differences between them are where the entire safety argument lives.

• Against phenylephrine. Phenylephrine is m-synephrine, the same molecule with the ring hydroxyl in the meta position instead of the para position. This is not a cosmetic difference. The meta hydroxyl hydrogen-bonds to a serine residue in the adrenergic binding pocket; the para hydroxyl cannot reach it. Measured against alpha-1 adrenoceptors, m-synephrine was 6-fold less potent than norepinephrine while p-synephrine was about 1,000-fold less potent. At alpha-2 the gap was 150-fold versus 1,000-fold. At beta-1, p-synephrine came in roughly 40,000-fold below norepinephrine. The industry's pharmacological defense is real and it is this data.

• Against ephedrine. The comparison the marketing implies is looser than it sounds. Ephedrine has no ring hydroxyl at all and carries an alpha-methyl group on the side chain, which is what gives it monoamine-releasing activity and central penetration that p-synephrine largely lacks. Bitter orange is not "ephedra with the hydroxyl moved." It is a weaker, more peripheral, less lipophilic compound in the same chemical family.

• Beta-3 adrenoceptors and lipolysis. The thermogenic story rests on beta-3 activation in adipocytes. Octopamine is a selective beta-3 agonist in rodent fat cells and produced a 19 percent reduction in body-weight gain in obese rats over four weeks. In human fat cells it was ineffective. p-Synephrine is only a partial beta-3 agonist. This is the gap between the rodent mechanism the marketing describes and the human outcome the trials measure.

• Fuel-partitioning effect. What p-synephrine does reproducibly in humans is shift substrate use during submaximal exercise toward fat and away from carbohydrate, without a measurable resting change in heart rate or blood pressure at 3 mg/kg. That is a real pharmacodynamic effect. It has not translated into weight change in any adequately powered trial.

• Metabolism. p-Synephrine is a substrate for both MAO-A and MAO-B with a Km of 250 micromolar, which is the mechanistic basis of the MAOI contraindication. CYP2D6 participates in its onward metabolism. Peak serum arrives at one to two hours, half-life is about two hours, roughly 66 percent appears in urine as a conjugate and about 2.5 percent unchanged.

• The furanocoumarin mechanism, which is a separate system entirely. Bergapten, bergamottin and 6',7'-dihydroxybergamottin in the fruit are mechanism-based inhibitors of intestinal CYP3A4. This has nothing to do with synephrine, adrenergic receptors, or weight loss. It is why the juice interacts with drugs.

Synergistic Supplements

The honest version of this section is short, because the only well-documented synergy is the one regulators keep trying to break up.

• Caffeine. The pairing is nearly universal in commercial products and it is the one with the clearest additive cardiovascular signal. Ratamess 2015 found the 100 mg synephrine plus 100 mg caffeine arm improved power and velocity where synephrine alone did not. Health Canada permits it only up to 40 mg synephrine with a maximum 320 mg caffeine; ANSES recommends against it entirely. Calling this a synergy for benefit and not for risk is not supported.

• Naringin and hesperidin. Kaats 2013 tested bitter orange extract with and without added naringin (600 mg) and hesperidin (100 mg) over 60 days. The flavonoid arm showed a small heart rate increase the extract-alone arm did not. These are the fruit's own flavanones, and their principal effect in this context is on absorption and enzyme inhibition rather than on thermogenesis.

• Green tea catechins and yohimbine. Commonly stacked, no controlled evidence of additive benefit for body composition, and yohimbine adds its own alpha-2 antagonism and its own case literature.

• What does not belong in the same product. Any MAO inhibitor, any additional sympathomimetic, and any product also carrying an unfamiliar amine.

Interactions & What NOT to Take

The interaction profile splits cleanly into two mechanisms that get conflated constantly.

Mechanism one: the furanocoumarins, which affect drug metabolism

• Malhotra and colleagues, Clinical Pharmacology and Therapeutics 2001: a single serving of Seville orange juice raised felodipine AUC by 76 percent and Cmax by 61 percent compared with water. The effect was attributed to furanocoumarins, and dilute grapefruit juice was the comparator.

