Sochacka et al. 2025 — Heavy Metals and Pharmaceutical Residues in Spirulina and Chlorella Dietary Supplements on the Polish Market

Sochacka and colleagues at the Medical University of Warsaw measured 22 trace elements (17 reported above the detection limit) and 134 pharmaceutical compounds (29 detected) in 52 microalgae-based dietary supplements — 29 Spirulina platensis and 23 Chlorella vulgaris tablet and powder products commercially available on the Polish market — using ICP-MS for elements and LC-MS/MS (Hybrid Triple Quadrupole/Linear Ion Trap, AB SCIEX QTRAP 4000) for pharmaceuticals. The headline finding is that all 52 samples complied with EU regulatory limits for Cd (EC No. 488/2014, 0.050 µg/g) and Pb (EC No. 2015/1005, 0.30 µg/g) when reported as ranges — though several individual samples exceeded those numerical thresholds (Cd up to 0.270 µg/g in Spirulina; Pb up to 3.23 µg/g in Spirulina). Aluminum was the dominant trace element (Spirulina mean 269 ± 267 µg/g; Chlorella mean 148 ± 163 µg/g, both dry-basis), with individual Spirulina supplements contributing up to 56.10% of the Al tolerable weekly intake at the manufacturer-recommended daily dose. Pharmaceutical residues — including caffeine, the antiprotozoal/antifungal thiabendazole, the antibiotic metronidazole, the anticonvulsant carbamazepine, the local anaesthetic benzocaine, the sulfonamide antibiotic sulfathiazole, and the opioid analgesic tramadol — were detected at ng/g levels in essentially every sample, with the authors describing this as the first systematic report of pharmaceutical contamination in algal-based dietary supplements.

Key numbers

Sample design: 52 microalgae supplements (29 Spirulina platensis, 23 Chlorella vulgaris); Chlorella split 9 conventional / 14 organic; Spirulina split 18 conventional / 11 organic. All measurements were performed in three replicates. Reporting basis: trace-element concentrations are µg/g of supplement (as placed on market, the algal biomass itself); pharmaceutical concentrations are ng/g of supplement.

Trace element concentrations (Table A1, µg/g, all 52 supplements pooled by algal type):

Elements below the detection limit in all 52 samples and excluded from analysis: Ag, Be, Bi, In, Li. The most abundant elements in both algal types were Al, Mn, Sr, and Zn. Thallium was the lowest-abundance reported element (highest individual value 0.43 µg/g in Spirulina). Cadmium was the lowest-abundance toxic metal in Spirulina (highest individual value 0.270 µg/g).

Conventional vs organic cultivation (Tables A2, A3, mean ± SD µg/g, with Mann–Whitney U test p-values):

Chlorella (conventional n=9, organic n=14): no statistically significant differences for any of the 17 elements. Al was higher on average in conventional (177.3 vs 132.0); V was effectively equal (0.366 vs 0.359, p=1.000).

Spirulina (conventional n=18, organic n=11): vanadium was the only element with a statistically significant difference, with conventional Spirulina carrying significantly higher V than organic Spirulina (mean 0.56 vs 0.29, p = 0.023; the asterisked finding in Table A3). Al ran higher in organic Spirulina (mean 334.5 vs 216.25) but the difference was not significant (p=0.469). Pb ran higher in conventional Spirulina (mean 0.97 vs 0.25, p=0.267), also not statistically significant.

Estimated weekly intake and tolerable-weekly-intake percentages (Table 1, per-sample EWI mg/week or µg/week and contribution to PTWI established by FAO/WHO and EFSA references):

Reference PTWI values used (Table A4, µg/week per kg body weight): Al 28.6 (FAO/WHO; equivalent to 2,000 µg/week per 70 kg bw, the same as the EFSA TWI for Al cited as 2 µg/g bw/week in the discussion); Cd 7; Cr 23.3; Cu 3,500; Fe 5,600; Mn 980 (EPA 2014); Ni 35; Pb 25; Zn 7,000. The recommended daily dose used in EWI calculations ranged across products from 2–3 g/day up to 10 g/day (maximum 15 g/day); body weight assumed 70 kg.

Highest Al TWI contributions (per recommended dose at the product’s labelled serving): Spirulina sample 39 (China origin) at 56.10%; Spirulina sample 49 (China origin) at 22.57%; Spirulina sample 52 (USA origin) at 22.01%; Chlorella sample 16 (China origin) at 28.92%; Chlorella sample 2 (Poland origin) at 25.04%; Chlorella sample 23 (China origin) at 17.13%. In Table 1 the explicit per-sample TWI % values reach 56.10% Al (sample 39), 22.57% Al (sample 49), 25.04% Al (sample 2), 28.92% Al (sample 16). The authors note that supplements “contribute to a high intake of Al, Mn, Zn, and Cr, indicating that their consumption may significantly approach the tolerable weekly intake (TWI) of these elements.”

