Heavy Metal Index

Heavy Metals Burden in Drinking Water: Global Patterns, Sources, and Public Health Implications

Source access

Publication
Water 18(8): 886
Year
2026
Authors
Olowoyo JO; Olaiya OO; Oharisi OL; Olusola JA
Source type
review
License
CC BY 4.0

Olowoyo et al. 2026 — Heavy metals in drinking water: global patterns across tap, bottled/sachet, surface, and groundwater (narrative review)

This narrative review of 66 studies published 2015–2024 compiles reported heavy-metal concentrations in four drinking-water source types — municipal tap water (Table 1), bottled/sachet water (Table 2), surface water (Table 3), and groundwater (Table 4) — and summarises maxima with WHO standard limits (Table 5). The review reports the highest concentrations of Pb, Cd, As, Cr, Ni, Hg, U, and Mn in groundwater or surface water and the lowest in tap and bottled water, but documents notable exceedances of WHO guidelines in every source type and across all six continents represented. The review is relevant to HMI as both a direct dietary exposure pathway (drinking water as consumed) and an upstream ingredient input (water used to reconstitute infant formula, dilute beverages, prepare processed foods). The review’s principal limitation is that it is a narrative tabulation, not a meta-analysis: per-study ranges are reproduced without weighting, sample sizes are not extracted, and the included papers use mixed units (µg/L and mg/L), mixed analytical methods (AAS, ICP-OES, ICP-MS, GFAAS, SWASV), and mixed quality-assurance regimes.

Key numbers

All ranges below are reproduced from the review’s Tables 1–4 and from the corresponding narrative. Where the review’s narrative and its Table 5 summary disagree on the highest-recorded value, both are stated and the disagreement is flagged (see Verification notes). WHO guideline values are taken from the review’s text and from WHO Guidelines for Drinking-Water Quality (4th edition with 1st and 2nd addenda, 2022; review ref [20]).

Lead (Pb). WHO guideline 10 µg/L (0.01 mg/L).

  • Tap water (Table 1): ND to 260.0 µg/L. Highest reported in Iran (ND–260.0 µg/L, ICP-MS); Saudi Arabia 99.00 ± 6.00 µg/L.
  • Bottled/sachet water (Table 2): ND to 123.15 µg/L. Highest in South Africa (5.98–123.15 µg/L, ICP-MS).
  • Surface water (Table 3): ND to 0.259 mg/L. Highest 0.259 mg/L (259 µg/L) from Ghana (Cobbina et al., 2015; ref [5], AAS, ** mg/L notation). The abstract confirms “ND to 0.259 mg/L (surface water)” as the Pb surface-water range. Bangladesh 8.000–64.00 µg/L is the next-highest by µg/L magnitude.
  • Groundwater (Table 4): ND to 0.791 mg/L. Highest 0.791 mg/L from Ghana (Cobbina et al., 2015; ref [5]).
  • Overall maximum: 0.791 mg/L (groundwater, Ghana). Review narrative and abstract both attribute this maximum to groundwater. The review’s Table 5 summary row lists “Lead — 0.791 mg/L — Ghana — Tap water”; this is a paper-internal contradiction (see Verification notes). Note: the same 0.259 mg/L Ghana figure that appears in Table 3 as a Pb surface-water value also appears in Table 5 as the Hg surface-water overall maximum (Ghana), a second Pb/Hg cross-attribution in the review’s summary table.

Cadmium (Cd). WHO guideline 3 µg/L (0.003 mg/L).

  • Tap water (Table 1): ND to 0.10 mg/L (100 µg/L) and ND in Saudi Arabia. The narrative states “Cadmium in tap water did not show any value of Cd above the recommended limit of 3 µg/L”; the Nigeria entry ND–0.10 mg/L (** mg/L notation) reaches 100 µg/L, contradicting the narrative claim. Flagged as paper-internal inconsistency.
  • Bottled/sachet water (Table 2): 0.055 µg/L to 5.51 µg/L; <LOD–10 µg/L in Saudi Arabia. Highest 5.51 µg/L (South Africa); Olowoyo et al. [4] reported 0.15–5.51 µg/L slightly above WHO.
  • Surface water (Table 3): 0.01–2.23 mg/L. Highest 2.23 mg/L (2,230 µg/L) from Ghana (ref [5]). The review’s narrative names this Ghana surface-water value as the overall Cd maximum.
  • Groundwater (Table 4): 0.005 mg/L to 0.920 µg/L; range as written includes mixed unit notation. South Africa 0.810 ± 0.03 mg/L. The narrative states “from the groundwater sources, the value ranged from 0.005 mg/L to 0.791 mg/L (Table 4)” — note 0.791 is the Pb maximum, suggesting a transposition; the highest Cd groundwater value per Table 4 is 4.6 mg/L (BDL–4.6 mg/L, South Africa, ICP-OES/IC). Flagged as paper-internal inconsistency.
  • Overall maximum per the review’s Table 5 and narrative: 2.23 mg/L surface water, Ghana.