• Di Marco and colleagues, 2002: Seville orange juice raised dextromethorphan bioavailability from 0.10 to 0.46, a marker of intestinal CYP3A and P-glycoprotein inhibition.

• Sildenafil AUC and Cmax rose 44 percent with Seville orange juice in a healthy-subject study.

• Edwards and colleagues, Clinical Pharmacology and Therapeutics 1999, is the finding that separates Seville orange from grapefruit: Seville orange juice significantly reduced enterocyte CYP3A4 protein yet did not alter cyclosporine bioavailability, while grapefruit juice did. Grapefruit is doing something to transporters that Seville orange is not doing to the same degree. The two juices are not interchangeable, and treating a bitter orange warning as a copy of the grapefruit warning gets the direction of risk wrong in both directions.

• Bitter orange juice extract also inhibits intestinal P-glycoprotein-related efflux carriers in vitro, and the fruit's naringin is an established clinical inhibitor of the uptake transporter OATP1A2.

Mechanism two: the amines, which affect the cardiovascular system

• Monoamine oxidase inhibitors, including phenelzine, tranylcypromine, isocarboxazid and the antibiotic linezolid. Tyramine plus MAOI is a hypertensive-crisis pairing and bitter orange extracts contain measurable tyramine.

• Any other sympathomimetic: pseudoephedrine, phenylephrine, ephedrine, DMAA, amphetamine-class stimulants.

• Beta-blockers and antihypertensives, where the pharmacological opposition is direct.

• Caffeine, in doses above roughly 320 mg alongside 40 mg synephrine, which is the only combination any regulator has explicitly permitted.

A note on which risk applies to whom. The CYP3A4 issue applies to anyone eating the fruit or drinking the juice, including people who never touch a supplement. The cardiovascular issue applies to people taking concentrated extracts. Marmalade is a drug-interaction question. A fat-burner is a blood-pressure question.

Quality, Testing & Adulteration

This is the section where bitter orange is worse than its reputation, and the primary source is the FDA's own laboratory.

The FDA's 2020 survey. Pawar and Sagi, Drug Testing and Analysis 2020, analyzed 59 bitter orange dietary supplements purchased between June 2016 and November 2018 using LC-MS/MS.

• Only 23 of the 59 products declared a synephrine amount on the label at all.

• Of those 23, only 5 contained an amount close to the declared value. That is a label-accuracy rate of 22 percent among the products that bothered to make a claim.

• Six products contained methylsynephrine, a synthetic amine also known as oxilofrine, at up to 240 mg per daily serving.

• One product contained isopropyloctopamine at up to 76 mg per daily serving.

• Elevated levels of octopamine, tyramine and hordenine were also found, above what the botanical alone would explain.

FDA had already ruled on methylsynephrine. On March 31, 2016 the agency issued 7 warning letters covering 8 products, stating that methylsynephrine does not meet the statutory definition of a dietary ingredient under the Federal Food, Drug, and Cosmetic Act. Four years later its own scientists found it in 10 percent of the bitter orange products they bought, and no enforcement action or consumer advisory followed the 2020 paper.

Earlier FDA analytical work on 48 supplements found the same pattern in the other direction: one product declared 19.5 percent synephrine and contained 13.4 percent; the highest measured concentration, 16.9 percent, came from a product declaring 16 percent; and two products declaring 20 percent and 25 percent synephrine contained none that could be detected.

What a certificate of analysis should show, and almost never does

• Species confirmation as Citrus aurantium by DNA or HPTLC, plus the plant part: immature whole fruit, peel, or a mixture.

• Quantitative p-synephrine by HPLC or LC-MS/MS with the actual mg per serving rather than a percentage of an unstated extract weight.

