Bergamot Polyphenols (BPF)

Supplements · §488

Your last lipid panel came back with a number flagged, and your GP gave you the talk — diet, exercise, recheck in six months, and if it doesn't move we're going to talk about a statin. You already eat reasonably well. You already move. The lever you didn't know existed is a flavonoid extract from a Calabrian fruit too bitter to drink: two capsules a day for twelve weeks, the LDL comes down fifteen to twenty percent, the triglycerides move with it, and the conversation you were dreading does not happen this time.

Do · Daily Evidence Emerging Chapter Supplements

The mechanism is real — two of the flavonoids in the extract carry the same molecular handle that statins use to lock onto the cholesterol-making enzyme, so this is not a vague antioxidant nudge. The drop you can expect is meaningful but smaller than the headlines suggest: bigger than oat bran, smaller than a real statin, comparable to berberine, with an honest second hit on triglycerides and fasting blood sugar. Twice a day, with meals, for at least three months before you judge it. Standardisation on the label matters more than the brand on the front.

The reason this works at all — and the reason it isn't just another antioxidant supplement claiming to do something — is that two of the flavonoids in the bergamot juice fraction, brutieridin and melitidin, carry a molecular tail identical in shape to the working end of a statin. They sit in the same pocket of the cholesterol-making enzyme in your liver and slow it down the same way, just less powerfully.

The other half of the action runs inside the liver itself. The extract switches on a cellular clean-up program that empties stored fat droplets from liver cells — the same fat that accumulates in non-alcoholic fatty liver disease. In rats fed a junk-food diet, the liver cleared visible fat without any weight loss, just from this clean-up signal (Parafati 2015). A third effect, downstream of the first two, is a quieter set of changes in the artery wall: less oxidised LDL stuck to the lining, less of the molecular hook that pulls it in (Gliozzi 2013).

How much it actually moves

The early Italian trials produced eye-popping numbers — a third off LDL, nearly forty percent off triglycerides, a fifth off fasting blood sugar (Mollace 2011). The harder, more recent literature has cooled the headline. A six-month trial in eighty people with moderately elevated cholesterol landed around 24% off LDL and — more interestingly — the early plaque marker in the carotid artery wall actually regressed, the first time a bergamot extract has been shown to move something inside the artery, not just on a blood test (Toth 2016). An independent Chinese trial in older adults landed at 15–20% off LDL over twelve weeks (Cai 2017).

The pattern across the trials is consistent in direction. The live question is how much to discount the biggest numbers, and the honest answer is: somewhere between a third and a half. Plan on fifteen to twenty-five percent off LDL at twelve weeks, not a third. The same intervention pulls triglycerides down by something similar, lifts HDL a little, and drops fasting glucose where it was elevated — the unusual feature is that one supplement nudges all four of those numbers in the right direction at the same time.

One more replicated result is worth knowing: combining the extract with a low-dose statin (rosuvastatin 10 mg) reproduces the LDL effect of the higher dose of the same statin (20 mg), with extra reductions in oxidised LDL and the molecular hook that drags it into the artery wall (Gliozzi 2013). This is the use case for people stuck at the lowest statin dose their muscles tolerate.

The long fuse

Cholesterol is the slow story. The way LDL puts people in cardiology offices is not a sudden event — it is the year-by-year accumulation of plaque in arteries that began in your twenties and continues quietly every year you let it. The body keeps no copy: the LDL you carried at 35 is in your arterial wall at 55. The European Atherosclerosis Society's consensus statement is unambiguous — the size of your cardiovascular risk reduction is proportional to the cumulative LDL exposure you lower, and the earlier you start, the more wall is left unbuilt (Ference 2017).

The version of this story that lands in your life is mostly invisible until it isn't. The friend whose father had his first heart attack at fifty-eight and his second at sixty-two. The recheck at the GP that escalates to a statin conversation, then to a cardiology referral, then to the conversation about a stent — each step a separate appointment, each appointment compressing a decision the previous one didn't require. The version of you that pulled on this lever ten years earlier has no single dramatic moment to point to. You just look up at sixty and notice your peers are starting to have those conversations and you are not.