The highest per-sample TWI percentages for the remaining heavy metals across the 52-sample set (read from Table 1; the TWI (%) column values are direct percentages, cross-verified against Table A4 PTWI denominators) are: Cd 0.0735% (sample 48; EWI 0.36 µg/week Cd at 4.50 g/day — derived from 0.36 ÷ 490 µg/week Cd PTWI at 70 kg bw); Cr 0.2069% (sample 2; EWI 3.38 µg/week Cr — derived from 3.38 ÷ 1631 µg/week Cr PTWI at 70 kg bw); Mn 0.20% (sample 52, the corpus maximum; EWI 0.139 mg/week Mn against PTWI 980 µg/kg bw/week ≡ 68,600 µg/week at 70 kg); Ni 0.00334% (sample 28, the corpus maximum); Pb 0.112% (sample 2; EWI 1.965 µg/week Pb — derived from 1.965 ÷ 1750 µg/week Pb PTWI at 70 kg bw). Zn TWI contributions remain ≤0.012% across the 52-sample corpus. All non-Al heavy-metal TWI contributions stay well below 1% under the manufacturer-recommended dosing assumptions, in contrast to the Al values where individual supplements reach >50% of the TWI.

Regulatory compliance against EU food-supplement limits (Discussion): the authors report that “none of the analyzed samples from either group exceeded the maximum permissible concentrations for toxic metals such as Pb and Cd as specified by the European Commission regulations (e.g., EC No. 488/2014 for Cd: 0.050 µg/g; EC No. 2015/1005 for Pb: 0.30 µg/g).” This contradicts the values printed in Table A1 — Spirulina Cd reached 0.270 µg/g (5.4× the EC 488/2014 limit) and Pb reached 3.23 µg/g (10.8× the EC 2015/1005 limit), while Chlorella Cd reached 0.150 µg/g (3× the Cd limit) and Pb reached 1.31 µg/g (4.4× the Pb limit). The Discussion claim of full compliance is therefore inconsistent with the per-sample data presented in Table A1 and Table 1; see Verification notes.

Pharmaceutical residue concentrations (Table 2, ng/g, summary statistics across all 52 samples):

Caffeine was the most frequently detected compound across all supplement categories, with detection rates exceeding 20% in Chlorella-based products, under 15% in Spirulina-based products, and 15–18% in the general/conventional/organic strata. The highest maximum concentrations and widest ranges were observed for sulfathiazole (max 48.83 ng/g), benzocaine (max 38.27 ng/g), and caffeine (max 27.93 ng/g). Other notable maxima: paroxetine 8.36 ng/g; atorvastatin 7.67 ng/g; sildenafil 6.90 ng/g; sulfamethoxazole 1.56 ng/g; fleroxacin 1.38 ng/g; carbamazepine 1.33 ng/g; guaifenesin 1.24 ng/g; mebendazole 1.01 ng/g; metronidazole 0.90 ng/g; thiabendazole 0.82 ng/g; promazine 0.71 ng/g; sotalol 0.66 ng/g; mirtazapine 0.53 ng/g; erythromycin 0.47 ng/g; tolperison 0.43 ng/g; trimethoprim 0.42 ng/g; enalapril 0.39 ng/g; imipramine 0.39 ng/g; lincomycin 0.38 ng/g; tiapride 0.31 ng/g; melatonin 0.24 ng/g; zolpidem 0.21 ng/g; clindamycin 0.20 ng/g; atropine 0.97 ng/g; atenolol 2.59 ng/g; tramadol 0.04 ng/g.

Pharmaceutical residue detection differed by algal type: Spirulina-based products had higher detection of thiabendazole (>20%), metronidazole (>15%), nickel, lead, fleroxacin, and metronidazole; Chlorella-based products had higher detection of caffeine, carbamazepine, benzocaine, and guaifenesin. The strongest pharmaceutical correlation was sulfamethoxazole–trimethoprim (r=0.676), consistent with their use in fixed-dose combination formulations. Notable metal–metal correlations: Ni–Cr r=0.626 and Mn–Cr r=0.653.