• Chiral analysis showing the l-enantiomer fraction. Natural material runs 92 to 100 percent l-form. Racemic material means synthetic synephrine was added.

• A screen for methylsynephrine, isopropyloctopamine, DMAA, DMHA and higenamine, none of which should be present at any level.

• Quantified octopamine, tyramine, N-methyltyramine and hordenine, so the total adrenergic amine load is visible rather than hidden behind a single marker.

• Furanocoumarin content, which no consumer-facing COA reports and which drives the drug interaction.

What a buyer can actually verify. Very little without a lab. Bitter orange has no USP monograph for the finished supplement, and third-party certification programs that do sport-focused banned-substance testing will generally reject synephrine products outright rather than certify them, because the NCAA bans the compound. The practical filter is negative: a product carrying a proprietary blend, no mg figure for p-synephrine, and no chiral or contaminant data has told you it did not test, and the FDA's 22 percent label-accuracy figure is the base rate you should assume applies.

Special Considerations

Athletes. Synephrine sits on the 2026 WADA Monitoring Program, meaning it is measured but not prohibited. Octopamine is prohibited under S6 as a specified stimulant, as is oxilofrine, which is methylsynephrine. The NCAA bans both synephrine and octopamine. A bitter orange extract naturally contains octopamine at 0.023 to 0.028 percent, which means a WADA-tested athlete taking a synephrine product is consuming a prohibited substance as an unavoidable minor constituent of a permitted one. That is a positive test waiting for a sensitive enough assay.

Service members. Octopamine is on the DoD prohibited list. Operation Supplement Safety flags synephrine products as high-risk and cites at least two case reports involving service members. Two of the published cardiovascular cases were published in Military Medicine.

People with hypertension. The 8-week pooled effect, plus 6.37 mmHg systolic, is roughly the size of the reduction a single first-line antihypertensive is expected to produce. Taking this while treating hypertension is working against your own prescription.

People who eat marmalade or drink Seville orange juice while on medication. This is a different population with a different risk. If you take felodipine, nifedipine, simvastatin, cyclosporine, tacrolimus, sildenafil, or any other narrow-margin CYP3A4 substrate, the juice matters and the supplement question is beside the point.

Photosensitivity. Bitter orange peel oil is phototoxic. All subjects in a controlled test exhibited phototoxic reactions to 5 microliters per square centimeter of 100 percent oil under light exposure. IFRA caps it at 1.25 percent for products applied to sun-exposed skin. This applies to the essential oil in cosmetics, not to oral supplements.

Regulatory contradiction worth noticing. Citrus aurantium appears in the FDA's Poisonous Plant Database, while its peel, flowers, leaf, oils and extracts are simultaneously listed as GRAS food additives across 21 CFR parts 172, 182, 184 and 186. Both entries are defensible on their own terms. Together they describe an agency that has never resolved what this plant is.

Research Status & Evidence Quality

Strong Evidence For

• p-Synephrine increases fat oxidation rate during submaximal exercise at 3 mg/kg. Demonstrated in multiple crossover trials by the same research group in different populations.

• Seville orange juice inhibits intestinal CYP3A4 and raises exposure to CYP3A4 substrate drugs. Felodipine AUC plus 76 percent, dextromethorphan bioavailability 0.10 to 0.46, sildenafil AUC plus 44 percent.

• Commercial bitter orange supplements do not reliably contain what their labels claim. 5 of 23 label-declaring products matched, in an FDA laboratory analysis of 59 products.

Moderate Evidence For

• Single doses of roughly 50 mg raise systolic blood pressure by about 7 mmHg for up to 5 hours. One well-designed 15-subject crossover, consistent with the pooled 8-week finding.

• Eight weeks of use raises systolic and diastolic blood pressure. Pooled from the 2022 meta-analysis, p = 0.02 and p = 0.03.

• Products combining synephrine with caffeine are associated with serious cardiovascular events. Thirty case reports, 35 patients, 80 percent of circulatory cases involving the combination.