Arsenic (As — total; no inorganic/organic speciation by the included studies). WHO guideline 10 µg/L (0.01 mg/L); US FDA 10 µg/L.

  • Tap water (Table 1): 0.013 to 99 µg/L. Highest 99 µg/L from Iran (AAS).
  • Bottled/sachet water (Table 2): ND to 9.05 µg/L; Croatia 1.44–9.05 µg/L (ACP-AES).
  • Surface water (Table 3): 0.001–92 µg/L. Bangladesh 10–92 µg/L; Bolivia 2.40–34.30 µg/L; China 0.29–33.11 µg/L.
  • Groundwater (Table 4): 0.0001–692 µg/L. Highest 692 µg/L from China (ref [43]). Bangladesh 6–581 µg/L (ref [151]); Canada Alberta <1–179.0 µg/L; Bihar India 6.250–135.0 µg/L; Bangladesh (refs [149], [22]) 0.030–148 µg/L.
  • Overall maximum: 692 µg/L groundwater, China.

Chromium (Cr — total; the review explicitly states “none of the articles used for this review reported on the two most prominent states of Cr separately, but rather on the total Cr”). WHO guideline 50 µg/L; US FDA/USEPA 100 µg/L.

  • Tap water (Table 1): ND to 10.50 µg/L; Uganda 10.50 µg/L (highest in Table 1); Saudi Arabia 1.00–6.00 µg/L.
  • Bottled/sachet water (Table 2): ND to 190.26 µg/L. Highest 190.26 µg/L from South Africa (ref [4]).
  • Surface water (Table 3): ND to 126.0 µg/L (per narrative summary). Highest 126.0 µg/L from Bangladesh (ref [29]).
  • Groundwater (Table 4): “0.090–244.4 **” for China (ref [43]; ** is mg/L per the table key). Table 4 thus prints 244.4 mg/L for China groundwater Cr.
  • Overall maximum: Tables 4 and 5 agree on 244.4 mg/L for the China entry, but disagree on water-source attribution (Table 4 groundwater; Table 5 “Surface water”). The narrative summary on p6 reports “not detectable to 244.43 µg/L in groundwater” — disagreeing with both tables on unit (µg/L vs mg/L). A 244.4 mg/L groundwater Cr concentration would be five orders of magnitude above the WHO 50 µg/L guideline and is implausible without further documentation; a 244.43 µg/L groundwater value is consistent with the narrative summary on p6 and with the per-row Chinese surface-water range (0.09–106.82 µg/L, ref [43]) reported elsewhere in Table 3. Flagged as paper-internal contradiction.

Nickel (Ni). WHO guideline 70 µg/L (0.07 mg/L); EU 20 µg/L for natural mineral water.

  • Tap water (Table 1): 0.02 to 3.5 µg/L for most entries; Nigeria 1.021 ± 0.1 mg/L (1,021 µg/L) — highest entry in Table 1. The review narrative claims the 1.021 ± 0.1 mg/L Nigeria value is from a “surface water” study; Table 1 places it in the tap-water table. Flagged as paper-internal inconsistency.
  • Bottled/sachet water (Table 2): ND to 2.33 µg/L (the review summarises the bottled-water maximum as 2.33 µg/L from Saudi Arabia).
  • Surface water (Table 3): 0.13–516.5 µg/L. Highest 516.5 µg/L from China (ref [43]).
  • Groundwater (Table 4): BDL–516.5 µg/L; China 0.870–516.5 µg/L; Nigeria 0.953 ± 0.07 mg/L (953 µg/L); Bangladesh 13–71 µg/L.
  • Table 5 overall maximum: “Nickel — 516.5 µg/L — China — Surface water”.

