Place · Level 3 · Debunk
Weight-loss supplements debunked · what works, what's a scam
$30B/年市场 · 多数 RCT 证据微弱或缺失 · 没有补剂减重超过 1-2 kg · 与 GLP-1 阶梯相差一个数量级
Story path
- 1$30B myth · regulatory truth$30B myth · regulatory truth
- 2'Some signal' group · meta-analysis review'Some signal' group · meta-analysis review
- 3Pure-fraud group · ~ 0 clinical evidencePure-fraud group · ~ 0 clinical evidence
- 4'Natural = safe' trap · adaptogens'Natural = safe' trap · adaptogens
- 5What's actually worth your moneyWhat's actually worth your money
Chapter 1
$30B myth · regulatory truth
$30B myth · regulatory truth
Global weight-loss supplement market > $30B/year, ~ 60% of US adults have tried at least one (NIH ODS 2023). The marketing playbook never changes:
'Natural = safe = effective''No diet, no exercise needed''Burns stubborn fat' / 'Blocks carbs/fat absorption''Celebrity X lost 20 lbs in 30 days'
FDA does not pre-approve supplements (DSHEA 1994): supplements are classified as 'food'; manufacturers don't need to prove efficacy or safety before selling. FDA can only recall after harm (e.g., ephedra was banned in 2004 only after deaths). This means most fat-loss miracle pills on the shelf have never passed a single RCT.
The reference table (memorize before reading any ad):
Bottom line: no marketed supplement has achieved ≥ 2 kg / 6-month clinically meaningful weight loss in rigorous RCTs. You're probably buying a placebo + marketing + packaging. The next three pages dissect the categories: ① things with weak signals (green tea / CLA / vinegar), ② pure fraud (raspberry ketone / synephrine / Garcinia), ③ the 'natural adaptogen' trap (Ashwagandha et al.).
'Natural = safe = effective''No diet, no exercise needed''Burns stubborn fat' / 'Blocks carbs/fat absorption''Celebrity X lost 20 lbs in 30 days'
FDA does not pre-approve supplements (DSHEA 1994): supplements are classified as 'food'; manufacturers don't need to prove efficacy or safety before selling. FDA can only recall after harm (e.g., ephedra was banned in 2004 only after deaths). This means most fat-loss miracle pills on the shelf have never passed a single RCT.
The reference table (memorize before reading any ad):
| Intervention | 6-12 mo weight loss | Evidence |
|---|---|---|
| Any supplement | 0-2 kg | weak / none |
| Diet + exercise (lifestyle) | 5-10 kg | strong |
| GLP-1 (semaglutide 2.4 mg) | ~15 kg | very strong (STEP-1) |
| GIP/GLP-1 (tirzepatide 15 mg) | ~21 kg | very strong (SURMOUNT-1) |
| Bariatric surgery (sleeve/RYGB) | 25-35% TBW | very strong |
Bottom line: no marketed supplement has achieved ≥ 2 kg / 6-month clinically meaningful weight loss in rigorous RCTs. You're probably buying a placebo + marketing + packaging. The next three pages dissect the categories: ① things with weak signals (green tea / CLA / vinegar), ② pure fraud (raspberry ketone / synephrine / Garcinia), ③ the 'natural adaptogen' trap (Ashwagandha et al.).
Chapter 2
'Some signal' group · meta-analysis review
'Some signal' group · meta-analysis review
This group has a few statistically significant meta-analyses, but all effect sizes fall below the clinically meaningful threshold (≥ 5% TBW loss is the usual bar, i.e. ≥ 3.5 kg for a 70 kg person).
