Safety Profile
Known Safety Concerns
- Class IV (E150d) contains 4-MEI -- California Prop 65 carcinogen listing
- Type of caramel color rarely disclosed on supplement labels
- 4-MEI caused cancer in animal studies at high doses -- human relevance debated
- CSPI has petitioned FDA to ban certain caramel colors
Contraindications
- Class IV (E150d) contains 4-MEI -- California Prop 65 carcinogen listing
- Type of caramel color rarely disclosed on supplement labels
Interactions
Information not yet available for this ingredient profile.
Evidence and Scientific Findings
Ingredient Overview
Caramel color is the most widely consumed food coloring globally. Class IV caramel (E150d — ammonia-sulfite process) contains 4-methylimidazole (4-MEI), which the state of California lists as a probable carcinogen (Prop 65). This class is commonly used in cola drinks and some supplement syrups. Classes I and II have better safety profiles. The type of caramel color is rarely specified on supplement labels.
Biological and Chemical Classification
- Scientific Name
- Caramel color (E150a/b/c/d)
Mechanism of Action
Information not yet available for this ingredient profile.
Clinical Evidence of Effectiveness
Information not yet available for this ingredient profile.
Pharmacokinetics
Information not yet available for this ingredient profile.
Recommended Dosage
Information not yet available for this ingredient profile.
SETI — Scientific Evidence Transparency Index
Executive Summary — Ingredient Assessment
- 10 studies reviewed
- 0 high-quality studies (meta-analysis or RCT)
- Main clinical benefit observed: Excipient
- Evidence consistency: High consistency across studies (100%)
- Class IV (E150d) contains 4-MEI -- California Prop 65 carcinogen listing
- Type of caramel color rarely disclosed on supplement labels
- 4-MEI caused cancer in animal studies at high doses -- human relevance debated
- CSPI has petitioned FDA to ban certain caramel colors
The available scientific evidence for Caramel Color indicates notable safety signals that warrant caution. Use should be considered carefully and monitored, particularly in sensitive populations or alongside other medications.
Total SETI Score
High risk| Evidence quality | 10/40 |
| Evidence consistency | 20/20 |
| Safety signals | 0/20 |
| Study recency | 10/10 |
| Evidence transparency | 10/10 |
Evidence Summary
- 10 studies reviewed
- 0 high-quality studies (meta-analysis or systematic review)
- 0 studies identified benefits or no safety concern (GREEN)
- 10 studies reported limited or advisory safety evidence (YELLOW)
Evidence Policy
Only peer-reviewed scientific literature indexed in PubMed or comparable databases is included in this evaluation. Commercial websites, blogs, and marketing materials are excluded. All references include direct traceable links to source documents.
Last updated: 24 მარ 2026, 09:34
Evidence Distribution
-
Observational / other LOW evidence YELLOWDual-functional carbon dots for fluorescent detection of L-Glutamic acid and 2-Methylimidazole in complex matrices: Applications in food safety and environmental monitoring. ↗Ju Y et al.. Dual-functional carbon dots for fluorescent detection of L-Glutamic acid and 2-Methylimidazole in complex matrices: Applications in food safety and environmental monitoring.. Anal Chim Acta. 2025. PMID:41167902.PMID 41167902 ↗Journal Anal Chim ActaYear 2025Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/41167902/
-
Observational / other LOW evidence YELLOWReproductive and developmental toxicity risk assessment for 4-methylimidazole. ↗Scialli AR. Reproductive and developmental toxicity risk assessment for 4-methylimidazole.. Regul Toxicol Pharmacol. 2025. PMID:40412543.PMID 40412543 ↗Journal Regul Toxicol PharmacolYear 2025Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/40412543/
-
Observational / other LOW evidence YELLOWUnraveling the Physicochemical, Nutritional and Antioxidant Properties of the Honey Produced from the Fallopia japonica Plant. ↗Cucu AA et al.. Unraveling the Physicochemical, Nutritional and Antioxidant Properties of the Honey Produced from the Fallopia japonica Plant.. Foods. 2024. PMID:38998468.PMID 38998468 ↗Journal FoodsYear 2024Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/38998468/
-
Observational / other LOW evidence YELLOWImmunotoxicity of 2-Acetyl-4-tetrahydroxybutylimidazole in BALB/c mice with different vitamin B6 nutritional statuses. ↗Yuan L et al.. Immunotoxicity of 2-Acetyl-4-tetrahydroxybutylimidazole in BALB/c mice with different vitamin B6 nutritional statuses.. Toxicology. 2024. PMID:38768702.PMID 38768702 ↗Journal ToxicologyYear 2024Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/38768702/
-
Observational / other LOW evidence YELLOWLevel Optimization of Beet Powder and Caramel Color for Beef Color Simulation in Meat Analogs before and after Cooking. ↗Lee S et al.. Level Optimization of Beet Powder and Caramel Color for Beef Color Simulation in Meat Analogs before and after Cooking.. Food Sci Anim Resour. 2023. PMID:37701740.PMID 37701740 ↗Journal Food Sci Anim ResourYear 2023Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/37701740/
-
Observational / other LOW evidence YELLOWEffects of 2-acetyl-4-tetrahydroxybutylimidazole exposure during gestation and lactation on maternity and offspring immune function in Balb/c mice. ↗Zhao Z et al.. Effects of 2-acetyl-4-tetrahydroxybutylimidazole exposure during gestation and lactation on maternity and offspring immune function in Balb/c mice.. Toxicology. 2023. PMID:37531992.PMID 37531992 ↗Journal ToxicologyYear 2023Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/37531992/