Iron (Fe). No WHO health-based guideline; WHO aesthetic limit 300 µg/L (0.3 mg/L).

  • Tap water (Table 1): 0.9 to 1073 µg/L. Mean highest 139 ± 13 µg/L (Nigeria, Table 1 — 139.00 ± 13.00 ** mg/L per the published table; double asterisk in this study is mg/L, but the value as printed in mg/L would be 139 mg/L which is implausible — Table 1 is internally ambiguous on units for the Nigeria Fe entry; the narrative quotes 0.139 ± 0.013 mg/L as the tap-water mean of the highest concentration; flagged as paper-internal unit ambiguity).
  • Bottled/sachet water (Table 2): ND to 1346.91 µg/L. Highest 1346.91 µg/L from South Africa (ref [4]); narrative attributes this to Table 2 (bottled water), not tap water.
  • Surface water (Table 3): 0.01 to 1548 µg/L; Algeria 1065–1248 mg/L (Tebessa, ref [73]); China 4.31–1548 µg/L.
  • Groundwater (Table 4): 0.0001 mg/L to 20,843.2 µg/L. Highest 20,843.2 µg/L from South Africa (ref [72]); Bangladesh 850–10,860 µg/L; India 112.61–15,708.11 µg/L; Tuscany Italy 11–1337 µg/L.
  • Table 5 overall maximum: 20,843.2 µg/L groundwater, South Africa.

Copper (Cu). WHO guideline 2,000 µg/L (2 mg/L).

  • Tap water (Table 1): 0.013 to 62 µg/L; Nigeria 47.00 ± 3.00 mg/L (this is mg/L per the ** notation — 47 mg/L is enormous and far exceeds the WHO 2 mg/L guideline; printed as mg/L in Table 1).
  • Bottled/sachet water (Table 2): ND to 7.530 µg/L; South Africa 21.33–407.32 µg/L (ICP-MS, ref [4]).
  • Surface water (Table 3): 0.00045 mg/L to 119 µg/L; Bangladesh 23–119 µg/L (AAS Flame, ref [29]); Bolivia 0.7–14.90 µg/L; Costa Rica 0.48–52.83 µg/L.
  • Groundwater (Table 4): BDL to 996.6 µg/L. Highest 996.6 µg/L from China (ref [43]); Sri Lanka 0.592–432 µg/L; Tuscany Italy ND–6.00 µg/L.
  • Narrative additionally cites China “up to 3816.50 mg/L” and Pakistan/India “up to 2325 mg/L” from individual papers in the review’s text; these mg/L attributions are not reproduced in Tables 1–4 at the corresponding magnitudes and are likely unit-reporting errors in the original studies (flagged in Verification notes — the wiki reproduces the table values faithfully).
  • Table 5 overall maximum: 996.6 µg/L groundwater, China.

Zinc (Zn). No WHO health-based guideline; aesthetic value 3,000 µg/L.

  • Tap water (Table 1): ND to 369.0 µg/L. Highest 369.0 µg/L from Ecuador (ICP-OES, ref [92]).
  • Bottled/sachet water (Table 2): <LOD to 2,978.31 µg/L. Highest 2,978.31 µg/L from South Africa (ref [4]).
  • Surface water (Table 3): 0.001 mg/L to 894.75 µg/L; China 2.32–894.75 µg/L (ref [43]); Pakistan 9.0–74.0 µg/L.
  • Groundwater (Table 4): 0.0010 mg/L to 19,964.5 µg/L. Highest 19,964.5 µg/L from South Africa (ref [72]); China 2.32–894.75 µg/L; Sri Lanka 8.107–241 µg/L.
  • Table 5 overall maximum: 19,964.5 µg/L groundwater, South Africa.

Uranium (U). WHO guideline 30 µg/L (0.03 mg/L).

  • Tap water (Table 1): ND to 5.590 µg/L; Iran ND–5.590 µg/L, Lebanon 0.227 µg/L. Detected in only two tap-water studies.
  • Bottled/sachet water (Table 2): ND to 1.92 µg/L; Saudi Arabia <LOD–1.92 µg/L; Lebanon ND–0.073 µg/L.
  • Surface water (Table 3): <LOD to 41.00 µg/L; Hungary <LOD–41.00 µg/L; Tebessa Algeria 14.20–18.00 µg/L.
  • Groundwater (Table 4): 0.001 to 579 µg/L. Highest 579 µg/L from SW-Punjab India (Bajwa et al., ref [50], laser fluorimeter); Tuscany Italy 0.210–0.51 µg/L; Mexico 0.001–0.049 µg/L. The narrative reports Bajwa et al. found 338 of 498 samples (~68%) above the WHO permissible limit and a separate study reporting Kabul, Afghanistan groundwater at ~190 µg/L (17.2× the WHO limit).
  • Table 5 overall maximum: 579 µg/L groundwater, India.