① Green tea extract (EGCG) — Cochrane review (Jurgens 2012):
18 RCTs, n = 1945, mean weight loss 0.95 kg / 12-13 weeksTranslation: < 2 lbs in three months — clinically weakHepatotoxicity: US DILIN network has logged > 80 cases of acute liver injury from high-dose EGCG (> 800 mg/day), some requiring transplant'Green tea burns fat' is at best a mild beverage side effect, not a supplement target
② CLA (conjugated linoleic acid) — Onakpoya 2012 meta:
18 RCTs, weight loss rate 0.05 kg/week (~ 1.3 kg / 6 mo)Clinical meaning near zeroSome trials: ↓ HDL + ↑ LDL + ↑ insulin resistance (adverse cardiometabolic effects)Animal models: induces hepatic steatosis (opposite of marketing)
③ Apple cider vinegar:
Kondo 2009 (Japan 12-wk RCT, n = 175): 15 mL/d -1.2 kg, 30 mL/d -1.9 kgSubsequent replications mostly small samples + methodological issues (no blinding / self-reported weight / short duration)2024 Khezri Lebanon RCT (n = 120) showed larger effect, but manufacturer-funded + design controversiesMechanism hypothesis (satiety + slowed gastric emptying) is weak; clinical effect likely driven by concurrent diet changes
④ Berberine → see existing story berberine:
Primarily glucose-lowering + lipid improvement, weak weight effect (meta ~ 1-2 kg)'Nature's Ozempic' marketing is misleading: 5-10× weaker than semaglutide for weight loss
Why even the 'somewhat effective' ones aren't worth it: the same money spent on +1500 steps/day + 20 g/day protein would give an expected 3-5 kg loss over 6 months — 2-3× the sum of everything above, with no hepatotoxic risk.
① Green tea extract (EGCG) — Cochrane review (Jurgens 2012):
18 RCTs, n = 1945, mean weight loss 0.95 kg / 12-13 weeksTranslation: < 2 lbs in three months — clinically weakHepatotoxicity: US DILIN network has logged > 80 cases of acute liver injury from high-dose EGCG (> 800 mg/day), some requiring transplant'Green tea burns fat' is at best a mild beverage side effect, not a supplement target
② CLA (conjugated linoleic acid) — Onakpoya 2012 meta:
18 RCTs, weight loss rate 0.05 kg/week (~ 1.3 kg / 6 mo)Clinical meaning near zeroSome trials: ↓ HDL + ↑ LDL + ↑ insulin resistance (adverse cardiometabolic effects)Animal models: induces hepatic steatosis (opposite of marketing)
③ Apple cider vinegar:
Kondo 2009 (Japan 12-wk RCT, n = 175): 15 mL/d -1.2 kg, 30 mL/d -1.9 kgSubsequent replications mostly small samples + methodological issues (no blinding / self-reported weight / short duration)2024 Khezri Lebanon RCT (n = 120) showed larger effect, but manufacturer-funded + design controversiesMechanism hypothesis (satiety + slowed gastric emptying) is weak; clinical effect likely driven by concurrent diet changes
④ Berberine → see existing story berberine:
Primarily glucose-lowering + lipid improvement, weak weight effect (meta ~ 1-2 kg)'Nature's Ozempic' marketing is misleading: 5-10× weaker than semaglutide for weight loss
Why even the 'somewhat effective' ones aren't worth it: the same money spent on +1500 steps/day + 20 g/day protein would give an expected 3-5 kg loss over 6 months — 2-3× the sum of everything above, with no hepatotoxic risk.
Chapter 3
Pure-fraud group · ~ 0 clinical evidence
Pure-fraud group · ~ 0 clinical evidence
Common thread: no high-quality RCT shows meaningful weight loss, but in-vitro / animal data get amplified by marketing, keeping these alive on e-commerce and TikTok.
① Raspberry ketone
Went viral after Dr. Oz called it a 'miracle fat burner' in 2012Human RCT count: 0 (literally zero)Sole supporting data: rats / isolated adipocytes + one 8-week open-label study (no control)Rat equivalent dose scaled to humans: ~ 90 mg/kg/day, i.e. 6.3 g/day for a 70 kg adult (commercial supplements are 100-200 mg, 1/30 to 1/60 of that)Verdict: textbook 'marketing > evidence' case
② Bitter orange (synephrine / Citrus aurantium)
The 'legal replacement' after ephedra ban (2004)Same mechanism: α/β adrenergic agonism, CNS stimulation + ↑ HR + ↑ BPWeight effect: meta ~ 1.3 kg / 6-12 weeks (mostly from co-administered caffeine, not synephrine itself)Adverse events: hypertensive crisis / arrhythmia / stroke / MI case reports (especially combined with caffeine)Banned for athletes in Canada and other sport bodiesRisk > benefit
③ Garcinia cambogia (HCA, hydroxycitric acid)
Marketing: 'blocks fat-synthesis enzyme (adenosine triphosphate: The cell's universal energy currency — almost everything that costs energy spends it.-citrate lyase)'2011 Onakpoya meta-analysis (12 RCTs): mean weight difference 0.88 kg vs placebo, borderline clinical meaning, high heterogeneityMajor rigorous RCT (Heymsfield 1998 JAMA, 12 weeks, n = 135) negativeLiver injury: US DILIN has logged dozens of Garcinia-associated acute liver failure cases, some requiring transplantFDA 2017 warning
④ White kidney bean extract (Phaseolus vulgaris, α-amylase inhibitor)
Marketed as 'carb blocker — eat rice and bread without gaining'In vitro: inhibits α-amylase → reduced starch digestionHuman RCTs: a few small (n = 60-100) studies show 12-week loss 1-3 kg vs placebo, high heterogeneityBorderline clinical meaning, long-term data missing, GI bloating common (undigested starch reaches colon)
⑤ Chromium picolinate → see existing story chromium:
Meta: mean loss ~ 0.5-1 kg, near-zero clinical meaningHigh-dose liver/kidney toxicity case reports
Common pattern: ① in-vitro / animal effects amplified into human marketing claims ② 1-2 small studies cited endlessly ③ large / independent / long-term RCTs nearly all negative ④ side effects downplayed under the 'natural' label.
