Magnesium hydroxide

What it is

Magnesium hydroxide is an inorganic magnesium salt, Mg(OH)2, sold as a white aqueous suspension (milk of magnesia) and in chewable antacid tablets. In the stomach it reacts with hydrochloric acid, which is why it is sold as an antacid, and the magnesium that is not absorbed stays in the gut and draws water in, which is why the same product is also sold as a saline laxative. A US Drug Facts label for a 30 mL dose states the strength as magnesium hydroxide 2400 mg and the magnesium content as 1 g.

What the research says

The evidence is strongest for short-term, symptomatic use, and it is thin. As a laxative it has a human dose-response trial in six volunteers with normal bowel habits showing more bowel movements, wetter and bulkier stools. In children, head-to-head randomised trials against polyethylene glycol detected no difference in success rates, and a Cochrane review pooling three of them found slightly more stools per week on polyethylene glycol on low to very low certainty evidence; acceptance by taste was worse in two paediatric trials but better than polyethylene glycol in a crossover trial in anorectal malformation. As an antacid, the human evidence is mostly pH studies and trials of aluminium/magnesium hydroxide combinations rather than magnesium hydroxide alone, and the pH study's own authors call its gastric effect minimal. The serious harm is hypermagnesaemia, documented in case reports including deaths, one of them in a woman with normal kidney function. The percentage rates quoted on this page come from cohort and cross-sectional studies of the closely related salt magnesium oxide, not of magnesium hydroxide; we found no comparable rate literature for magnesium hydroxide itself.

Evidence grade: Limited evidence.

How it works

Drug class: Inorganic magnesium salt; antacid and osmotic (saline) laxative

Magnesium hydroxide neutralises stomach acid, and the magnesium that stays in the gut holds water in the bowel so stools become wetter and more frequent. A dose-response crossover trial in volunteers over 55 found that rising doses produced more bowel movements, a higher percentage of stool water, larger stool volume and more stool magnesium, alongside a rise in stool prostaglandin E2 whose causal contribution the authors said was unknown. (Source 1)

What it is used for

• A dose-response randomised crossover trial in six volunteers over 55 with NORMAL bowel habits shows the laxative effect is real and dose-related; it did not test treatment success in constipated people. Against polyethylene glycol the head-to-head trials in children detected no difference in success rates, and a Cochrane review pooling three of them found slightly more stools per week on polyethylene glycol, on low to very low certainty evidence. Acceptance is mixed, not uniformly worse: two paediatric trials found significantly more children refusing magnesium hydroxide, while a crossover trial in anorectal malformation scored magnesium hydroxide ABOVE polyethylene glycol on user preference. The trials are small and mostly open-label. Evidence: limited. (Source 1)
• Antacids containing magnesium hydroxide raise oesophageal pH more than placebo in people with heartburn, in a single-blind crossover trial whose abstract does not state whether treatment order was randomised, and a 565-patient randomised, double-blind, placebo-controlled trial found an antacid relieved a higher median proportion of a patient's heartburn episodes than placebo (62% against 41%). The trials tested an aluminium/magnesium hydroxide combination or an unspecified antacid, not magnesium hydroxide on its own, and the pH trial's own authors call the gastric effect minimal and locate the action in the oesophagus. Evidence: limited. (Source 2)
• A randomised crossover trial in 15 patients with surgically corrected anorectal malformation detected no statistically significant difference in radiographic faecal clearance between magnesium hydroxide, senna and polyethylene glycol. With 15 patients the trial had almost no power, so this is an absence of evidence rather than evidence of no difference. Evidence: not-supported. (Source 3)