-
Observational / other LOW evidence YELLOW4-Methylimidazole, a carcinogenic component in food, amount, methods used for measurement; a systematic review. ↗Akbari N et al.. 4-Methylimidazole, a carcinogenic component in food, amount, methods used for measurement; a systematic review.. Food Chem X. 2023. PMID:37397204.PMID 37397204 ↗Journal Food Chem XYear 2023Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/37397204/
-
Observational / other LOW evidence YELLOWMechanisms of Rebaudioside A Degradation and Ingredient-Sweetener Interactions in Beverages during Storage. ↗Gelinas BS et al.. Mechanisms of Rebaudioside A Degradation and Ingredient-Sweetener Interactions in Beverages during Storage.. Molecules. 2022. PMID:35209174.PMID 35209174 ↗Journal MoleculesYear 2022Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/35209174/
-
Observational / other LOW evidence YELLOWProcessing of coconut sap into sugar syrup using rotary evaporation, microwave, and open-heat evaporation techniques. ↗Asghar MT et al.. Processing of coconut sap into sugar syrup using rotary evaporation, microwave, and open-heat evaporation techniques.. J Sci Food Agric. 2020. PMID:32337729.PMID 32337729 ↗Journal J Sci Food AgricYear 2020Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/32337729/
-
Observational / other LOW evidence YELLOWMultigenerational reproductive assessment of 4-methylimidazole administered in the diet to Hsd:Sprague Dawley SD rats. ↗Behl M et al.. Multigenerational reproductive assessment of 4-methylimidazole administered in the diet to Hsd:Sprague Dawley SD rats.. Reprod Toxicol. 2020. PMID:32229253.PMID 32229253 ↗Journal Reprod ToxicolYear 2020Study type Observational / otherEvidence strength LOW evidencePubMed link https://pubmed.ncbi.nlm.nih.gov/32229253/
Score Transparency
0 of 10 approved references (score saturates at 10). More peer-reviewed studies = stronger evidence base.
Method: Q = number of approved references ÷ 10 (capped at 1.0)
Limited — mostly case reports or animal studies
Method: L = mean study-level weight across approved references. Level 1 (meta-analysis / systematic review) = 1.0; Level 2 (RCT) = 0.8; Level 3 (cohort/case-control) = 0.6; Level 4 (case report) = 0.4; Level 5 (animal / in-vitro) = 0.2.
Mixed or neutral — roughly equal benefit and risk signals
Method: D = (sum of risk-scored references − sum of benefit-scored references) ÷ total evidence score, then scaled from [−1, 1] to [0, 1]. 0.0 = pure benefit; 0.5 = neutral; 1.0 = pure risk.
One or more monitoring-level safety signals active
Method: S = 0.5 (neutral baseline) + sum of active signal severity deltas ÷ 10. Severity deltas: Critical = +2.0, High = +1.5, Moderate = +1.0, Low = +0.5. Capped at 1.0.
Final GIRI Score for Caramel Color. Risk level thresholds: Low 0–3.0 · Moderate 3.0–5.5 · High 5.5–7.5 · Critical 7.5–10.
Full methodology & data sources
The GIRI Score is computed entirely from structured data — no editorial scoring or subjective weighting is applied at any step.
- References: Only approved references are counted. Each reference is assigned an evidence level (L1–L5) and a direction (risk / neutral / benefit) by the reference manager or AI classifier.
- Safety Signals: Sourced from regulatory agencies (FDA, EMA, Health Canada, TGA, and others) and pharmacovigilance databases. Only active signals count toward the score.
- Formula version: GIRI Score v3.7.0 — Q × L × D × S × 10.
- Limitations: The score reflects published evidence and recorded signals as of the last update date. It is not a clinical risk assessment and should not replace advice from a qualified healthcare professional.
Risk Level Classification
Based on available regulatory signals and scientific evidence, this ingredient presents a low safety concern under normal conditions of use.
0–3.0
3.0–5.5
5.5–7.5
7.5–10
The score pin shows exactly where this ingredient falls on the fixed risk scale.
What drove the Low classification for Caramel Color
A score of 3.5 places this ingredient in the Low band. Thresholds: Low 0–3.0 · Moderate 3.0–5.5 · High 5.5–7.5 · Critical 7.5–10.
0 approved references.
Limited — mostly case reports or animal studies (Level 4–5).
Neutral or mixed — benefit and risk signals roughly balanced.
No active signals — S component is at neutral baseline (0.5), contributing no extra risk weight.
No major regulatory restrictions or advisories recorded across monitored jurisdictions (FDA, EMA, Health Canada, TGA, and others).
How are the Low / Moderate / High / Critical thresholds defined?
The four risk levels are fixed score bands. A score is assigned to exactly one level based on where it falls:
| Level | Score | Meaning |
|---|---|---|
| LOW | 0.0 – 2.9 | Sparse or predominantly beneficial evidence. No active safety alerts. |
| MODERATE | 3.0 – 5.4 | Mixed signals — some risk alongside benefit. Caution at high doses or in sensitive groups. |
| HIGH | 5.5 – 7.4 | Multiple studies or regulatory alerts documenting adverse effects. Professional oversight recommended. |
| CRITICAL | 7.5 – 10 | Regulatory restrictions in one or more major jurisdictions. Serious documented harm. Avoid without specialist supervision. |
Thresholds are fixed constants (GIRI_Score_Utils::LEVEL_THRESHOLDS). They do not change per ingredient and are never subject to editorial adjustment.