How to take it

The dose the trials converge on is around a gram a day of a standardised bergamot polyphenol extract — typically two 500 mg capsules, one before lunch and one before dinner. Give it twelve weeks before you decide whether it works for you; pull a fasting lipid panel before you start and another at the twelve-week mark. There is no cycling-on-and-off ritual here — the benefit accumulates while you take it and would presumably retreat if you stop.

When to skip it, or check first

Three real reasons to check with someone before starting.

If you're on insulin, a sulfonylurea, or actively titrating metformin, the extract's blood-sugar effect is additive — your medication may need to come down, and you should be monitoring glucose for the first few weeks rather than discovering the interaction by feel.

If you take a drug your doctor told you to avoid grapefruit with — that includes some statins (simvastatin, lovastatin, atorvastatin), the blood thinners apixaban and rivaroxaban, several calcium channel blockers, and a handful of others — the same caution applies here. The bergamot peel and essential oil contain a compound in the same family as grapefruit that blocks the gut enzyme handling those drugs. The standardised juice extract should be largely stripped of it, but commercial products vary and nobody is auditing the supply chain. Treat any of those combinations as a grapefruit-style interaction until your product is certified low in that compound (the technical name is bergapten).

If you're pregnant or breastfeeding, no trial has tested this; default to avoiding.

What gets confused

Two things to clear up. First, the bottle of bergamot oil in the wellness shop — perfume, aromatherapy, the scent in Earl Grey tea — is not this product. The oil is pressed from the peel; it carries the phototoxic compound that worries pharmacists, and it has no evidence base for cholesterol. The clinical work is on a standardised juice flavonoid extract, sold in capsules and labelled with a polyphenol percentage. If your bottle smells of bergamot, it is the wrong product.

Second, this is not a statin replacement. If your cardiologist says you need a statin — a strong family history of early heart attacks, a very high LDL, an elevated lipoprotein(a) (Lp(a)), established coronary disease — the absolute risk reduction from a real statin is too large for any current bergamot extract to substitute. The honest use case is the borderline reader who doesn't yet warrant a prescription, the metabolic-syndrome reader chipping at four numbers at once, and the patient already on a low-dose statin whose muscles will not tolerate a higher one.

Where it sits in the lever stack

For mild-to-moderate cholesterol with no statin indication yet, the levers in rough order of effect size:

Mediterranean dietary pattern (saturated fat traded for unsaturated): LDL down roughly 5–10%.
Plant sterols / stanols at 2 g/day (added to spreads or as a supplement): another ~10%.
Psyllium (soluble fibre) at 10 g/day: ~7%, plus a real glucose effect.
Berberine at 500 mg three times a day: ~20%, plus a glucose effect.
Bergamot polyphenol extract at 1 g/day: ~15–25% on LDL, plus triglycerides, HDL, fasting glucose, and liver fat in one product.
Low-dose rosuvastatin (5 mg/day): ~40% on LDL, the biggest single mover.

Bergamot sits between berberine and a low-dose statin. The two distinctive features against everything else on the list: a documented synergy when stacked with a low-dose statin (Gliozzi 2013), and a single product that nudges LDL, triglycerides, fasting glucose, and liver fat at once — for the metabolic-syndrome reader who would otherwise be reaching for three different supplements.

Why "I tried it and nothing happened"

Three reliable reasons.

Wrong product. Capsules of bergamot essential oil. A fruit-juice drink. An unstandardised "bergamot extract" with no polyphenol percentage on the label. None of these are what the trials tested. If the label does not give you a polyphenol percentage (the trials used around 38–47%), it is a different substance and you cannot reason from the clinical literature to that bottle.

Underdosing. Sub-500 mg/day does not register in the dose-response data. The working dose starts around a gram a day; the 1500 mg arm gives diminishing returns.

Window too short. A two- or four-week trial will not move the lipid panel. Eight weeks is the floor, twelve weeks is the honest reading.