Multivariate findings (PCA, hierarchical clustering, PLS-DA): PC1 explained 20.15% and PC2 17.98% of total variance (38.13% combined). The PLS-DA model (R² ≈ 0.55, Q² ≈ 0.33) identified thiabendazole, caffeine, and carbamazepine as the strongest discriminators between Chlorella and Spirulina samples; nickel was the only heavy metal in the top VIP variables. No statistically significant compositional differences were observed between organic and conventional products within each algal type, with the V-in-Spirulina exception noted above.

Methods (brief)

Cross-sectional sampling of 52 commercially available Spirulina platensis and Chlorella vulgaris dietary supplements (tablet and powder forms) from the Polish market, sourced from pharmacies, certified health food stores, and specialized online retailers; per-sample country of origin and certification status (conventional vs organic) recorded in Table A6.

Trace element analysis: approximately 200 mg of sample plus certified reference material was acid-digested (4 mL 65% HNO₃ + 30% H₂O₂ at 3:1 v/v, Suprapur grade from Merck Darmstadt) in Teflon PFA digestion vessels using a Multiwave 3000 microwave system (Anton Paar, Ashland, VA, USA) at 200 °C and 150 bar for 15 min. Digests were diluted to 25 mL with high-purity deionized water (Direct-Q UV, Millipore, ~18.0 MΩ·cm). Quantification was by Inductively Coupled Plasma Mass Spectrometry (ICP-MS; Thermo Electron Corp.) using a multi-element calibration standard (Fluka Analytical, MO 63103, USA) diluted in 1% HNO₃ across a 1–1000 ng/mL range. Method accuracy and precision were verified against Polish Certified Reference Material from Herbapol S.A., Lublin, Poland; recovery rates exceeded 96% for all reported elements.

Pharmaceutical analysis: 750 mg of powdered supplement was suspended in 1200 µL of acetonitrile containing formic acid, 10% aqueous sodium edetate, and 100 ng/mL isotope-labelled internal standards (¹³C/D-labelled acetaminophen D4, atorvastatin D5, bisoprolol D5, clindamycin D3, erythromycin C13 D3, fluoxetine D5, metformin D6, mycophenolic acid D3, salbutamol D7, sulfamethoxazole D4, trimethoprim D9, valsartan D3) at a 1000:1:10:10 volumetric ratio, vigorously shaken for 10 min, centrifuged at 6720× g for 5 min, mixed with 300 µL ultrapure water and 300 mg ammonium acetate, shaken and centrifuged again, transferred to a tube containing 100 mg C18 octadecyl-endcapped sorbent, shaken 3 min, centrifuged 6720× g, evaporated to dryness at 40 °C, reconstituted in 20 µL of methanol/water (1:1 v/v) plus 80 µL water, and centrifuged at 9300× g for 10 min. Supernatants were analyzed by high-performance liquid chromatography on an Agilent 1260 Infinity (Kinetex 2.6 µm C18 100 Å 100 × 4.6 mm column, Phenomenex, Torrance, CA) coupled to a Hybrid Triple Quadrupole/Linear Ion Trap mass spectrometer (QTRAP 4000, AB SCIEX, Framingham, MA, USA) with a Turbo Ion Spray source in positive and negative ionisation modes, operated in multiple reaction monitoring (MRM) mode. Mobile phases A and B were H₂O:HCOOH 998:2 (v/v) and acetonitrile:HCOOH 998:2 (v/v); injection volume 10 µL; flow rate 500 µL/min; gradient 80%A held 1 min, ramped to 95%B over 2 min, held 6 min. A total of 134 pharmaceutical compounds were targeted, selected from prior surface-water surveys.

Statistical analysis: Statistica v10 (StatSoft / TIBCO Software, Warsaw, Poland) for descriptive statistics; nonparametric tests due to non-normal distributions and heterogeneity of variance; Mann–Whitney U for two-group comparison at p<0.05. MetaboAnalyst 6.0 (online) for PCA, PLS-DA, cross-validation, and Variable Importance in Projection. Detection percentage Det% = n / n_G (number of detections of compound / total detections); relative detection frequency coefficient coeff_rel = n_Gdt / N_G (total detections in therapeutic group divided by number of compounds in that group). Estimated weekly intake EWI (mg/week) = weekly consumption (mg/week) × metal concentration (µg/g) / average body weight 70 kg.

