Overview
Kowalczyk et al. carried out an EU-FORA Fellowship project at the National Food Institute, Technical University of Denmark (DTU). The work measured rare earth elements (REEs), antimony, barium, boron, lithium, tellurium, thallium, and vanadium in a selection of dry teas commercially available in Denmark and estimated the dietary exposure of adult Danish tea consumers. For each element except tellurium, a risk assessment was performed by comparing estimated intakes against health-based guidance values. The study was published in the EFSA Journal in January 2022 as part of the EU-FORA Series 4 (EFSA Journal 2022;20(S1):e200410).
Study design
Three tea types were analysed: black tea (Camellia sinensis), green tea (Camellia sinensis), and rooibos (Aspalathus linearis). Samples were sourced from the Danish market. The paper does not state the number of commercial products sampled. Samples were ground and homogenised before analysis. A 0.3 g test portion was subjected to acid digestion with microwave heating. Following digestion, samples were analysed by inductively coupled plasma triple-quadrupole mass spectrometry (TQ ICP-MS, iCAP TQ, Thermo Scientific) using Standard, Kinetic Energy Discrimination, and Oxygen Reaction modes (Methods section 2.2.2, PDF p. 5).
Because teas are consumed as infusions, the transfer rate of each element from dry leaf to aqueous infusion was also measured. Brewing conditions and infusion time used for the transfer rate experiments are not described in the main text; a 3-minute brewing time is noted in the uncertainty section. Transfer-rate-corrected concentrations were used for all exposure calculations.
Exposure was estimated for adult Danish consumers only (age greater than 18 years), defined as individuals consuming at least one cup of tea per week. Average daily consumption was 350 mL/day; the high consumption scenario used the 95th percentile of 1,084 mL/day. An average body weight of 70 kg was adopted for all scenarios. Three scenarios were modeled:
- Scenario 1: average consumption (350 mL/day), average element concentration
- Scenario 2: average consumption (350 mL/day), high concentration (P95)
- Scenario 3: high consumption (1,084 mL/day), high concentration (P95)
Scenario 3 was included to account for brand loyalty, the possibility that high consumers may consistently purchase the same high-contamination product.
Methods (brief)
Dry tea samples were ground, homogenised, and digested by microwave-assisted acid digestion (0.3 g test portion). Element analysis was by TQ ICP-MS (iCAP TQ, Thermo Scientific) with three measurement modes: Standard, KED, and Oxygen Reaction. Quality assurance included the certified reference material BCR-670 for REEs and adherence to European standard methods EN 13805:2014 and EN 15763:2009. Tellurium concentrations were often at or below the limit of quantification and are reported as not determined or very low (Methods section 2.2.2, PDF p. 5). LOD and LOQ values for the individual analytes are not tabulated in the main text or appendices. Spike recovery data beyond the BCR-670 CRM are not reported.
Key numbers
Table A.1 (Appendix A, PDF p. 11) reports estimated exposure in ng/kg body weight per day for each element and each tea type, by scenario. The analyte columns in Table A.1 are: sum of REEs, boron (B), barium (Ba), lithium (Li), antimony (Sb), thallium (Tl), and vanadium (V). Tellurium is not included.
| Scenario | REEs (ng/kg bw/day) | B (ng/kg bw/day) | Ba (ng/kg bw/day) | Li (ng/kg bw/day) | Sb (ng/kg bw/day) | Tl (ng/kg bw/day) | V (ng/kg bw/day) |
|---|---|---|---|---|---|---|---|
| 1 (avg/avg) | 9 | 482 | 132 | 2 | 0.4 | 0.2 | 0.1 |
| 2 (avg/P95) | 18 | 578 | 196 | 5 | 0.4 | 0.5 | 0.3 |
| 3 (P95/P95) | 56 | 1,789 | 606 | 16 | 1 | 1.5 | 0.8 |
Green tea:
| Scenario | REEs (ng/kg bw/day) | B (ng/kg bw/day) | Ba (ng/kg bw/day) | Li (ng/kg bw/day) | Sb (ng/kg bw/day) | Tl (ng/kg bw/day) | V (ng/kg bw/day) |
|---|---|---|---|---|---|---|---|
| 1 (avg/avg) | 8 | 636 | 109 | 2 | 0.3 | 0.03 | 0.5 |
| 2 (avg/P95) | 9 | 714 | 133 | 5 | 0.5 | 0.6 | 0.8 |
| 3 (P95/P95) | 27 | 2,211 | 412 | 14 | 2 | 0.2 | 2.5 |
Rooibos tea:
| Scenario | REEs (ng/kg bw/day) | B (ng/kg bw/day) | Ba (ng/kg bw/day) | Li (ng/kg bw/day) | Sb (ng/kg bw/day) | Tl (ng/kg bw/day) | V (ng/kg bw/day) |
|---|---|---|---|---|---|---|---|
| 1 (avg/avg) | 15 | 991 | 44 | 43 | 0.05 | 0.01 | 0.6 |
| 2 (avg/P95) | 19 | 1,345 | 64 | 53 | 0.07 | 0.02 | 0.8 |
| 3 (P95/P95) | 58 | 4,167 | 199 | 163 | 0.2 | 0.04 | 2.6 |
Rooibos carried the highest boron exposure (up to 4,167 ng/kg bw/day in Scenario 3) and the highest lithium exposure (up to 163 ng/kg bw/day in Scenario 3). Black tea carried the highest thallium exposure in Scenario 3 (1.5 ng/kg bw/day). Antimony exposure was low across all tea types and scenarios, ranging from 0.05 to 2 ng/kg bw/day.