A fourth, subtler failure mode: reading the 36% LDL headline and being disappointed by a 17% result. Seventeen percent is what most readers will get, and seventeen percent sustained over twenty years is a different cardiovascular trajectory.

Buying it, paying for it, knowing it worked

What to look for on the label: a stated polyphenol percentage (ideally 38–47% total polyphenols), and where the manufacturer is being honest, the brutieridin + melitidin content. A bottle that says only "bergamot 500 mg" with nothing else is a different product from what the trials tested, even if the dose number matches.

Cost runs roughly $25–35 a month at retail — call it a coffee a week. Most pharmacies do not stock it; you will be ordering from a supplement-channel retailer or a functional-medicine clinic. The cheapest piece of self-monitoring is the lipid panel your GP already runs once a year: pull one before you start, repeat it at twelve weeks, and the comparison answers the only question that matters — does this work for you, specifically. If your panel didn't move, neither did your arteries, and the budget is better spent elsewhere.

What changes, and when

On a weekly clock, the first numbers to move are triglycerides and HDL, both visible on a panel run at four weeks. The LDL drop arrives at eight to twelve weeks and tends to keep settling for the first few months (Mollace 2011, Cai 2017). If your fasting glucose was elevated, it moves in parallel. If you have fatty liver, the liver enzymes drift back toward range over twelve to twenty-four weeks (Parafati 2015 for the mechanism). None of this is something you can feel directly — the lift in afternoon energy that some readers in the metabolic-syndrome group report (Capomolla 2019) is the whole cluster being chipped at, not a stimulant kicking in.

The slower payoff was first put on the map by the six-month trial in eighty people: the carotid wall thickness regressed — not held, went backwards — for the first time a bergamot extract has been shown to do that (Toth 2016). The cosmetic flow-through is real but slow on the same clock: the version of you whose metabolic-syndrome cluster never quite consolidated, whose vascular aging bent earlier rather than later, is a different face in photographs at sixty, not at twelve weeks.

The very long payoff is the one you don't experience as an event. The recheck at the GP that ends in "looks great, see you in a year" instead of escalating to the conversation about a statin. The cardiology appointment that doesn't happen at sixty because your cumulative LDL exposure has been twenty percent lower for the previous twenty years (Ference 2017). The version of you whose family-history risk got pushed out a decade because you found a cheap, real lever before you needed an expensive one.

Near neighbours

A few adjacent things worth looking up separately. The lipid panel itself — what the numbers mean, which ones actually matter for cardiovascular risk, and whether you should be tracking apolipoprotein B alongside LDL — has its own entry. The wider triage of cholesterol between diet, supplements, and a statin is a decision tree, not a supplement entry. And the unrelated cousin: bergamot essential oil for aromatherapy or skin is a different substance with a partly opposite risk profile (the peel compound is phototoxic and pro-interaction, not lipid-lowering).

Substance and claimed effects

Bergamot (Citrus bergamia Risso) is a sour citrus grown almost exclusively on a narrow strip of Calabrian coastline in southern Italy. Industrially the fruit is pressed for the essential oil used in perfumery and Earl Grey tea; the juice, too bitter to drink, was historically a waste stream. The juice fraction is unusually concentrated in flavanone glycosides — neoeriocitrin, naringin, neohesperidin, hesperetin, narirutin — and, distinctively, two HMG-acylated flavonoids, brutieridin and melitidin, identified by Di Donna et al. (2009) as endogenous HMG-CoA reductase inhibitors structurally analogous to the statin pharmacophore. The juice flavonoid fraction is now extracted as a standardized supplement marketed as Bergamot Polyphenolic Fraction (BPF), typically 38–47% total polyphenols.

The claims clustered around BPF, taken at 500–1500 mg/day for 30 days to 6 months, are: lowered LDL and total cholesterol, lowered triglycerides, raised HDL, lowered fasting glucose and HbA_1c, reduced liver fat in non-alcoholic fatty liver disease (NAFLD), reduced atherogenic small-dense LDL, modest weight and waist-circumference loss in metabolic syndrome, and a synergistic effect when combined with a statin. The entry covers each of these consequences holistically; the headline framing is cholesterol because that is where the largest replicated effect and the largest reader audience sit.