① Raspberry ketone
Went viral after Dr. Oz called it a 'miracle fat burner' in 2012Human RCT count: 0 (literally zero)Sole supporting data: rats / isolated adipocytes + one 8-week open-label study (no control)Rat equivalent dose scaled to humans: ~ 90 mg/kg/day, i.e. 6.3 g/day for a 70 kg adult (commercial supplements are 100-200 mg, 1/30 to 1/60 of that)Verdict: textbook 'marketing > evidence' case
② Bitter orange (synephrine / Citrus aurantium)
The 'legal replacement' after ephedra ban (2004)Same mechanism: α/β adrenergic agonism, CNS stimulation + ↑ HR + ↑ BPWeight effect: meta ~ 1.3 kg / 6-12 weeks (mostly from co-administered caffeine, not synephrine itself)Adverse events: hypertensive crisis / arrhythmia / stroke / MI case reports (especially combined with caffeine)Banned for athletes in Canada and other sport bodiesRisk > benefit
③ Garcinia cambogia (HCA, hydroxycitric acid)
Marketing: 'blocks fat-synthesis enzyme (adenosine triphosphate: The cell's universal energy currency — almost everything that costs energy spends it.-citrate lyase)'2011 Onakpoya meta-analysis (12 RCTs): mean weight difference 0.88 kg vs placebo, borderline clinical meaning, high heterogeneityMajor rigorous RCT (Heymsfield 1998 JAMA, 12 weeks, n = 135) negativeLiver injury: US DILIN has logged dozens of Garcinia-associated acute liver failure cases, some requiring transplantFDA 2017 warning
④ White kidney bean extract (Phaseolus vulgaris, α-amylase inhibitor)
Marketed as 'carb blocker — eat rice and bread without gaining'In vitro: inhibits α-amylase → reduced starch digestionHuman RCTs: a few small (n = 60-100) studies show 12-week loss 1-3 kg vs placebo, high heterogeneityBorderline clinical meaning, long-term data missing, GI bloating common (undigested starch reaches colon)
⑤ Chromium picolinate → see existing story chromium:
Meta: mean loss ~ 0.5-1 kg, near-zero clinical meaningHigh-dose liver/kidney toxicity case reports
Common pattern: ① in-vitro / animal effects amplified into human marketing claims ② 1-2 small studies cited endlessly ③ large / independent / long-term RCTs nearly all negative ④ side effects downplayed under the 'natural' label.
Chapter 4
'Natural = safe' trap · adaptogens
'Natural = safe' trap · adaptogens
'Natural' is the strongest hook in modern supplement marketing. Three common logic traps must be exposed:
Trap 1: natural ≠ effective
Ashwagandha: moderate stress/sleep data (1-2 RCTs showing cortisol ↓), weight evidence very weak: just 1-2 small studies (n < 50) showing 1-2 kg loss / 8 weeks, not conclusiveIts 'fat loss' claim rides on the 'stress ↓ → cortisol ↓ → abdominal fat ↓' syllogism — every step is unvalidated in RCTGreen coffee bean extract (chlorogenic acid): 2012 Vinson RCT (n = 16) showed 22-wk -8 kg, but the study was retracted and the author sued by FTC for fraud (2014). Independent replications negative.