Manganese (Mn). WHO guideline 400 µg/L (0.4 mg/L); USEPA health advisory 30 µg/L (secondary contaminant).

  • Tap water (Table 1): 0.234 to 122.0 µg/L. Highest 3.000–122.0 µg/L from Ecuador (ICP-OES, ref [92]); Costa Rica 0.34–86.8 µg/L; Ethiopia 2.080–19.78 µg/L; Lebanon 0.234 µg/L; Saudi Arabia ND.
  • Bottled/sachet water (Table 2): <LOD to 296 µg/L; Croatia 1.13–296 µg/L (ICP-AES, ref [86]); South Africa 16.95–268.91 µg/L.
  • Surface water (Table 3): 0.040 to 1,941 µg/L; China 0.040–1,941 µg/L (ICP-MS, ref [43]); Bolivia 0.040–780.5 µg/L.
  • Groundwater (Table 4): 0.0001 mg/L to 2,339.90 µg/L. Highest 2,339.90 µg/L from India (ref [107]); Sri Lanka 2.477–2,241 µg/L (* µg/L notation per Table 4); Bangladesh 50.0–2,020 µg/L.
  • Table 5 overall maximum: 2,339.9 µg/L groundwater, India.

Mercury (Hg — total; the review notes “though the studies used for this review did not indicate the Hg form”). WHO guideline 6 µg/L per Table 5; 1 µg/L per the narrative text (see Verification notes).

  • Tap water (Table 1): ND to 78.00 µg/L; Iran ND–1.05 µg/L; Saudi Arabia 18.60–78.00 µg/L (ICP-MS, ref [70]).
  • Bottled/sachet water (Table 2): 1.622–13.51 µg/L from Yemen (ICP-MS, ref [94]).
  • Surface water (Table 3): 0.001 mg/L to 0.191 mg/L; Ghana 0.001–0.191 mg/L (ref [5]); Sri Lanka 0.003–0.090 mg/L. Narrative: highest Hg surface-water reading 0.259 mg/L Ghana per Table 5 and narrative (the 0.001–0.191 mg/L range in Table 3 ends at 0.191 mg/L, slightly below the 0.259 mg/L overall summary; reproduced as written).
  • Groundwater (Table 4): 0.0001 mg/L to 0.191 mg/L; Ghana 0.0001–0.191 mg/L; Sri Lanka 0.001–0.06 mg/L; India 0.001–0.010 mg/L. Narrative reports the highest groundwater Hg as 0.191 mg/L.
  • Table 5 overall maximum: 0.259 mg/L (= 259 µg/L) surface water, Ghana. The narrative is in agreement.
  • Note: the narrative says “WHO has set the permissible limit of Hg in drinking water to 1 µg/L”; Table 5 lists the WHO limit as 6 µg/L (0.006 mg/L). The 6 µg/L value in Table 5 corresponds to inorganic mercury in WHO 2017 Guidelines (4th edition, 1st addendum) and the 1 µg/L value to the older guideline. Flagged as paper-internal inconsistency in the cited WHO limit.

Antimony (Sb). WHO guideline 20 µg/L (0.02 mg/L). Not reported in Tables 1 or 4 in this review.

  • Bottled/sachet water (Table 2): 0.410 to 7.23 µg/L; Poland 0.410–2.145 µg/L (HPLC/ICP-DRC-MS); Croatia 0.876–3.25 µg/L; South Africa ND–7.23 µg/L (ref [4]).
  • Surface water (Table 3): ND to 3.082 µg/L; China ND–2.18 µg/L; Turkey 3.082 µg/L (ref [124]).
  • Table 5 overall maximum: 7.23 µg/L bottled water, South Africa.

Selenium (Se). WHO guideline 40 µg/L (0.04 mg/L). Not reported in Tables 1, 2, or 4 in this review.