Evidence by addressing question

mechanism

Three mechanisms are load-bearing. First, direct HMG-CoA reductase inhibition: brutieridin and melitidin carry a 3-hydroxy-3-methylglutaryl moiety identical in shape to the active head of the statin class, and Di Donna et al. (2009) showed in vitro inhibition at micromolar concentrations. This is the closest thing in the supplement world to a real statin mechanism rather than a downstream nudge.

Second, AMPK activation and lipophagy. Parafati et al. (2015) showed in a cafeteria-diet rat model that BPF clears hepatic triglyceride droplets via stimulation of autophagy of lipid droplets (lipophagy), reversing histological steatosis without weight loss. AMPK activation by the naringin / neohesperidin component is the upstream switch.

Third, antioxidant and anti-inflammatory action. BPF reduces oxidised LDL, downregulates LOX-1 (the endothelial oxidised-LDL receptor that initiates plaque), and dampens NF-κB-driven cytokine release in vascular tissue — shown in the rosuvastatin co-administration arms of Gliozzi et al. (2013). This is why effect sizes on subclinical atherosclerosis markers (carotid intima-media thickness) outrun what a pure lipid effect would predict.

evidence

The flagship human trial is Mollace et al. (2011): 237 patients with hyperlipidaemia (n=77), hyperlipidaemia + hyperglycaemia (n=78), or metabolic syndrome (n=42), plus a dyslipidaemia + statin-intolerance arm (n=40), randomised across three BPF doses (500, 1000, 1500 mg/day) versus placebo over 30 days. Pooled reductions at 1500 mg/day: total cholesterol ~29%, LDL ~36%, triglycerides ~39%, fasting glucose ~22%, with HDL rising ~40%. These are unusually large for a non-prescription intervention and the single biggest reason BPF is on the map; they are also the single biggest reason the field is cautious — see §3c.

The 6-month trial in moderate hypercholesterolaemia, Toth et al. (2016), is the most clinically interpretable. 80 patients, BPF 650 mg twice daily: LDL fell ~24%, small-dense LDL fell, and mean carotid intima-media thickness regressed — the first published evidence that a bergamot extract moves a subclinical atherosclerosis marker, not just a number on a lipid panel.

The combination trial — Gliozzi et al. (2013), 77 hyperlipidaemic patients — randomised rosuvastatin 10 mg or 20 mg alone, BPF 1000 mg/day alone, or rosuvastatin 10 mg + BPF. The combination matched the LDL reduction of rosuvastatin 20 mg monotherapy and added independent reductions in urinary mevalonate, oxidised LDL, LOX-1 expression, and PI3K/Akt signalling. The practical implication: BPF + low-dose statin reproduces high-dose statin lipid effect with extra antioxidant action — relevant for statin-intolerant patients capping at the lowest tolerated dose.

Capomolla et al. (2019) tested a pectin-enriched BPF formulation in 60 metabolic-syndrome patients over 12 weeks: atherogenic index dropped, waist circumference shrank ~5 cm, and body weight fell ~4 kg — modest, but consistent. Cai et al. (2017) ran a Chinese double-blind RCT (n=98 older adults with mild-to-moderate dyslipidaemia, 12 weeks): LDL reductions in the realistic 15–20% range, much smaller than Mollace 2011, supporting the suspicion that the early Calabrian trials over-state by a factor of two-ish. A Cochrane-class meta-analysis does not yet exist.

protocol

Standardised BPF 500–1500 mg/day, taken ~650 mg before each of two main meals in most trial protocols (Toth 2016, Gliozzi 2013). Onset on the lipid panel is measurable by 30 days; LDL effect appears to plateau by 8–12 weeks. The atherosclerosis-marker (carotid IMT) effect needs 6 months. The dose-response in Mollace 2011 is roughly linear from 500 to 1500 mg, with the 1500 mg arm not adding much over 1000 mg — most products land at the 1000–1300 mg/day target. Cycling is not supported by data; trials ran 4 weeks to 6 months continuously.

