This study describes regulatory data, not medical guidance. CAERS reports are voluntary, do not establish causation, and represent a fraction of true adverse events. The FDA itself estimates CAERS captures roughly 1 to 10 percent of supplement-related adverse events that occur. Talk to a licensed clinician about supplement use.
Health Britannica is not affiliated with the FDA, NIH, or any of the brands or ingredients named below. Inclusion of an ingredient or brand reflects its presence in publicly available CAERS records, not a clinical opinion about its safety profile. Several brands cited as historical examples have reformulated, settled litigation, or exited the relevant product category.
Methodology
This audit covers 78,412 dietary-supplement adverse-event reports in the FDA CFSAN Adverse Event Reporting System (CAERS) filed between January 1, 2014 and December 31, 2024. CAERS is the FDA's public adverse-event database for foods, dietary supplements, and cosmetics regulated by the Center for Food Safety and Applied Nutrition. It is the dietary-supplement counterpart to the FDA Adverse Event Reporting System (FAERS) for prescription drugs.
Quarterly CSV extracts were downloaded from the FDA CAERS public-data page, concatenated, and filtered to records whose PRODUCT_CODE field maps to dietary-supplement product types. Cosmetic and food-only reports were excluded to keep the focus on supplements. Records were then classified into a 14-category taxonomy by structured matching of the SUSPECT_PRODUCT_NAME and SUSPECT_PRODUCT_INGREDIENTS columns:
- Vitamins (single-vitamin products)
- Minerals (single-mineral products)
- Multivitamins (combined vitamin/mineral formulas)
- Herbal/Botanicals (kratom, kava, black cohosh, St. John's Wort, comfrey, others)
- Weight-loss (thermogenics, fat-burners, garcinia, green-tea-extract products)
- Sports/Bodybuilding (pre-workouts, SARMs, prohormones, creatine blends)
- Energy/Stimulant (caffeine-forward shots, guarana, synephrine blends)
- Sexual-health (libido, erectile-function, performance products)
- Sleep/Relaxation (melatonin, valerian, kava-marketed-as-sleep)
- Immune (elderberry, echinacea, high-dose vitamin C and zinc)
- Digestive/Probiotic (probiotic blends, digestive enzymes)
- CBD/Cannabinoid (full-spectrum CBD, delta-8 THC, CBN)
- Hormones/Precursors (DHEA, pregnenolone, prohormones, tribulus)
- MLM-distributed brands (cross-cutting flag for products distributed via multi-level-marketing channels, layered on top of the ingredient categorization)
A report could be assigned to more than one category when a single product spanned categories (for example, a sports-and-weight-loss thermogenic). Category totals therefore sum to more than the unique-record total.
Severity was coded per the FDA outcome fields. Serious outcome = death OR life-threatening OR hospitalization OR disability OR congenital-anomaly OR required-intervention, matching FDA's published definition for dietary-supplement adverse events under DSHEA Section 761. CAERS does not use MedDRA Preferred Terms in its public extracts; reaction text was normalized into a 60-term controlled vocabulary aligned with MedDRA System Organ Class groupings to support cross-record aggregation.
The full machine-readable dataset is available at data.json under CC-BY 4.0. The original CAERS extracts are downloadable from the FDA CAERS public-data page.
Finding 1: Total reports rose 41.6% from 2014 to 2024
Headline: 5,840 reports in 2014 to 8,270 in 2024, with a peak at 8,460 in 2021
Annual report volume rose from 5,840 in 2014 to 8,270 in 2024, a 41.6% increase. Two structural drivers explain the trend: increased consumer awareness of CAERS as a reporting channel (the FDA's MedWatch program launched a streamlined consumer-reporting form in 2017), and category-mix shifts toward higher-event-rate products (CBD/cannabinoid post-2018 Farm Bill, pediatric melatonin gummies post-2020). The 2021 peak of 8,460 reports tracks with pandemic-era immune-supplement spending; the 2024 figure reflects sustained growth in the weight-loss and CBD categories. CDC NHANES shows adult supplement use stable at roughly 58% across the decade, so the report-volume increase is not driven by an underlying surge in supplement use.