  • Surface water (Table 3): <LOD to 12.20 mg/L; Tebessa Algeria 7.70–12.20 mg/L (ICP-MS, ref [73]); Turkey 0.283 µg/L; Hungary <LOD–7.00 µg/L.
  • Groundwater (from the narrative, not in Table 4): Bihar India 0.210–0.900 µg/L; Ethiopia 0.015–2.64 mg/L. Both ranges reproduced as written; the mg/L attribution for the Algeria and Ethiopia Se values, if correct, would be three orders of magnitude above the WHO guideline; the underlying studies almost certainly report in µg/L. Flagged.
  • Table 5 overall maximum: 12.20 mg/L surface water, Algeria. The value as printed is implausibly large and likely reflects a unit-transcription error in the review; flagged.

Cross-source synthesis (review-level summaries, not new data):

  • Highest concentrations of Pb, Cd, As, Cr, Ni, Fe, Cu, Zn, U, Mn, Hg generally occur in groundwater or surface water; bottled and tap water generally show the lowest concentrations.
  • Bottled/sachet water shows occasional Pb, Cd, Cr, Mn, and Hg exceedances of WHO guidelines (the review highlights this as “inconsistent regulatory enforcement and possible leaching from packaging materials”).
  • Geographic concentration of high values: Ghana, Bangladesh, China, India, South Africa, Saudi Arabia, Nigeria are repeatedly cited as the highest-burden regions across multiple analytes.
  • Analytical methods used by the included studies: ICP-MS dominant in developed-country studies; AAS, ICP-OES, ICP-AES, ICP-DRC-MS, GFAAS, FAAS, SWASV, and laser fluorimetry in others. The review does not extract LODs/LOQs from the included studies.

Methods (brief)

Narrative review. The authors searched ScienceDirect, Google Scholar, and PubMed using Boolean keyword strings (e.g., “Heavy metals AND drinking water”, “trace metals AND drinking water”, “heavy metals NOT wastewater”). Initial returns: 4,026 articles (ScienceDirect, “heavy metals AND drinking water”); 26,640 (Google Scholar, “heavy metals and drinking water”); 696 (PubMed, “heavy metals and drinking water”). After screening to remove duplicates, off-topic studies (non-drinking-water sources), and articles reporting on heavy-metal removal rather than concentrations, 66 articles published 2015–2024 were retained for tabulation.

Concentrations from the 66 included studies are reproduced in four tables (Table 1 tap water, Table 2 bottled/sachet water, Table 3 surface water, Table 4 groundwater), each listing the analyte, the concentration range (µg/L * or mg/L **), the country/region, the analytical instrument used by the original study, and the reference number. Table 5 summarises the overall highest value, source type, and country per analyte against the WHO standard limit.

No risk-of-bias appraisal, no quality scoring, no weighting by sample size, no meta-analysis, no statistical synthesis. Per-study sample sizes are not extracted into the tables. WHO reference values are taken from the 4th edition Guidelines for Drinking-Water Quality (Geneva, 2022; ref [20]); the narrative cites both the current edition and historical (2011) editions in different sections.

Implications

Certification (HMTc context, not threshold-setting): This review documents heavy-metal contamination across all four major drinking-water source types in 2015–2024 literature, with Pb, As, Cd, Cr, Ni, Mn, Hg, and U exceeding WHO guidelines in some studies. For HMTc product standards, the review is relevant in two ways: (i) drinking water is itself a directly-consumed matrix that falls under HMTc Category 8 (Water and Water-Based Products), and the review confirms that bottled and sachet water cannot be assumed clean by default; (ii) water is an ingredient input for reconstituted infant formula, beverages, soups, and many processed foods, so water-source contamination contributes to total dietary exposure independent of the food matrix. The review does not provide individual-sample data that could feed a percentile-pooled HMTc threshold; it is occurrence-context evidence, not row-pool material.

Courses: Useful as a global synoptic of heavy-metals-in-drinking-water occurrence over the 2015–2024 decade. Specifically useful for the equity dimension (developing-country water systems and rural/well populations bear disproportionate burden), for explaining why tap and bottled water are not equivalently safe across regions, and for the cross-analyte pattern that groundwater and surface water generally exceed tap and bottled water.

App: Drinking water is a concurrent exposure pathway for any consumer-facing exposure estimate, especially for reconstituted infant formula and beverages. The review’s documentation that bottled/sachet water is not reliably below WHO guidelines for Pb, Cd, Cr, and Mn warrants caveating “bottled water is safer” framing in any consumer-facing translation.