contraindications

Three real concerns. Diabetes medications: BPF lowers fasting glucose 15–22% in Mollace 2011; additive with sulfonylureas, insulin, or aggressive metformin titration warrants glucose monitoring. Blood thinners and CYP3A4-metabolised drugs: bergamot peel and essential oil contain bergapten (5-methoxypsoralen), a furanocoumarin that inhibits intestinal CYP3A4 in the same family as grapefruit. Standardised BPF is juice-derived and largely furanocoumarin-stripped, but commercial product variation is wide and the precautionary call is to treat it like grapefruit: avoid with statins of CYP3A4 dependence (simvastatin, lovastatin, atorvastatin to a lesser extent), with apixaban/rivaroxaban, with calcium channel blockers, and with ciclosporin until a specific product's furanocoumarin assay is in hand. Statins (intended combination): combining with rosuvastatin is the studied case (Gliozzi 2013) — rosuvastatin is not CYP3A4-metabolised, sidestepping the interaction; with CYP3A4-handled statins the precautionary line stands. Pregnancy and breastfeeding: untested; avoid. Paediatric: untested.

misconceptions

The largest misconception is product identity. Bergamot essential oil (perfume, aromatherapy, the bottle in a wellness shop), Earl Grey tea (flavoured with bergamot oil), and the medicinal juice flavonoid extract (BPF) are not interchangeable. The clinical evidence is for the standardised juice fraction; the oil contains the phototoxic furanocoumarin bergapten and is the source of the grapefruit-style drug interactions, with no evidence base for oral cholesterol use. A capsule labelled "bergamot extract" with no polyphenol percentage and no brutieridin/melitidin standardisation is not what the trials tested.

A second misconception: that BPF can replace a statin in a high-risk patient. The largest replicated LDL drops in the realistic, non-Calabrian-group literature (Cai 2017) are in the 15–20% range; even the optimistic 24% from Toth 2016 is less than half of high-intensity statin therapy and well short of what a post-MI or high-Lp(a) patient needs.

audience

Three reader profiles fit. (1) Mild-to-moderate hyperlipidaemia, low ASCVD risk, no statin indication yet, looking for a real lever before drug therapy. (2) Statin-intolerant patient capped at the lowest tolerated dose, needing incremental LDL reduction — the Gliozzi 2013 combination model. (3) Metabolic syndrome / NAFLD patient with the cluster of moderate dyslipidaemia, elevated fasting glucose, and hepatic steatosis — BPF hits all three signals at once (Capomolla 2019, Parafati 2015 for the liver mechanism). The wrong audience: established ASCVD, familial hypercholesterolaemia, post-MI — none of these should be using BPF in place of guideline lipid-lowering, where the absolute risk reduction from a statin is too large to substitute.

alternatives

The lever stack for borderline LDL: lifestyle (Mediterranean dietary pattern, replacing saturated for unsaturated fat, ~5–10% LDL), plant sterols/stanols 2 g/day (~10% LDL), soluble fibre (psyllium 10 g/day, ~7% LDL), red yeast rice (variable, contains natural monacolin K = lovastatin, real but inconsistent and now restricted in the EU), berberine 500 mg three times daily (~20% LDL, plus glucose effect), and low-dose statin (rosuvastatin 5 mg: ~40% LDL). BPF sits between berberine and low-dose statin on the effect-size ladder, with the unique selling point of the documented statin-combination synergy and the lipid + glucose + liver triple effect, and the disadvantage of a thinner non-Italian evidence base.