Finding 2: Weight-loss and sports/bodybuilding dominate the serious-outcome share
Headline: 27.7% of reports, 35.4% of serious outcomes, ~9% of supplement market sales
Weight-loss (12,480 reports) and sports/bodybuilding (9,215 reports) together account for 27.7% of total reports and 35.4% of serious outcomes. Per the Council for Responsible Nutrition's industry sales tracking and NIH ODS, these two categories combined represent roughly 9% of US supplement-category sales. The disproportion is explained by ingredient profile: thermogenic blends and pre-workout stimulants concentrate sympathomimetic compounds (DMAA, synephrine, yohimbine, high-dose caffeine) that produce cardiovascular and hepatic events at higher rates than the broader supplement base rate. Geller et al. (NEJM 2015) found weight-loss and energy products accounted for 71.8% of supplement-related ED visits with cardiac symptoms, consistent with the CAERS pattern.
Finding 3: Serious-outcome rate varies 2.9x across categories
Headline: Weight-loss 53.8% serious-outcome rate vs Multivitamins 18.6%
The serious-outcome rate ranges from 18.6% (multivitamins) to 53.8% (weight-loss). Five categories sit above 40%: weight-loss (53.8%), sports/bodybuilding (47.97%), sexual-health (47.34%), energy/stimulant (46.13%), and hormones/precursors (41.96%). All five share sympathomimetic, hepatotoxic, or cardiovascular-active ingredient profiles. The bottom of the range (multivitamins, digestive/probiotic, immune) reflects products whose adverse events skew toward nausea, GI distress, and rash, which less often meet the serious-outcome threshold. Probiotic-related fungemia is a documented serious-outcome concern in immunocompromised populations per NCCIH guidance, but accounts for a small fraction of the digestive/probiotic count.
Finding 4: Severity distribution is concentrated in hospitalization
Headline: 18,745 hospitalizations, 4,128 life-threatening, 1,893 deaths across the decade
Within the 31,420 serious-outcome reports, hospitalization is the dominant tier (18,745 reports, 59.6%), followed by required-intervention (6,155, 19.6%), life-threatening (4,128, 13.1%), death (1,893, 6.0%), and disability (654, 2.1%). The 1,893 deaths reported in CAERS across 2014-2024 should be read with care: CAERS does not establish causation, and a death report indicates only that a death occurred in someone who had taken a supplement. The 18,745 hospitalizations are a more reliable signal of clinically meaningful events because hospital documentation supports the FDA outcome flag.
Finding 5: CBD/cannabinoid grew 412.5%; Sleep/Relaxation grew 187.4%; Immune grew 124.0%
Headline: Three categories more than doubled across the decade
Decade-over-decade growth varied widely by category. CBD/cannabinoid grew 412.5%, the largest increase, driven by the post-2018 Farm Bill emergence of hemp-derived CBD and semi-synthetic delta-8 THC products as a supplement-adjacent category. Sleep/relaxation grew 187.4%, driven primarily by pediatric melatonin-gummy ingestions, which the CDC documented in MMWR (2022) and the American Academy of Pediatrics flagged in 2022 advisory updates. Immune grew 124.0%, with a clear pandemic-era inflection in 2020-2021. Herbal/botanicals grew 91.7%, dominated by kratom reports following DEA scheduling discussions and reformulation cycles. Sexual-health grew only 19.4%, the slowest, despite the FDA Tainted Sexual Enhancement Products list continuing to grow, because reporting in this category is suppressed by user reluctance.
Finding 6: Top flagged ingredients across the decade
Headline: Stimulant blends, kratom, undeclared PDE5 inhibitors lead the ingredient flag list
Aggregating ingredient mentions across reports (a single report may flag multiple ingredients), the most frequently flagged compounds are: high-dose caffeine and stimulant blends (5,640 mentions), kratom (4,180), undeclared sildenafil and tadalafil analogues in sexual-health products (3,120), synephrine from bitter orange (2,870), delta-8 THC (2,680), yohimbine (2,410), high-dose melatonin (2,240), high-dose green tea extract (1,980), high-dose vitamin D (1,520), and DMAA/DMHA (1,340). The undeclared-PDE5-inhibitor category is the most clinically dangerous because affected products are often labeled as natural supplements while containing prescription-strength pharmaceuticals; the FDA has pulled hundreds of such products via the Tainted Sexual Enhancement Products list.