Provenance notes

PDF retrieved via the Manual Fetch Kimi pipeline (P0316). MDPI open-access publication under CC BY 4.0; the publisher’s redistribution terms permit retention and reproduction with attribution. Published online 8 April 2026 (Water 18(8): 886). DOI 10.3390/w18080886. SHA256 of raw PDF: 4e7dfb46e3a19a5be1f1859e697e95c37a1df8102aeb5fdb39bf1076dd36aa05.

Verification notes

  • 2026-05-25 merge-enhance (Claude Code, autonomous ingest skill v2.0). The pre-existing version of this page (updated 2026-05-14) had the following defects, all corrected here:

    1. raw_handle: manual-fetch-kimi — generic, not file-specific. Updated to MFK_p0316-heavy-metals-burden-in-drinking-water.
    2. raw_path pointed to P0316.pdf (filename stub); the actual file is P0316_Heavy Metals Burden in Drinking Water.pdf. Path corrected.
    3. metals: [Pb, Cd, tAs, tHg, Ni, Cr, Al] listed Al. The review does not discuss aluminum anywhere (not in Tables 1–5, not in Table 5 summary, not in the narrative). Al removed. Cu, Fe, Zn, U, Mn, Sb, Se added — all are discussed in the review with per-source ranges.
    4. access_url, raw_sha256, no_doi_assigned, publication volume/issue, sample_n absent. Added.
    5. Cd Key numbers stated “highest in groundwater from Ghana (2.23 mg/L)“. This is incorrect: the review’s narrative and Table 5 both attribute the 2.23 mg/L Ghana Cd value to surface water (Cobbina et al., ref [5]), not groundwater. Corrected.
    6. Mercury Key numbers contained “Groundwater: ND to ~0.026 mg/L tap water; ND to ~126.0 µg/L surface water” — internally muddled (tap-water value mislabeled as groundwater; the 126 µg/L attribution belongs to Cr, not Hg). Replaced with the four matrix-type ranges as actually reported in Tables 1–4 and the overall maximum from Table 5.
    7. Cr, Fe, Cu, Zn, U, Mn, Sb, Se absent from Key numbers despite being in (or warranting addition to) the metals list and discussed extensively in the review. Added.
    8. matrices: [tap-water, groundwater, surface-water, bottled-water] used non-canonical splits. Collapsed to [drinking-water] (the controlled-vocabulary umbrella per docs/gpt-collaboration/system-prompt.md); the four sub-types are now described in the body and in Key numbers per the review’s own Table 1–4 structure.

  • Paper-internal contradictions reproduced as flagged here (faithful to source, not silently corrected):

    • Pb overall maximum: review narrative and abstract attribute 0.791 mg/L to groundwater (Ghana); Table 5 summary row lists it as “Tap water.” The narrative/abstract reading is internally consistent with Table 4 (groundwater table). The wiki reproduces the narrative attribution as the primary statement and flags the Table 5 disagreement.
    • Pb/Hg cross-attribution at Ghana surface water (0.259 mg/L): the value 0.005–0.259 mg/L appears as a Pb surface-water entry for Ghana in Table 3, and the same 0.259 mg/L Ghana surface-water figure appears in Table 5 as the overall Hg maximum (Ghana, Surface Water). The narrative on p11 also names 0.259 mg/L as the highest Hg surface-water concentration. Table 3’s Hg row for Ghana caps at 0.001–0.191 mg/L. So 0.259 mg/L is simultaneously attributed to Pb in Table 3 and to Hg in Table 5 and the narrative — a second paper-internal cross-attribution beyond the well-known Pb Table 5 row.
    • Cd in tap water: narrative says no tap-water entry exceeded 3 µg/L; Table 1 Nigeria entry ND–0.10 mg/L (100 µg/L) contradicts.
    • Cd in groundwater: narrative reports 0.005–0.791 mg/L groundwater range, but 0.791 mg/L is the Pb maximum (Ghana groundwater) — likely transposition in the narrative; Table 4 South Africa entry reaches 4.6 mg/L.
    • Cr overall maximum: Tables 4 and 5 both print 244.4 in mg/L (China), but disagree on water-source type (Table 4 groundwater; Table 5 “Surface water”). The narrative summary on p6 prints “244.43 µg/L in groundwater” — disagreeing with both tables on unit. A 244.4 mg/L groundwater Cr value would be ~5,000× the WHO 50 µg/L guideline and is implausible; the narrative’s µg/L groundwater attribution is the consistent reading.
    • Ni: narrative claims 1.021 ± 0.1 mg/L Nigeria is from a surface-water study; Table 1 places that exact value in the tap-water table for Nigeria (ref [53]). Reproduced both; flagged.
    • Fe tap water Nigeria entry 139.00 ± 13.00 ** mg/L: as written this would be 139 mg/L (≈460× the 300 µg/L WHO aesthetic limit), which is implausible; the narrative says the mean tap-water highest was 0.139 ± 0.013 mg/L, suggesting an * (µg/L) vs ** (mg/L) typesetting error in Table 1 or a missing decimal in the narrative.
    • WHO Hg guideline: narrative says 1 µg/L; Table 5 says 6 µg/L. Both reproduced as written.
    • Se units in Table 3 and the narrative: Algeria 7.70–12.20 mg/L and Ethiopia 0.015–2.64 mg/L would be 192× and 66× the WHO 40 µg/L guideline. Almost certainly a µg/L source paper reproduced as mg/L by the review; flagged.
    • Cu in narrative: China “up to 3816.50 mg/L” and Pakistan/India “up to 2325 mg/L” (paragraph on page 8) are not in Tables 1–4 at these magnitudes and are almost certainly mg/L vs µg/L unit-transcription errors in the original studies that the review reproduced uncritically; flagged.