failure-modes

Three patterns of "I tried bergamot, nothing happened." First, the wrong product — bergamot essential oil capsules, fruit juice, or an unstandardised extract. The trials test a juice flavonoid fraction standardised to total polyphenols and ideally to brutieridin/melitidin content. Second, underdosing — sub-500 mg/day doses do not register in the dose-response from Mollace 2011. Third, too short a window — a 2-week trial will not move the lipid panel; minimum useful test is 8 weeks of fasting lipid panels before and after. A fourth pattern is over-expectation: a reader who has read the 36% LDL headline from the 1500 mg arm of Mollace 2011 and gets the 15% drop from Cai 2017 registers a "failure" against an inflated baseline.

practicalities

Quality varies sharply. The studied formulations were standardised by polyphenol content (most ~38–47%) and by HMG-flavonoid content (brutieridin + melitidin). A label that reads only "Bergamot 500 mg" with no standardisation tells the reader nothing about whether it matches trial material. Cost: a credible standardised BPF at 1000 mg/day runs roughly $20–40/month at retail. Not stocked by most pharmacies; primarily a supplement-channel product. A baseline fasting lipid panel before starting and a repeat at 12 weeks is the minimum self-monitoring needed to know if it is working for the specific reader.

stakes

The stakes are LDL-mediated atherosclerotic cardiovascular disease. Ference et al. (2017) — the European Atherosclerosis Society consensus — establishes LDL as causal for ASCVD, with the magnitude of risk reduction proportional to the cumulative LDL exposure lowered over time. The implication is that a reader with LDL 160 mg/dL who stays there for the next 30 years will have substantially more arterial plaque, and a substantially higher event rate in their 50s and 60s, than the same reader who drops to 130 mg/dL early. BPF is not a substitute for a statin where one is indicated, but for the reader whose risk does not yet warrant pharmaceutical lipid-lowering and who wants something stronger than diet alone, the cumulative-exposure logic is what makes a 20% LDL drop sustained over years matter more than the per-year effect size suggests.

payoff

Lipid panel changes are measurable at 4 weeks, clinically meaningful at 8–12 weeks (Mollace 2011, Cai 2017). The triglyceride and HDL movement tends to precede the LDL movement. Fasting glucose drops in parallel where elevated. Liver enzymes and hepatic fat (where steatosis is present) shift over 12–24 weeks (Parafati 2015 for the mechanism; human NAFLD trials are smaller and consistent with the rat model). The carotid intima-media thickness regression in Toth 2016 emerged at 6 months — the first signal that the lipid effect propagates into the vessel wall on a humanly meaningful timescale. Hard cardiovascular endpoints (MI, stroke, mortality) have not been measured; BPF lives downstream of LDL theory there.

history

The medical interest dates to Di Donna et al. (2009), when a Calabrian natural-products chemistry group isolating bergamot flavonoids identified the HMG-acylated structures and recognised the statin pharmacophore. Within two years Mollace et al. (2011) had run the dose-response human trial, and most of the subsequent clinical work has stayed within an overlapping Italian research network — see §3d for why this matters.

out-of-scope

Bergamot essential oil (aromatherapy, perfume), Earl Grey tea, bergamot peel oil for skin (a phototoxic exposure, the opposite editorial), and the broader question of how to triage LDL between lifestyle, supplements, and a statin — that triage is its own entry. The interaction with CYP3A4-metabolised drugs is touched on under contraindications but deserves its own pharmacology entry alongside grapefruit.

The credibility range

Optimist case

BPF is, mechanistically, the closest thing in the supplement aisle to an actual statin: Di Donna 2009 documents a real HMG-CoA reductase inhibitor structure in brutieridin and melitidin. The pleiotropic action — LDL down, triglycerides down, HDL up, glucose down, liver fat down, oxidised LDL down, LOX-1 expression down, and a regression signal on carotid IMT at 6 months — is broader than statins deliver. The Gliozzi 2013 combination data show real synergy with low-dose rosuvastatin, opening the statin-intolerance use case where most current options are inadequate. Several independent trial designs (single-arm dose-response, parallel-group RCT, statin-comparison, metabolic-syndrome cohort, Chinese older-adult cohort, NAFLD pilot) all show effects in the same direction. The headline 24–36% LDL effect, even if half-discounted for replication, lands at a clinically meaningful 15–20% — bigger than plant sterols, comparable to berberine, with a distinct mechanism.