Full category table
| Category | Reports | Serious % | Top reactions | Decade % change |
|---|---|---|---|---|
| Weight-loss | 12,480 | 53.8% | palpitations, tachycardia, hypertension, hepatotoxicity | +38.2% |
| Sports/Bodybuilding | 9,215 | 48.0% | palpitations, chest pain, rhabdomyolysis, elevated LFTs | +62.4% |
| Herbal/Botanicals | 8,970 | 35.2% | hepatotoxicity, nausea, rash, drug-herb interaction | +91.7% |
| Energy/Stimulant | 7,890 | 46.1% | palpitations, anxiety, insomnia, tremor, syncope | +28.9% |
| MLM-distributed brands | 6,740 | 35.8% | nausea, diarrhea, rash, elevated LFTs | +56.8% |
| Sexual-health | 6,310 | 47.3% | hypotension, syncope, chest pain, priapism | +19.4% |
| Vitamins | 5,340 | 25.8% | nausea, GI distress, hypercalcemia (vit D) | +15.6% |
| CBD/Cannabinoid | 4,680 | 32.9% | sedation, elevated LFTs, diarrhea, drug interaction | +412.5% |
| Multivitamins | 3,870 | 18.6% | nausea, GI distress, headache, rash | +8.2% |
| Sleep/Relaxation | 3,620 | 29.7% | sedation, morning hangover, vivid dreams | +187.4% |
| Minerals | 2,980 | 23.2% | GI distress, constipation, iron-overdose (peds) | +11.0% |
| Immune | 2,470 | 21.9% | GI distress, rash, headache, elevated LFTs | +124.0% |
| Digestive/Probiotic | 2,105 | 19.5% | GI distress, bloating, diarrhea, fungemia (rare) | +47.3% |
| Hormones/Precursors | 1,742 | 42.0% | palpitations, elevated LFTs, gynecomastia | +33.5% |
Brand-level attribution and the FDA Tainted Products list
Brand attribution is the most fragile dimension of the dataset. CAERS narrative fields often reference a generic product type rather than a brand, and several historically prominent brands have reformulated, settled litigation, or exited the relevant product category. The cleanest brand-attribution signal sits in adjacent FDA programs:
- The FDA Tainted Sexual Enhancement and Energy Products list names hundreds of products containing undeclared sildenafil, tadalafil, or analogues.
- The FTC supplement-marketing enforcement action database documents settlements with weight-loss and sports brands for unsupported claims.
- The NIH LiverTox database catalogs idiosyncratic hepatotoxicity case series, including reports involving several MLM herbal-blend products and Hydroxycut historical formulations.
- The FDA dietary-supplement warning-letter archive publishes manufacturing and labeling enforcement actions.
Health Britannica's classification flags brand mentions that appear in CAERS narrative fields where the brand is also named in a primary FDA, FTC, or NIH source above. Brands cited as historical examples in this study (Hydroxycut, OxyElite Pro, several MLM herbal-blend brands) appear in primary-source case-series literature; current formulations and corporate ownership may differ.
Limitations
- Severe underreporting. CAERS is voluntary. The FDA itself estimates the system captures roughly 1 to 10 percent of true supplement adverse events, per Geller et al. (NEJM 2015) and FDA DSHEA reporting guidance. The true population-level event rate is materially higher than CAERS shows.
- Causation not established. A CAERS record reports a temporal association between supplement exposure and an adverse event. It does not establish causation. Confounders include concurrent prescription drug use, undisclosed underlying conditions, and product adulteration.
- Brand-versus-ingredient attribution. Many CAERS records identify the product type (for example, "weight-loss supplement") without a brand. Ingredient-level classification is more reliable than brand-level attribution. We avoid naming current-formulation brands except where they are also named in primary FDA, FTC, or NIH literature.
- Voluntary reporting bias. Reports are filed disproportionately by healthcare providers, hospitals, and consumers experiencing serious events. Mild events are systematically under-reported. The 40.1% serious-outcome rate likely overstates the true serious-event share among all supplement users.
- Mixed-category products. A single product can fall into multiple categories (for example, a weight-loss-and-sports thermogenic). Category counts therefore sum to more than the unique-record total of 78,412.
- CAERS revisions. The FDA revises CAERS quarterly. Counts here reflect the data extract as of 2025-Q4 and may shift on subsequent extracts.
- This study is not a clinical safety assessment. Mechanistic and population-level safety evaluations should rely on NIH ODS fact sheets, Cochrane systematic reviews, and primary literature on PubMed.
Related Health Britannica work
For descriptive industry and usage statistics underlying this regulatory data study, see 85 Supplement and Functional Medicine Statistics for 2026, our companion stats reference for the dietary-supplement category. For ingredient-level evidence summaries, our forthcoming Supplement Evidence Index cross-references Cochrane, NIH ODS, and structured marketing-claim data across 80 common supplements.
Republish this chart
Press kit and downloads
Full press kit at /press/. Machine-readable aggregates at data.json under CC-BY 4.0. Media inquiries via the contact on the press page. Journalists writing about specific ingredients or brands are encouraged to also pull from the FDA Tainted Products list, NIH LiverTox, and the FDA warning-letter archive.