Overview
This systematic review and meta-analysis examined evidence for an association between mercury exposure and thyroid cancer (TC) risk. The authors searched PubMed, Google Scholar, Embase, and Web of Science for studies published from 1995 through 2022 and included 53 unique studies in their broader literature review. A quantitative meta-analysis was conducted on a subset of 10 studies that measured urinary mercury, thyroid tissue mercury, blood mercury, or fingernail mercury alongside TC incidence or diagnosis. The review also includes an Ingenuity Pathway Analysis (IPA) bioinformatics assessment and a comparison of regional mercury emissions data to global TC incidence rates from the 2020 GLOBOCAN database. The paper was published in Environmental Science and Pollution Research (2024) 31:12841-12855 and is open access under CC BY.
Study design
The quantitative analysis followed PRISMA guidelines. Literature searches identified 112 unique studies; after screening, 25 underwent full-text review and 10 were included in the final quantitative analysis. Included studies were cohort studies and case series containing data on urinary or tissue mercury levels alongside TC incidence, spanning 1995-2022. Studies with non-human subjects, without tissue mercury data, focused on non-thyroidal cancers, or focused on non-cancerous thyroid conditions were excluded. Fifteen studies were excluded from quantitative analysis for improper study design or insufficient data.
The broader literature review (45 unique studies) examined environmental mercury exposure (volcanic activity, industrial pollution), occupational exposure (gold mining, dental practice, crematorium work), dietary exposure (seafood consumption, well water), and cosmetic product exposure, alongside biochemical and molecular mechanisms relevant to thyroid carcinogenesis. The quantitative analysis was restricted to studies with numeric mercury measurements and TC outcomes; it did not evaluate dietary mercury intake studies separately unless those studies also reported biomonitoring measurements.
Samples
For the quantitative meta-analysis:
- Urine studies: 5 studies total; 3 compared mean urine mercury between TC patients and non-TC controls (Kim et al. 2022, Liu et al. 2021, Zhang et al. 2019); 2 examined TC risk in mercury-exposed versus non-exposed groups (Correia et al. 2020, Malandrino et al. 2016)
- Thyroid tissue studies: 3 studies (Zaichick 2022, Chung et al. 2016, Zaichick et al. 1995); tissue TC cohort N = 178; normal thyroid cohort N = 257
- Blood level study: 1 (Chung et al. 2016)
- Fingernail study: 1 (Zaichick 2022)
No pooled total sample size across all studies is reported in the paper; study-level sample sizes are shown only in the forest plot figures. The authors note that the overall power of the quantitative analysis is low.
Results
Urinary mercury: TC versus non-TC comparison (3 studies)
Mean urine mercury was higher in TC patients than in matched controls. Meta-analysis using the inverse variance method and Hunter-Schmidt estimator:
| Outcome | Mean difference | 95% CI |
|---|---|---|
| Urine mercury in TC versus controls | +1.86 ug/g creatinine | 0.32 to 3.41 |
TC risk in high mercury exposure groups (2 studies)
One volcanic environment study (Malandrino et al. 2016, Sicily) and one occupational cohort study (Kim et al. 2022, South Korea) compared TC risk in highly exposed versus non-exposed groups. Meta-analysis using the inverse variance method and DerSimonian-Laird estimator:
| Outcome | Relative risk | 95% CI |
|---|---|---|
| TC risk in exposed versus unexposed | 1.90 | 1.76 to 2.06 |
The volcanic area study reported TC incidence of 18.5 per 100,000 residents (95% CI 16.2-20.8) in the exposed group versus 9.6 per 100,000 (95% CI 8.4-10.8) in matched controls. The South Korean industrial cohort (Kim et al. 2022, nearly 3 million individuals) reported HR = 2.11 (95% CI 1.11-4.03) in the highest versus lowest tertile of urine mercury.
Thyroid tissue mercury: cancer versus normal (3 studies)
Mercury concentrations in thyroid tissue were higher in cancerous tissue than in normal thyroid tissue across three studies. Meta-analysis using the inverse variance method and DerSimonian-Laird estimator:
| Tissue type | Mean mercury (ppm) | 95% CI | N |
|---|---|---|---|
| Cancer thyroid tissue | 0.14 | 0.06 to 0.22 | 178 |
| Normal thyroid tissue | 0.08 | 0.04 to 0.11 | 257 |
Individual studies that did not find an association (dietary mercury)
Zidane et al. 2019 (French Polynesia, n = 229 TC patients, n = 373 controls): dietary mercury and TC risk OR = 0.97 (95% CI 0.41-2.29), not statistically significant.
Chung et al. 2016 (South Korea): did not find an association between mercury levels and TC.
Regional emissions and TC incidence comparison
The paper compared UNEP 2013 Global Mercury Modelling Report data to 2020 GLOBOCAN TC incidence rates across 11 regions. Selected data from Table 2 of the paper:
| Region | Stationary combustion (tons/yr) | Industrial sources (tons/yr) | TC incidence per 100,000 | TC mortality per 100,000 |
|---|---|---|---|---|
| North America | 32.4 | 34.8 | 12.4 | 0.30 |
| East Asia | 56.9 | 71.6 | 11.5 | 0.40 |
| South America | 11.3 | 34.1 | 9.6 | 0.50 |
| Europe | 22.3 | 15.6 | 8.3 | 0.33 |
| Australia | 4.7 | 11.1 | 10.5 | 0.27 |
| Middle East | 3.7 | 8.3 | 8.6 | 0.83 |
| Africa | 33.7 | 50.0 | 2.0 | 0.62 |
| South Asia | 27.7 | 21.2 | 1.6 | 0.38 |
| Southeast Asia | 7.9 | 14.1 | 4.4 | 0.63 |
| Central America | 5.2 | 13.2 | 7.2 | 0.65 |
| CIS countries | 27.3 | 42.9 | 7.3 | 0.40 |
Incidence rates are age-standardized from GLOBOCAN 2020. The correlation between high mercury emissions and high TC incidence is inconsistent; Africa and South Asia have high emissions but low TC incidence, which the authors attribute to genetic, cultural, diagnostic, and environmental confounding factors.