Emerging / Preliminary Evidence For

• Improved resistance-exercise volume at 100 mg. A single 12-subject industry-funded crossover.

• Increased resting metabolic rate. Reported across several small studies collected in an industry-affiliated review.

• Any effect on appetite. Reported only from multi-ingredient products where synephrine cannot be isolated.

Research Limitations

• Only 6 of the 18 trials in the 2022 meta-analysis used isolated p-synephrine. The rest tested blends, which is why "synephrine causes X" is a claim the literature is structurally unable to test at scale.

• The largest and longest safety study, 75 subjects for 60 days at 98 mg/day, was funded by the manufacturer of the extract used, and the principal author of the most-cited favorable reviews has disclosed consulting for that same company. This does not invalidate the work. It means the safety literature and the commercial interest are not independent of each other.

• Total FDA adverse event reports for bitter orange stood at 169 from 1969 through October 2009, with 22 filed with CFSAN between April 2004 and October 2009, according to an industry-affiliated analysis. No comparably detailed public tally has been published for the 16 years since, which is itself a gap.

• No study has assessed cardiotoxicity with dedicated cardiac endpoints. No chronic exposure, carcinogenicity or reproductive-toxicity study in humans exists at all.

• Case reports cannot establish causation, particularly when 80 percent of them involve two stimulants at once. This limitation cuts against alarm and against reassurance equally.

Summary & Key Takeaways

Bitter orange entered the American supplement market as a regulatory answer rather than a scientific one. When FDA declared ephedrine-alkaloid supplements adulterated effective April 12, 2004, the category needed a stimulant that was not ephedra, and Citrus aurantium was standing there with a GRAS food history and a pharmacology paper trail thin enough to be marketed as safety. Twenty-two years later the weight-loss claim has been tested and failed: 18 double-blind trials, 341 subjects, a pooled difference of 0.6 kg with a p value of 0.85. What survives is a genuine fuel-partitioning effect during exercise, a measurable blood pressure increase, a drug-interaction problem inherited from the fruit rather than the alkaloid, and a supply chain in which the FDA's own laboratory could match label to content in 5 of 23 attempts.

Bottom Line

The pharmacology defense for p-synephrine is legitimate on its own terms. The para hydroxyl really does cost it 1,000-fold at alpha receptors and 40,000-fold at beta-1 relative to norepinephrine, and it really is a weaker, more peripheral compound than ephedrine. That defense describes a molecule almost nobody buys. What people buy is 10 to 40 mg of p-synephrine stacked with caffeine, in a bottle whose actual content the label does not predict, sometimes with a synthetic amine the FDA has ruled is not a dietary ingredient at all. Judged on the product rather than the molecule, the case for taking it is that it may shift your fuel mix during a workout, and the case against is a blood pressure increase, a case series with 5 people left disabled, and a null result on the reason you were buying it.

Key Safety Points

• Do not combine with an MAO inhibitor. Tyramine plus MAOI is a hypertensive-crisis pairing and this extract contains tyramine.

• Do not take it if you have hypertension, coronary disease, an arrhythmia, or a prior stroke or infarction.

• Do not take it in pregnancy or while breastfeeding. The rat developmental data are conflicting and there are no human data.

• NCAA athletes should avoid it outright; the compound is banned and its natural octopamine content is banned by WADA as well.

• Assume the label is wrong about the dose. Nothing about the FDA's 2020 survey suggests otherwise.

• Keep the juice question separate from the pill question. Seville orange juice interacts with CYP3A4 drugs at ordinary food servings, and that risk applies to people who never open a supplement bottle.

Special Note

This pattern repeats on any shelf where a category loses its lead ingredient to a regulator. What replaces a banned compound is chosen for its legal position first and its evidence second, and the phrase that markets it, "ephedra-free," describes what the product does not contain rather than what it does. The absence of a ban is not a safety finding. It usually just means nobody has looked yet, and by the time the case reports accumulate, the label has already moved on to the next word.