Regulatory context
The risk characterization compared estimated intakes against health-based guidance values, as reported in section 3.1 (PDF pp. 5-7):
Antimony: The WHO (2003) TDI for antimony is 6 µg/kg body weight per day, derived from a NOAEL of 6.0 mg/kg body weight per day in a subchronic drinking-water study in rats (endpoint: decreased body weight gain and reduced food and water intake), with an uncertainty factor of 1,000 (100 for intra- and interspecies variation, 10 for use of a subchronic study). The highest antimony exposure from tea in this study — 1 ng/kg bw/day for black tea Scenario 3 — represents 0.03% of that TDI. The authors characterize the antimony risk from tea consumption as negligible.
Barium: The Scientific Committee on Health and Environmental Risks (SCHER, 2012) derived a TDI of 0.2 mg/kg body weight per day based on a BMDL05 for nephropathy, with an assessment factor of 300. The highest barium exposure from tea (black tea, Scenario 3: 606 ng/kg bw/day) covers 0.3% of this TDI.
Boron: EFSA (2004) established a tolerable upper intake level (UL) of 0.16 mg/kg body weight per day based on a NOAEL for decreased fetal body weight in rats, with an uncertainty factor of 60. The highest boron exposure from tea (rooibos, Scenario 3: 4,167 ng/kg bw/day) covers 2.6% of this UL.
Lithium: The US EPA (2008) provisional reference dose (p-RfD) for lithium is 2 µg/kg per day, derived from a LOAEL of 2.1 mg/kg per day for adverse organ effects with an assessment factor of 1,000. The highest lithium exposure from tea (rooibos, Scenario 3: 163 ng/kg bw/day) covers 8.14% of this p-RfD. This is the highest percentage TDI coverage reported for any element in this study.
Thallium: The US EPA (2009, 2012) provisional reference dose for thallium is 1 x 10^-5 mg/kg per day (10 ng/kg per day), based on hair follicle atrophy. The highest thallium exposure from tea (black tea, Scenario 3: 1.5 ng/kg bw/day) covers 14.8% of this p-RfD.
Vanadium: The US EPA (2006) reference dose for vanadium is 7 µg/kg body weight per day, based on gastrointestinal disturbance in humans. The highest vanadium exposure from tea (2.6 ng/kg bw/day) covers 0.04% of this RfD.
Rare earth elements: No health-based guidance values for REEs have been established by any regulatory authority. The authors adopted the lowest TDI derived for an individual REE, that for lanthanum (51.3 µg/kg bw/day, based on NOAEL for decreased body weight in a 90-day OECD study), as the reference value for the sum of all REEs. The highest sum-REE exposure from tea (rooibos, Scenario 3: 58 ng/kg bw/day) constitutes at most 0.1% of this reference value.
Tellurium: Concentrations in tea samples were very low and often below the LOQ; tellurium was not included in the exposure assessment.
Evidence fitness
This source supports characterization of antimony, barium, boron, lithium, thallium, and vanadium exposure from tea consumption by adult Danish consumers. For antimony specifically, the study directly quantifies tea as a minor contributor: the highest modeled intake under worst-case conditions (high consumption and high contamination, 1 ng/kg bw/day for black tea) corresponds to 0.03% of the WHO TDI. This is the central number the page previously reported and it is confirmed verbatim by Table A.1 and the risk characterization in section 3.1 (PDF pp. 5-7). The source does not report elemental concentrations in tea leaf on a dry-weight or wet-weight basis; all values are expressed as dietary exposure in ng/kg body weight per day, so the source cannot directly populate a contamination profile in µg/kg. Transfer rates from leaf to infusion were measured but the numerical values are not tabulated in the main text or appendices.
Limitations
Dry-weight concentrations not reported: The paper reports dietary exposure estimates in ng/kg bw/day rather than elemental concentrations in tea leaf or infusion in µg/kg or µg/L. This prevents direct comparison to food regulatory limits expressed as maximum levels in food.
Transfer rates not tabulated: The methods describe measurement of transfer rates from dry leaf to infusion, but the numerical transfer rate values are not given in the main text or appendices. Only the final exposure estimates derived from them are shown.
LOD and LOQ not reported: The main text states that tellurium was often below or at the LOQ, implying that LOQ values were defined, but they are not tabulated for any analyte.
No CRM recovery statistics reported: BCR-670 was included as a quality-control reference material for REEs, but percent recovery values are not given.
Sample count not stated: The paper does not state how many commercial tea products were purchased or how many individual samples were analysed per tea type. The P95 concentrations in Scenarios 2 and 3 therefore cannot be traced to a specific sample size or distribution.
Adults only: The exposure assessment covers adults aged over 18 years consuming at least one cup per week. Exposures for children or infants are not assessed.
Danish market: Sampling is limited to commercially available teas in Denmark; results may not generalise to other markets or regions.
Brewing time: The uncertainty section notes that consumers may brew tea for longer than 3 minutes, which was used in the transfer rate study, potentially increasing element leaching and actual intakes.
Related evidence
Update history
No substantive edit history is available in this build. The full commit record is available in git.