Interactions

• Prescription medicines generally, through the rise in stomach pH (label): Under the ANTACID warnings the label states that antacids may interact with prescription drugs and tells anyone taking one to ask a doctor or pharmacist first. The laxative block carries the same warning in its own words. (Source 4)
• Ibuprofen (pharmacokinetic study): Magnesium hydroxide taken with ibuprofen speeds its absorption, raising early blood levels, without increasing the total amount absorbed. Limit: The quote reproduces the MEDLINE record, which renders the journal's printed "P < 0.05" and "P < 0.01" in 1991-era ASCII as "P less than 0.05" and "P less than 0.01". The figures are the journal's; only the way the inequality is typed is the indexing database's. Note also the authors' own recommendation is specific: they recommend an antacid "which contains magnesium hydroxide without aluminium" when a fast analgesic effect is wanted, so this does not extend to combined aluminium/magnesium antacids. (Source 5)
• Ketoprofen and diclofenac (pharmacokinetic study): In the same crossover studies no significant change was detected in the rate or the extent of absorption of these two anti-inflammatories. With six volunteers that is an unpowered null rather than a measured absence of effect. Limit: Six subjects. A non-significant result in a trial that size does not show the absence of an interaction. The quote also reproduces MEDLINE's 1991-era ASCII rendering of the journal's inequalities elsewhere in the same abstract. (Source 5)
• Iron (ferrous sulfate) (clinical trial): Giving magnesium hydroxide half an hour after a large iron dose did not reduce iron absorption in healthy men, so the feared binding interaction was not seen at that timing and ratio. (Source 6)
• Tebipenem pivoxil hydrobromide (an oral antibiotic) (pharmacokinetic study): An aluminium hydroxide/magnesium hydroxide/simethicone suspension taken at the same time lowered peak blood levels of the antibiotic by about a fifth (the omeprazole arm, by comparison, lowered them by about two fifths), with about an 11% fall in total exposure. Limit: The quote now starts at "Following co-administration" so you can see which figure belongs to which arm: 22% is the antacid and 43% the omeprazole. One number in it is misprinted in the journal itself - the lower confidence bound is given as "83,2", with a comma where a decimal point belongs, in both the PubMed record and the publisher's own abstract. It should read 83.2. We have left the source as printed. The study was in healthy adults given single doses, not in patients being treated for an infection. (Source 7)
• Riociguat (pharmacokinetic study): An aluminium/magnesium hydroxide antacid cut riociguat absorption by about a third and its peak level by over half in HEALTHY MALE volunteers, not in patients, because riociguat dissolves poorly without stomach acid. The authors' own instruction from this result is that antacids should not be taken within an hour of riociguat; they say no dose adjustment is needed for a proton pump inhibitor. Limit: Two open-label crossover studies of 12 healthy males each, with single doses. The finding has not been tested in people actually being treated for pulmonary hypertension. The separation interval in the plain-language summary is the study authors' own instruction, quoted from the same abstract, and is not advice from us. (Source 8)
• Reduced kidney function and a magnesium-restricted diet (label): Under the SALINE LAXATIVE warnings - the laxative block, not the antacid one - the label lists kidney disease and a magnesium-restricted diet among four reasons to ask a doctor before use. The antacid block lists the first two of those only. The cohort and case literature shows this is where hypermagnesaemia concentrates. (Source 4)
• Alcohol (label): UNTESTED, not tested and found safe. We found no study of a magnesium hydroxide and alcohol interaction in this search. The label's own interaction warning is general - it says antacids may interact with prescription drugs and to ask first - and says nothing about alcohol either way. Limit: Read this as a gap in the literature, not as a negative result. Nobody has run the study, so there is no evidence either that alcohol matters here or that it does not. (Source 4)

Stopping it

• In the reported hypermagnesaemia case, stopping magnesium hydroxide alongside continued haemodialysis was the treatment that led to recovery, so stopping is itself the intervention when magnesium rises. (Source 9)
• The product's US label sets a stopping rule for the LAXATIVE use specifically: stop and ask a doctor if there is rectal bleeding or no bowel movement after use, or if a laxative is needed for more than one week. The two-week figure that appears elsewhere on the label belongs to the antacid use, not this one. (Source 4)
• There is no trial of weaning magnesium hydroxide specifically. The nearest evidence is an open-label pilot randomised trial of weaning children off a different osmotic laxative, lactulose, in which neither reducing the dose nor reducing the frequency affected the weaning success rate - a null in a 16-patient pilot whose own authors say the sample size could not be calculated, so it does not show the two methods are equally bad. The same trial reports that a pre-weaning lactulose dose above 1.5 g/kg/day may lead to better outcomes at week 12, and that success stayed low even after a gradual reduction over more than three months. (Source 10)

What goes wrong

Two published cases describe severe hypermagnesaemia in people taking magnesium hydroxide for constipation, one with end-stage kidney disease who recovered and one with normal kidney function who died. (Source 9)

• Case report, Very low certainty.
• Size: 2 patients.
• Who: An 82-year-old woman with end-stage renal disease and a 50-year-old woman with normal renal function, both taking magnesium hydroxide for constipation.
• How long: Not stated for either case.
• Result: Serum magnesium 9.9 mg/dL in the first case (metabolic encephalopathy, recovered after stopping the drug and haemodialysis) and 11.0 mg/dL in the second (cardiac arrest, death)
• Funding: not stated.

The first case involved an 82-year-old woman with end-stage renal disease who developed metabolic encephalopathy due to hypermagnesemia, after taking 3,000 mg of magnesium hydroxide daily for constipation.

A case report with a literature review reported that severe hypermagnesaemia is rare but, when it happens, frequently ends in death even in people with normal kidney function and despite renal replacement therapy. (Source 11)

• Case report, Very low certainty.
• Size: 1 patient plus a literature review.
• Who: A 53-year-old woman with chronic constipation treated regularly with magnesium-containing laxatives, admitted for a COPD exacerbation.
• How long: Died 2 days after renal replacement therapy began.
• Result: Serum magnesium rose from 2.0mg/dL on admission to a peak of 10.8mg/dL despite hydration and diuresis with normal kidney function; renal replacement therapy lowered the level but the patient deteriorated and died 2 days later. The review sentence gives no frequency figure - it says only "rare and often unsuspected" and "frequently results in death"
• Funding: not stated.