  • Speciation. Paper measures total As (no inorganic/total split), total Hg (no methylmercury speciation), total Cr (the review explicitly says no included study separated Cr-III from Cr-VI). Frontmatter metals: uses tAs, tHg, Cr accordingly.

  • Brand firewall (Part 12): no consumer-brand contamination values reported (the included studies report by country/region and analytical instrument, not by brand). No content stripped on this pass. Instrument vendor mentions (ICP-MS, ICP-OES, AAS, GFAAS, FAAS, SWASV, HPLC/ICP-DRC-MS, ACP-AES, laser fluorimeter) are scientific-method references retained per the 2026-05-17 audit exception.

  • Wiki/HMTc firewall (Part 2): no HMTc threshold proposals; the Implications section flags drinking water as relevant to HMTc Category 8 occurrence context but does not propose threshold values, does not write consumer-facing advisories, and does not infer that any specific HMTc value should derive from this review’s tabulations. The review itself is a tabulation, not a percentile-pooled dataset, and is appropriately flagged as occurrence-context evidence rather than row-pool material.

  • Evidence tier. Tier B (peer-reviewed but narrative review without meta-analysis, formal quality appraisal, or risk-of-bias scoring). Useful for global occurrence context and for source-pointing to the 66 underlying primary studies, less useful as a standalone evidence source for any quantitative claim.

  • Audit subagent 2026-05-25 (REVISE) — three Check-1 findings verified against the PDF and applied: (1) Pb surface-water highest restated as Ghana 0.259 mg/L from Table 3 (the previous draft named Bangladesh 64.00 µg/L, but Table 3 row 1 shows Ghana 0.005–0.259 ** mg/L = up to 259 µg/L, and the abstract confirms “ND to 0.259 mg/L (surface water)” for Pb); the Pb/Hg cross-attribution at Ghana 0.259 mg/L is now documented as a second paper-internal contradiction. (2) Mn tap-water highest in Table 1 restated as Ecuador 3.000–122.0 µg/L (ref [92], ICP-OES); the previous draft incorrectly named Costa Rica 0.34–86.8 µg/L as highest while listing the higher Ecuador entry in the same bullet. (3) Sri Lanka Mn groundwater range corrected to 2.477–2,241 µg/L per Table 4’s * (µg/L) notation; the previous draft printed “2,477–2,241” with the leading decimal dropped, producing an inverted high-to-low range. Also tightened: the Cr-overall-maximum framing now correctly states that Tables 4 and 5 agree on mg/L (disagreeing only on water-source type) while only the narrative summary on p6 prints µg/L, rather than the previous “Table 4 µg/L vs Table 5 mg/L” framing. Audit verdict: REVISE; Checks 2–5 clean on first pass.

Wiki pages this source may touch

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.

CommitDateDescription
b0f3d382026-06-12batch | corpus rescreen b04 old terminal skips