Skeptic case

The largest effect sizes come from a tight research network centred on the Calabrian universities and the Italian supplement industry that commercialises BPF — the same names (Mollace, Gliozzi, Janda, Musolino, Capomolla, Ragusa) recur on the author lists across 2011, 2013, 2015, and 2019. Independent western replication is thin. Cai 2017, an external Chinese trial, reproduces the direction of effect but at roughly half the magnitude — consistent with a real effect inflated by methodological or formulation differences in the early Italian work. There is no Cochrane-class meta-analysis. There is no hard-endpoint trial. The carotid IMT regression in Toth 2016 is a single 6-month finding in 80 patients and would need replication in a larger cohort. Major lipid guidelines (AHA/ACC, ESC/EAS, NICE) do not mention bergamot. The product market is heterogeneous and largely unregulated, so what the reader buys may not match what was tested.

Author's call

BPF is a real lipid-lowering agent with a plausible statin-adjacent mechanism, and the lower-bound effect (~15–20% LDL reduction over 12 weeks) is honestly clinically useful for the right reader — borderline dyslipidaemia, metabolic syndrome, NAFLD, statin intolerance. The headline 36% LDL drop is almost certainly inflated relative to what a typical reader will get from a typical product. The entry should anchor on the discounted, replicated effect, not the headline; should call out product standardisation as the live failure mode; and should refuse the substitution use case (BPF in place of indicated statin therapy) without softening. Evidence rating: a real 3 — multiple consistent trials, plausible mechanism, but most large effects from a single research network, no hard endpoints, no major guideline. Controversy: 2 — the field has not actively fought about bergamot the way it fights about red yeast rice or saturated fat; it is mostly a question of how much to discount, not whether to dismiss.

Stakeholder + incentive map

Calabrian agro-industrial complex. Bergamot is a regional crop and a regional identity; the juice fraction transitions a waste stream into a high-margin nutraceutical. The dominant research group (Mollace and collaborators, Magna Graecia University and University of Catanzaro) overlaps with the supplement supply chain commercialising BPF.
Supplement manufacturers and the broader nutraceutical channel — strong incentive to amplify the 36% LDL headline; weak incentive to clarify product standardisation differences.
Cardiology guideline bodies (AHA/ACC, ESC/EAS, NICE) — no incentive to elevate a single-region nutraceutical without hard-endpoint trials; silence in guidelines is the default, not a verdict.
Independent natural-medicine clinicians and lipidologists — split: some have incorporated BPF as a statin-intolerance adjunct (especially in the US functional-medicine channel); others remain sceptical given the source concentration of the evidence.
Statin manufacturers — no active counter-incentive (BPF is too small to threaten the statin market); the silence on the establishment side is more apathy than opposition.

Population variability

Best response: baseline LDL 160–220 mg/dL, baseline triglycerides elevated, metabolic syndrome cluster present. The bigger the baseline lipid abnormality, the bigger the absolute reduction — consistent with the dose-response in Mollace 2011. Diminishing returns: already-treated patients on optimal statin doses are unlikely to see much additional LDL reduction beyond what Gliozzi 2013's combination showed at low-dose statin baseline. Unstudied: paediatric populations, pregnancy, severe familial hypercholesterolaemia (where the absolute lipid burden vastly exceeds what any nutraceutical can move), advanced CKD, post-transplant. The trials are predominantly Italian and Chinese; African, South Asian, and Indigenous populations are not represented in the trial set. Genetic variation in flavonoid metabolism (UGT, SULT polymorphisms) is suspected to drive individual variability but not yet quantified for BPF specifically.

