The WHO has discussed minimum calcium and magnesium levels and suggests remineralizing RO water. Germany's risk agency (BfR, 2020) found no long-term harm with a balanced diet. Observational studies link more magnesium in water to fewer heart outcomes. That is an association, not proof.
| Source | What it found (quoted) and its limits | Evidence |
|---|---|---|
| BfR. Health risk assessment of mineral water with very low mineral content. BfR Opinion No 041/2020. (2020) · new Agency risk assessment · Germany (BfR) · n = n/a |
'does not have any long-term adverse effects on health' assuming 'a balanced diet and normal drinking quantities'; at 30 mg/L Mg, 1 L/day supplies median 30 mg/day (women) and 31.5 mg/day (men), ~8.3% and 7.3% of total Mg intake; hyponatremia in study after 9.2 L over seven hours (judged 'limited relevance') Limits: Assesses bottled water <=50 mg/L TDS, not RO home units directly; at-risk groups (unbalanced diets, pregnancy, athletes >1.5 h) flagged |
Evidence: strong Free to read · source · added 2026-10-30 |
| WHO (Cotruvo J, Bartram J, eds). Calcium and Magnesium in Drinking-water: Public Health Significance. ISBN 978 92 4 156355 0. (2009) · new Agency expert review · International (WHO) · n = n/a |
POU RO and distilling devices 'remove virtually all the minerals from the input water'; suggests bypass or remineralizing unit; Mg benefits 'generally occurred at magnesium concentrations of about 10 mg/l and greater' (not proof of causation); no minimum Ca/Mg value set Limits: Observational basis; WHO explicitly declines causal claim |
Evidence: strong Free to read · source · added 2026-10-30 |
| WHO. Nutrient minerals in drinking water and the potential health consequences of long-term consumption of demineralized, remineralized and altered mineral content drinking waters: Rolling revision of the WHO Guidelines (WHO/SDE/WSH/04.01); Chapter 1 of WHO 'Nutrients in Drinking Water' (2005). (2005) · new Agency expert consensus report (2003 Rome meeting) · International (WHO) · n = n/a |
'several authors have suggested' minimum levels of 'approximately 20 to 30 mg/l calcium and 10 mg/l magnesium in drinking water'; 'Stabilization of demineralized and corrosive drinking water should be done where possible with additives'; 'Unless properly stabilized, demineralized and some natural waters are corrosive to plumbing' Limits: Suggested minima are cited author proposals, not WHO guideline values; no new data |
Evidence: strong Free to read · source · added 2026-10-30 |
| Gu H, Gong Y, Li Z, et al. Contribution of direct-drinking water to calcium and magnesium and the influence on the height in school-age children. Front Nutr. (2024) · new Retrospective cohort + water sampling · Taicang, China; 24 primary schools, 2019-2023 · n = 4850 children |
NF/RO water had 'very low calcium and magnesium contents'; direct-drinking water contributed 8.95% of dietary Ca and 2.78% of dietary Mg; height and height growth did not differ between UF and NF/RO groups (P>0.05) Limits: NF and RO pooled; school water only part of intake |
Evidence: moderate Free to read · source · added 2026-10-30 |
| Shlezinger M, Amitai Y, Akriv A, Gabay H, Shechter M, Leventer-Roberts M. Association between exposure to desalinated sea water and ischemic heart disease, diabetes mellitus and colorectal cancer; a population-based study in Israel. Environ Res 166:620-627. (2018) · new Historical cohort / ecologic (EMR 2004-2013) · Israel; municipal desalinated supply vs non-desalinated regions · n = 450147 (cohort); 177900 (adjusted model) |
Adjusted HR IHD 1.06 (95% CI 1.02-1.11), DM 1.10 (1.06-1.14), CRC 0.93 (0.86-1.02); abstract: 'no time trend was observed' (interaction p = 0.955) Limits: Ecologic exposure; baseline disease prevalence differed between regions; municipal desalination, not home RO |