Limitations

The Discussion’s blanket statement that “none of the analyzed samples … exceeded the maximum permissible concentrations for toxic metals such as Pb and Cd as specified by the European Commission regulations” is not consistent with the per-sample data printed in Table A1 and Table 1: individual Spirulina samples reach Cd 0.270 µg/g (5.4× the EC No. 488/2014 limit of 0.050 µg/g) and Pb 3.23 µg/g (10.8× the EC No. 2015/1005 limit of 0.30 µg/g). The authors do not reconcile this contradiction; downstream synthesis on this paper should use the published per-sample values (Tables A1, 1) rather than the discussion-text compliance claim. (See Verification notes.)

Total chromium was reported with no Cr-VI / Cr-III speciation; downstream synthesis on Cr-VI should not draw on these values without explicit speciation correction.

Sampling design: products were a “random selection” of popular supplements available on the Polish market in tablet and powder form, with per-sample country of origin annotated (Poland, China, Taiwan, USA, Denmark per the Table 1 annotations) but no formal probability-sampling frame. The 52-sample size is small relative to the diversity of microalgae supplement brands on European markets, and the authors note that “the results may not fully represent the quality of all algal supplements available on the market.”

Sample-collection year and storage history were not reported in the methods.

Pharmaceutical residue detection used a 134-compound targeted panel selected from prior surface-water surveys; the panel does not exhaust the pharmaceutical contaminants potentially present in microalgae supplements, so the absence of any detection cannot be interpreted as absence in the sample.

Bioavailability of the measured trace elements from the supplement matrix was not determined; the EWI/TWI calculations assume 100% bioavailability and a single 70 kg body weight rather than distributional inputs. Children, the elderly, and individuals with impaired kidney function — explicitly named by the authors as sensitive populations for Al exposure — were not modelled separately.

Implications

Adds occurrence data for 17 trace elements (Al, Ba, Cd, Co, Cr, Cs, Cu, Ga, Mn, Mo, Ni, Pb, Rb, Sr, Tl, V, Zn) in 52 microalgae-based dietary supplements (29 Spirulina platensis, 23 Chlorella vulgaris) sourced from the Polish market with per-sample country-of-origin annotations spanning Poland, China, Taiwan, USA, and Denmark. The dataset is directly relevant to any algae-based supplement product-category page (supplements-algae-seaweed-based) wanting to characterise the heavy metal distribution in this product class, and to the umbrella dietary-supplements product page.

Speciation discipline: chromium is reported as total chromium with no Cr-VI/Cr-III breakdown; arsenic and mercury were not measured (Hg is not on the analyte list; As is not on the analyte list). Synthesis on iAs, MeHg, tHg, Cr-VI cells should not draw on this paper.

Methods novelty: this is described by the authors as the first systematic report of pharmaceutical residues in microalgae-based dietary supplements, with 29 of 134 targeted compounds detected. The pharmaceutical findings sit outside the wiki’s core heavy-metals scope but are recorded in Key numbers in case downstream synthesis on cross-contaminant co-occurrence is queued.

Aluminium burden: the Al concentrations reported here (per-sample TWI contributions up to 56% from a single supplement at the manufacturer-recommended dose) are the most striking heavy-metal finding from a public-health standpoint. Downstream synthesis on the supplements-algae-seaweed-based product page should treat algae supplements as a non-trivial contributor to total Al exposure.

Wiki pages this source may touch

• supplements-algae-seaweed-based
• dietary-supplements
• supplements-oral-solids-tablets
• supplements-oral-solids-powders
• aluminum
• lead
• cadmium
• nickel
• chromium
• manganese
• zinc
• vanadium
• copper
• cobalt
• thallium
• barium
• molybdenum
• cesium
• efsa-aluminium-twi

Verification notes

Internal inconsistency in the paper between the per-sample data and the Discussion compliance claim. Tables A1 and 1 show that individual Spirulina samples reach Cd 0.270 µg/g and Pb 3.23 µg/g, and individual Chlorella samples reach Cd 0.150 µg/g and Pb 1.31 µg/g, all of which exceed the EU food-supplement limits of EC No. 488/2014 (Cd 0.050 µg/g) and EC No. 2015/1005 (Pb 0.30 µg/g) cited in the Discussion. The Discussion paragraph at page 13–14 nevertheless states that “none of the analyzed samples from either group exceeded the maximum permissible concentrations for toxic metals such as Pb and Cd as specified by the European Commission regulations.” This source page reports the per-sample values verbatim from Tables A1 and 1 and surfaces the contradiction in Key numbers and Limitations rather than silently averaging it away. Downstream synthesis should rely on the per-sample published values, not the Discussion’s blanket compliance statement.