Trap 2: natural ≠ safe
Ephedra: completely natural, used in TCM for millennia. FDA banned 2004 after > 155 deaths (arrhythmia / stroke / MI)Kava: traditional Pacific Islander drink, banned in Germany / France / Canada for hepatotoxicityBitter orange (synephrine): see previous pageGreen tea extract (high-dose EGCG): see meta pageGarcinia: see previous pageShared lesson: 'plant-derived' does not imply molecular safety. Strychnine / tetrodotoxin / ricin are also natural.
Trap 3: natural vs drug — orders-of-magnitude effect gap
Gap: prescription vs natural supplement weight effects differ by 10-20×. This isn't 'gentle and natural' — it's 'natural but essentially inert'.
The right framing:
If a natural product had a truly strong weight-loss effect, it would either have major side effects (ephedra) or get developed into a prescription drug (cf. paclitaxel / metformin from plants)A 'no side effect + highly effective' natural weight-loss agent does not exist — if it had, pharma would have patented it long agoglucagon-like peptide-1: A gut hormone released after eating that makes you feel full and helps lower blood sugar. itself is a 'natural' hormone — scientists just chemically modified it to last 30 days in the body. That's the proper 'natural molecule + rigorous development' route.
Trap 1: natural ≠ effective
Ashwagandha: moderate stress/sleep data (1-2 RCTs showing cortisol ↓), weight evidence very weak: just 1-2 small studies (n < 50) showing 1-2 kg loss / 8 weeks, not conclusiveIts 'fat loss' claim rides on the 'stress ↓ → cortisol ↓ → abdominal fat ↓' syllogism — every step is unvalidated in RCTGreen coffee bean extract (chlorogenic acid): 2012 Vinson RCT (n = 16) showed 22-wk -8 kg, but the study was retracted and the author sued by FTC for fraud (2014). Independent replications negative.
Trap 2: natural ≠ safe
Ephedra: completely natural, used in TCM for millennia. FDA banned 2004 after > 155 deaths (arrhythmia / stroke / MI)Kava: traditional Pacific Islander drink, banned in Germany / France / Canada for hepatotoxicityBitter orange (synephrine): see previous pageGreen tea extract (high-dose EGCG): see meta pageGarcinia: see previous pageShared lesson: 'plant-derived' does not imply molecular safety. Strychnine / tetrodotoxin / ricin are also natural.
Trap 3: natural vs drug — orders-of-magnitude effect gap
| Intervention | Weight loss (52 wk, kg) | Class |
|---|---|---|
| Ashwagandha (limited data) | ~ 1-2 | 'natural' |
| Green tea EGCG | ~ 1 | 'natural' |
| CLA | ~ 1.3 | 'natural' |
| Berberine | ~ 1-2 | 'natural' |
| Semaglutide 2.4 mg | ~ 15 | prescription |
| Tirzepatide 15 mg | ~ 21 | prescription |
Gap: prescription vs natural supplement weight effects differ by 10-20×. This isn't 'gentle and natural' — it's 'natural but essentially inert'.
The right framing:
If a natural product had a truly strong weight-loss effect, it would either have major side effects (ephedra) or get developed into a prescription drug (cf. paclitaxel / metformin from plants)A 'no side effect + highly effective' natural weight-loss agent does not exist — if it had, pharma would have patented it long agoglucagon-like peptide-1: A gut hormone released after eating that makes you feel full and helps lower blood sugar. itself is a 'natural' hormone — scientists just chemically modified it to last 30 days in the body. That's the proper 'natural molecule + rigorous development' route.
Chapter 5
What's actually worth your money
What's actually worth your money
If you really want to spend money on 'supplementation', these might be worth it — but understand: none of them is a 'fat-loss miracle'. They support the body indirectly, not burn fat directly.