Evidence: moderate Paywalled · source · added 2026-10-30 |
| Koren G, Shlezinger M, Katz R, Shalev V, Amitai Y. Seawater desalination and serum magnesium concentrations in Israel. J Water Health 15(2):296-299. (2017) · new Population-based pre/post serum survey · Israel; communities switched to desalinated water vs controls · n = 66764 |
Mean serum Mg '2.065 +/- 0.19 mg/dl before desalination' and '2.057 +/- 0.19 mg/dl thereafter (p < 0.0001)'; serum Mg <=1.6 mg/dl rose from 1.62% (2010-2013) to 2.01% (2015-2016), 'increased by 24%' (p = 0.0019) Limits: Abstract only; very small absolute change; lab-test population not random; year: page lists 2016, vol 15 issue |
Evidence: moderate Access unknown · source · added 2026-10-30 |
| Jiang L, He P, Chen J, et al. Magnesium levels in drinking water and coronary heart disease mortality risk: a meta-analysis. Nutrients 8(1):5. (2016) · new Meta-analysis of observational studies · International · n = 9 articles/10 studies; 77821 CHD cases |
Overall 'RR = 0.89, 95% CI = 0.79-0.99, I2 = 70.6'; European subgroup 'RR = 0.83, 95% CI = 0.69-0.98'; cohort studies RR 0.88 (0.57-1.34), not significant Limits: High heterogeneity; highest-vs-lowest only, no dose-response; case-control dominated |
Evidence: moderate Free to read · source · added 2026-10-30 |
| Catling LA, Abubakar I, Lake IR, Swift L, Hunter PR. A systematic review of analytical observational studies investigating the association between cardiovascular disease and drinking water hardness. J Water Health 6(4):433-442. (2008) · new Systematic review + meta-analysis of observational studies · International · n = 14 studies (9 case-control; 5 cohort) |
'Of 2,906 papers identified, 14 met the inclusion criteria'; pooled 'OR 0.75 (95%CI 0.68, 0.82), p < 0.001' for magnesium and cardiovascular mortality; 'Only two studies reported a statistically significant effect for calcium' Limits: Abstract only; observational/ecological exposure; heterogeneity for Ca |
Evidence: moderate Access unknown · source · added 2026-10-30 |
| Thippeswamy HM, Shanbhog R, Nanditha Kumar M, Prashanth SN, Smitha P. Comparison of serum calcium, magnesium, phosphate, alkaline phosphatase, and vitamin D levels in children consuming reverse osmosis, non reverse osmosis, and high fluoride drinking water. Sci Rep 15:10689. (2025) · new Cross-sectional · Mysore district, India; children 9-13 · n = 180 (60 per group) |
Serum calcium RO 9.19 (0.46) vs non-RO 10.02 (0.41) (units printed as mmol/L in paper, values consistent with mg/dl); serum Mg RO 2.29 (0.33) vs non-RO 2.11 (0.14) mg/dl; RO water Ca 5.24 (3.59) mg/l, TDS 13.06 (8.15) mg/l; vitamin D p = 0.482 Limits: Cross-sectional, self-reported exposure; Mg direction opposite to hypothesis; unit inconsistency |
Evidence: weak Free to read · source · added 2026-10-30 |
| Rosborg I, Kozisek F, Ferrante M. Health effects of de-mineralization of drinking water. In: Rosborg I, Kozisek F (eds). Drinking Water Minerals and Mineral Balance. Springer, pp. 149-160. (2019) · new Narrative review (book chapter) · International · n = n/a |
Czech regulation for treated water 'TDS > 150 mg/L, Ca > 30 mg/L and Mg > 10 mg/L'; Israel post-AMI 1-year survival '95% had survived' (hard-water region) vs ~90% (desalinated region); rat hemoglobin ~19% lower with non-supplemented demineralized water; 'no evidence that desalinated water causes cancer' Limits: Narrative, author advocacy for minimum mineral levels; secondary citations |
Evidence: weak Paywalled · source · added 2026-10-30 |
Evidence grade: strong = trials, large field studies or official datasets; moderate = smaller or lab studies with real water; weak = single small studies, makers' specs or secondary reports. "Free to read" means the source itself is free to read online.