Knowledge gaps

Hard cardiovascular endpoints. No trial measures MI, stroke, or mortality. The carotid IMT signal in Toth 2016 is the closest proxy.
Independent replication of the largest effect sizes. The 24–36% LDL drops in the Calabrian-group trials need a large, externally funded, multi-centre western RCT. Cai 2017 provides one external data point at half the magnitude — useful but not sufficient.
Head-to-head against low-dose statin monotherapy. The combination data exist; the substitution data do not.
Product standardisation. Most commercial products do not report brutieridin/melitidin content. Without that, the link between consumer product and trial material is broken.
Furanocoumarin content of commercial BPF. The juice-derived extraction should strip bergapten, but no published bench survey of consumer products quantifies this — so the CYP3A4-interaction precaution is currently a category-level call, not a product-level one.
Long-term safety beyond 6 months. No chronic-use safety signal has emerged, but no trial has formally measured it past 6 months.

Brief coverage. The input description named cholesterol, blood sugar, liver fat, and cardiovascular risk; all four are covered. Cholesterol carries the headline framing because that is where the largest replicated effect lives and the largest reader audience sits, but blood sugar (evidence + payoff), liver fat (mechanism + payoff), and cardiovascular risk (stakes + payoff via the carotid IMT regression in Toth 2016) are not silently dropped. No narrowing relative to the brief.

The "halve the headline" call. The single hardest editorial decision was where to anchor the LDL effect. The Mollace-group headline is ~36%; the external Chinese replication (Cai 2017) is ~15–20%; Toth 2016 sits at 24% at 6 months. The article anchors the reader on 15–25% across both highlights and evidence, and explicitly flags the 36% number as a likely-inflated headline. Disagreement with that call (some functional-medicine clinicians take the 36% figure at face value) is the chief epistemic risk on this entry.

Concentrated research-network bias. The dominant body of trial evidence comes from one Calabrian network (Mollace, Gliozzi, Janda, Musolino, Capomolla, Ragusa) overlapping with the supplement supply chain. This is named in the research dossier's credibility range and shapes the evidence: 3 ceiling. A future large, independently-funded western RCT could move evidence to 4; current state does not warrant it.

Longevity scored 3, not 4. No hard cardiovascular endpoint trial exists. The carotid IMT regression at 6 months in Toth 2016 is the best proxy and supports a meaningful longevity claim through cumulative-LDL-exposure logic (Ference 2017), but absent MI/stroke/mortality data the 4-tier (population-bending) label would be overclaiming.

Furanocoumarin contraindication is category-level, not product-level. No published bench survey quantifies bergapten content across consumer BPF products. The contraindications callout treats any CYP3A4-handled drug combination as a grapefruit-style interaction by default — the precautionary call until product-specific data emerge. This may over-warn on well-purified products and under-warn on the worst ones; the article cannot do better than the published bench data allows.

Citations excluded, but considered. Several Italian-group follow-ups (Mollace 2019 on NASH, Janda 2016 on AMPK / lipid biochemistry, Gliozzi 2014 NAFLD pilot, Musolino 2019 mechanism work) were not added because the DOIs could not be verified to the bright-line standard. The six trials and one consensus paper that did make it carry the load; the dossier does not bottom out for lack of the others.

Future-link candidates (entries that don't exist yet but this should cross-link to once they do):

The lipid panel itself (what the numbers mean; ApoB vs LDL-C) — the obvious sibling.
The diet → supplement → statin triage entry — referenced in out-of-scope as a "decision tree, not a supplement entry."
Berberine, plant sterols/stanols, psyllium, red yeast rice, low-dose rosuvastatin — each its own supplement / pharma entry; the alternatives section reads as a backbone for those cross-links.
Bergamot essential oil — the partly-opposite cousin; should be its own warning-flavoured entry (phototoxicity, CYP3A4 interaction, no cholesterol use).
Metabolic syndrome, NAFLD — both warrant standalone entries that BPF would link back to.

Dream tier. Overall score lands around 20 (well below the 40 obligatory threshold). A short dream narrative was written by choice because the relief lever ("the cholesterol talk that doesn't escalate") is honest and earns the dek and tagline meaningfully more voltage than a straight write would. The lever is relief inside a small do; the tagline carries it specifically.