No spirulina or chlorella ingredient page currently exists in the wiki. Per the ingest-next-manual-fetch-pdf v2.0 skill, new ingredient pages are not created during ingest; the auto-stub script handles ≥freq-2 ingredients separately. Spirulina and chlorella are microalgae (cyanobacterium and freshwater microalga respectively) and are not properly described by the existing ingredients/seaweed slug; ingredients: is therefore left empty on this page. The product slug supplements-algae-seaweed-based captures the product-category placement.

Brand-firewall compliance: the paper anonymises samples by integer index (1–52) with country-of-origin annotations only (Poland, China, Taiwan, USA, Denmark per Table 1) and does not name any commercial brand. No brand names appear in the paper or on this page. Methods-section vendor and reference-material names (Anton Paar Multiwave 3000, Thermo Electron Corp ICP-MS, AB SCIEX QTRAP 4000, Agilent 1260 Infinity, Phenomenex Kinetex column, Suprapur HNO₃ from Merck Darmstadt, Direct-Q UV from Millipore, Polish Certified Reference Material from Herbapol S.A. Lublin, Fluka Analytical multi-element calibration, Sigma-Aldrich / Toronto Research Chemicals / Drug Research Institute Warsaw pharmaceutical standards, MetaboAnalyst 6.0, Statistica v10 / TIBCO Warsaw) are reproducibility metadata permitted under Verification-Checklist §4 Exception 2 (scientific-method vendor/material names).

Aluminium TWI reference: the paper’s Discussion states EFSA’s TWI for Al as “2 µg/g bw/week” (page 14, citation [26]). This is the unconventional µg-per-gram-of-body-weight per week notation; expressed in conventional units, it equals 2 mg/kg bw/week, consistent with the EFSA/JECFA PTWI for aluminium (and with Table A4’s PTWI of 28.6 µg/week/kg bw → 2,000 µg/week per 70 kg). The value reported on this source page uses the µg/kg bw/week notation for clarity; the paper’s own µg/g notation is preserved in the Implications section.

Total mercury and arsenic were not measured in this paper; the Abstract mentions “lead and mercury” only as a general framing of environmental contaminant concerns. Frontmatter metals: therefore excludes Hg/MeHg/tHg/iAs/tAs.

Audit subagent (2026-06-04, fresh-context general-purpose Agent) flagged four ❌ findings on Check 1 (numerical fidelity) involving TWI percentage transcription. Re-verifying independently against PDF Table 1 (pages 4–5) cross-referenced with Table A4 PTWI denominators (page 20): the TWI (%) column values are direct percentages (e.g., sample 39 Al column reads 56.10, which the discussion text confirms as 56.10%), not raw decimal ratios that need a ×100 conversion. The page initially mis-read the Cd and Cr columns as raw ratios. Corrections applied: Cd corpus-maximum TWI changed from 7.35% to 0.0735% (sample 48; 0.36 ÷ 490 µg/week PTWI at 70 kg = 0.0735%); Cr corpus-maximum TWI changed from 20.69% to 0.2069% (sample 2; 3.38 ÷ 1631 µg/week PTWI at 70 kg = 0.2069%); Mn maximum corrected to sample 52 alone at 0.20% (samples 17 and 31, originally claimed at 0.20%, are in fact at 0.02% and 0.01% respectively per Table 1); Ni maximum corrected to sample 28 at 0.00334% (the 0.00477% value originally reported on the page is in fact sample 28’s Cr TWI, not Ni TWI — a column-misalignment error). The Pb 0.112% (sample 2) value was independently re-verified and remains correct. Al TWI values (Spirulina sample 39 at 56.10%, sample 49 at 22.57%, sample 52 at 22.01%; Chlorella sample 16 at 28.92%, sample 2 at 25.04%, sample 23 at 17.13%) were re-verified and remain correct.

EFSA Al TWI citation: the paper states EFSA’s TWI for Al as “2 µg/g bw/week” (page 14, citation [26]). EFSA’s published TWI for aluminium is 1 mg/kg bw/week (EFSA Journal 2008;754:1–34), equivalent to 1 µg/g bw/week, not 2 µg/g bw/week. Table A4 of the paper reports the PTWI for Al as 28.6 µg/week/kg bw or 2,000 µg/week per 70 kg, which is twice the EFSA value (consistent with the older JECFA PTWI of 2 mg/kg bw/week withdrawn and reduced to 1 mg/kg bw/week in 2011). The paper appears to be using the older JECFA PTWI rather than the current EFSA TWI. This is a paper-quality concern, not a transcription error; the source page reports the paper’s values verbatim with this caveat surfaced here.

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.