Instrument and analytical method
This is a meta-analysis, not a primary measurement study. No analytical instrument was used by the authors to measure mercury. The included primary studies used various analytical methods that were not harmonized or described in detail in this paper. The analysis used the inverse variance method with the DerSimonian-Laird estimator (tissue studies) and the Hunter-Schmidt estimator (urine studies). Statistical pooling was performed on untransformed (raw) means.
LOD/LOQ
Not reported. This is a meta-analysis; limits of detection and quantification apply to the primary studies, which are not described in detail in this paper.
CRM recovery
Not reported. Not applicable to a meta-analysis.
Mechanisms reviewed
The paper reviewed and summarized several molecular pathways through which mercury may promote thyroid carcinogenesis, based on in vitro and cell-line studies. The IPA bioinformatics analysis identified mercury interactions with TRPC4, TRPC5, ESR1, PGR, PLAT, MTR, MTA1, and SLC12A2, all of which have somatic mutations documented in TC in the COSMIC database. Summary from Table 1 of the paper:
| Pathway | Mercury mechanism | TC relevance |
|---|---|---|
| RAS/MAPK/PI3K | Mercury activates RAS, leading to MAPK and PI3K activation | H-RAS, N-RAS, K-RAS activating mutations in TC |
| MAPK/ERK | Mercury activates MAPK | RET/PTC fusion, RET activating mutation, BRAF mutations in TC |
| PTEN/Akt/CREB | Mercury inhibits PTEN, promoting Akt/CREB activation | PTEN-inhibiting mutations and AKT1-activating mutations in TC |
| Endocrine/HPA axis | Mercury reduces T3/T4 and elevates TSH | High TSH drives thyroid cell proliferation in TC |
| ROS/Nrf2 | Mercury produces ROS, activating Nrf2 signaling | Nrf2 signaling promotes PTC survival; Nrf2 inhibits apoptosis |
| ROS/oxidative stress | Mercury ROS production causes DNA damage | Oxidative stress causes DNA damage, contributing to carcinogenesis |
| DNA repair | Mercury inhibits PARP-1, DNA ligase III-alpha, and XRCC1 | Inability to repair DNA damage promotes carcinogenesis |
Evidence fitness
This paper is EF-4 (context only). It is a systematic review and meta-analysis of human epidemiological studies and in vitro mechanistic evidence. It does not provide primary food concentration measurements. The quantitative findings (mean urine mercury difference of 1.86 ug/g creatinine, relative risk of 1.90 in high-exposure groups, tissue mercury 0.14 ppm in TC) are summary statistics derived from included primary studies, not independently measurable data from this paper itself. The paper can support claims about biological plausibility and epidemiological association between mercury exposure and TC risk, but it cannot establish causation and explicitly states that insufficient evidence precludes definitive conclusions on dietary mercury as a TC risk factor. The in vitro and bioinformatics evidence supplements but does not replace human epidemiological data. The paper is published in Environmental Science and Pollution Research (Springer Nature) under CC BY.
Regulatory context
The paper notes the following regulatory reference values and positions:
- US EPA (2005 Cancer Guidelines): “mercury and methylmercury are not likely to cause cancer in humans”; EPA states no human studies have definitively linked mercury to cancer, though high doses caused increased tumor rates in rodents
- US FDA: maximum 1 ppm mercury in cosmetics; maximum 65 ppm in eye creams only when no alternative preservative is available (as of February 2022)
- US regulatory agencies: methylmercury (MeHg) daily reference values range from 0.1 to 0.3 ug/kg/day
- WHO mercury in drinking water: atmospheric mercury levels typically too low to affect drinking water safety, with exceptions in contaminated well water areas
Limitations
The authors state the following limitations:
- The power of the quantitative analysis is low
- The 10 included studies are heterogeneous in design, population, and exposure measurement
- Two of the nine primary quantitative studies found no association between mercury and TC
- The correlation between regional mercury emissions and regional TC incidence is inconsistent across all 11 regions analyzed (South Asia and Africa are outliers with high emissions but low incidence)
- Most dietary mercury and TC studies are case-control designs, limiting causal inference
- No study specifically isolated dietary mercury exposure from other routes (environmental, occupational)
- The exact contribution of different mercury sources (soil, aquatic, atmospheric) to tissue mercury levels in the included primary studies was not determined
- Potential confounders including genetic, cultural, diagnostic, and socioeconomic factors were not controlled for in the emissions-incidence comparison
- The IPA analysis is bioinformatics-based and predictive, not derived from thyroid tissue data
Implications
Certification: Provides mechanistic context for mercury and thyroid endocrine disruption relevant to MeHg and total mercury (tHg) threshold discussions. The finding that elevated TSH (linked to mercury-disrupted thyroid hormone synthesis) is associated with thyroid cell proliferation is relevant to the clinical justification for mercury limits in dietary categories. Does not establish food concentration thresholds.
Courses: Useful for health-effects modules on mercury carcinogenicity status, the distinction between dietary mercury as a potential risk factor versus established neurotoxicant, and the mechanistic basis linking mercury to endocrine disruption and thyroid pathology.
App: Not directly applicable. This paper contains no food matrix concentration data.
Microbiome: Not addressed in this paper.
Related evidence
Update history
No substantive edit history is available in this build. The full commit record is available in git.