① Protein powder (whey / casein / plant)
Core role: helps you hit 1.6-2.4 g/kg/day total protein, preserving muscle during deficitMorton 2018 meta (49 RCTs, n = 1863): resistance training + protein adds 0.3 kg lean mass, +9% strengthLongland 2016: high protein (2.4 g/kg) + training → simultaneous fat loss 4.8 kg + muscle gain 1.2 kg (4 weeks)Not 'fat burning', but 'muscle preservation during cut'See protein-during-deficit
② Caffeine
Modestly raises BMR (~ 3-11%) + improves exercise performance (endurance + strength + HIIT)Source preference: coffee / tea / dark chocolate (with polyphenols) > pure caffeine pillsDoesn't require 'supplement dosing' — 200-400 mg/day (1-3 cups) is plenty
③ Creatine monohydrate
Strongest evidence base in sports nutrition, > 1000 RCTsTraining performance + muscle mass + muscle preservation during cut, indirect body-composition help3-5 g/day, no 'loading' needed, cheap (~ $0.10/day)
④ Fish oil (omega-3, EPA + DHA)
Cardiovascular + anti-inflammatory + neurocognitive, not a fat-loss toolREDUCE-IT (EPA 4 g/day) cut MACE 25% in high-TG / high-CV-risk patientsGeneral population 1-2 g/day EPA+DHA is reasonable
True weight-loss ROI ranking (same dollars spent):
1. Good food (quality protein + vegetables + whole grains, replacing UPF) — vastly beats any supplement
2. Gym membership / a pair of running shoes — strength + aerobic, dual metabolic benefit
3. Sleep (blackout curtains / earplugs / cool bedroom) — sleep < 6h has RR 1.5 for obesity
4. Bathroom scale + food scale — self-quantification, core tool for behavior change
5. (If medically indicated) glucagon-like peptide-1: A gut hormone released after eating that makes you feel full and helps lower blood sugar. — physician evaluation, $300-1300/month but 5-15 kg effect, cost-per-kg far below supplements
Atlas connections:
weight-management-foundations — weight-management frameworkglp1-agonists-deep — the real drug comparisonprotein-during-deficit — protein strategy in deficitberberine · chromium — already debunked in depth
One-line summary: cut the $50-100/month supplement budget, replace it with 2 kg more protein + a gym membership + a good pillow — six months later you'll have lost more weight, and you'll understand why.
① Protein powder (whey / casein / plant)
Core role: helps you hit 1.6-2.4 g/kg/day total protein, preserving muscle during deficitMorton 2018 meta (49 RCTs, n = 1863): resistance training + protein adds 0.3 kg lean mass, +9% strengthLongland 2016: high protein (2.4 g/kg) + training → simultaneous fat loss 4.8 kg + muscle gain 1.2 kg (4 weeks)Not 'fat burning', but 'muscle preservation during cut'See protein-during-deficit
② Caffeine
Modestly raises BMR (~ 3-11%) + improves exercise performance (endurance + strength + HIIT)Source preference: coffee / tea / dark chocolate (with polyphenols) > pure caffeine pillsDoesn't require 'supplement dosing' — 200-400 mg/day (1-3 cups) is plenty
③ Creatine monohydrate
Strongest evidence base in sports nutrition, > 1000 RCTsTraining performance + muscle mass + muscle preservation during cut, indirect body-composition help3-5 g/day, no 'loading' needed, cheap (~ $0.10/day)
④ Fish oil (omega-3, EPA + DHA)
Cardiovascular + anti-inflammatory + neurocognitive, not a fat-loss toolREDUCE-IT (EPA 4 g/day) cut MACE 25% in high-TG / high-CV-risk patientsGeneral population 1-2 g/day EPA+DHA is reasonable
True weight-loss ROI ranking (same dollars spent):
1. Good food (quality protein + vegetables + whole grains, replacing UPF) — vastly beats any supplement
2. Gym membership / a pair of running shoes — strength + aerobic, dual metabolic benefit
3. Sleep (blackout curtains / earplugs / cool bedroom) — sleep < 6h has RR 1.5 for obesity
4. Bathroom scale + food scale — self-quantification, core tool for behavior change
5. (If medically indicated) glucagon-like peptide-1: A gut hormone released after eating that makes you feel full and helps lower blood sugar. — physician evaluation, $300-1300/month but 5-15 kg effect, cost-per-kg far below supplements
Atlas connections:
weight-management-foundations — weight-management frameworkglp1-agonists-deep — the real drug comparisonprotein-during-deficit — protein strategy in deficitberberine · chromium — already debunked in depth
One-line summary: cut the $50-100/month supplement budget, replace it with 2 kg more protein + a gym membership + a good pillow — six months later you'll have lost more weight, and you'll understand why.