Limit of this finding: Read the scope carefully. The sentence about frequent death is about severe hypermagnesaemia in general, not specifically about laxative cases: the source does not put a "from laxatives" qualifier on it. The link to magnesium-containing laxatives is made by the case itself and by the next sentence of the abstract, which says retained laxative in the gut acts as a reservoir for continuing magnesium absorption. The source also calls severe hypermagnesaemia "rare and often unsuspected" and says the deaths happen "despite renal replacement therapy", so this is a statement about how bad the rare event is, not about how often it occurs. One case plus a narrative literature review cannot give a rate.

A review of the literature shows that though rare and often unsuspected, severe hypermagnesemia frequently results in death even in individuals with normal renal function despite renal replacement therapy.

In a hospital cohort of patients prescribed the related salt MAGNESIUM OXIDE - not magnesium hydroxide - 23% developed hypermagnesaemia, and four factors were independently associated with it: a low estimated glomerular filtration rate, a raised blood urea nitrogen, a higher magnesium oxide dose and a longer duration. (Source 12)

• Cohort study, Low certainty.
• Size: 320 of 3258 hospitalised patients prescribed magnesium oxide.
• Who: Hospital inpatients in Japan prescribed magnesium oxide tablets. This is magnesium oxide, a different salt: it carries about 1.5 times the elemental magnesium per mg that magnesium hydroxide does, so the dose thresholds here do not transfer directly to magnesium hydroxide.
• How long: Records over October 2015 to September 2017.
• Result: 75/320 (23%) developed hypermagnesaemia (grade 1 in 62, 4% at grade 3); independent risk factors were eGFR 55.4 mL/min or less (OR 3.105, P = 0.001), blood urea nitrogen 22.4 mg/dL or more (OR 3.490, P < 0.001), dose 1650 mg/day or more (OR 1.914, P = 0.039) and duration 36 days or more (OR 2.198, P = 0.012)
• Funding: not stated.

Limit of this finding: These rates are magnesium oxide, not magnesium hydroxide. We found no cohort of comparable size reporting hypermagnesaemia rates for magnesium hydroxide specifically, so the rate literature used on this page is borrowed from the oxide salt and the thresholds are not interchangeable.

Seventy-five patients out of 320 (23%) developed grade 1 and grade 3 hypermagnesemia, with the occurrence of grade 1 and grade 3 in 62 (19%) and 13 (4%) patients, respectively.

In a cross-sectional review of records of patients on daily magnesium oxide for functional constipation, 16.6% of those who had a magnesium measured had a high serum magnesium and 5.2% met the study's definition of hypermagnesaemia; in the paper's own table the only factor significantly associated with hypermagnesaemia was chronic kidney disease grade 4. (Source 13)

• Survey study, Low certainty.
• Size: 193 of 2,176 patients on daily magnesium oxide who had serum magnesium measured.
• Who: Patients whose records were reviewed retrospectively at National Hospital Organization Tokyo Medical Center from September to December 2017 after being given magnesium oxide as a laxative for functional constipation; mean age in the analysed set 73.2 +/- 14.6 years. This is magnesium OXIDE, not magnesium hydroxide.
• How long: Retrospective, cross-sectional.
• Result: 32/193 (16.6%) had serum magnesium at or above 2.5 mg/dl and 10/193 (5.2%) at or above 3.0 mg/dl, with the highest level seen 4.5 mg/dl. In the paper's Table 2 the only factor significantly associated with hypermagnesaemia was chronic kidney disease grade 4 (eGFR under 30 ml/min/1.73 m2, p = 0.014); the sentence quoted here groups magnesium oxide dose over 1,000 mg/day (p = 0.053) and age over 75 years (p = 0.098) together as only "marginally associated", and Table 2 marks both with the same marginal-significance symbol, although the paper's abstract and Conclusion go further and say age was not associated at all. Dose WAS significantly associated with the lower threshold, high serum magnesium at or above 2.5 mg/dl (p = 0.004). None of the patients with a raised magnesium had symptoms of hypermagnesaemia.
• Funding: not stated.

Limit of this finding: The paper's abstract and its own Table 2 do not agree about the dose. The abstract prints p = 0.009 for magnesium oxide dose and hypermagnesaemia; the full text's Table 2 and Results put that figure at p = 0.053 and call dose only "marginally associated", and the string 0.009 appears nowhere in the full text. We have followed the paper's own table. So do not read this study as showing that a higher dose significantly raises the risk of hypermagnesaemia at the 3.0 mg/dl threshold - at that threshold only reduced kidney function reached significance. Dose did reach significance at the lower 2.5 mg/dl threshold. Two further oddities are the source's own and have been left exactly as printed: the abstract and the full text both print "p = 0.000" for chronic kidney disease grade 4, and a p-value is never exactly zero - the authors mean p < 0.001; and the same sentence prints the grammatical slip "CKD grades 4". Also note the rates here are for magnesium oxide, a different salt from magnesium hydroxide, and that none of the patients with a raised magnesium had symptoms.

The only factor significantly associated with hypermagnesemia among patients in the Per Protocol Set was CKD grade 4 (eGFR <30 ml/min/1.73 m2, p = 0.014), whereas factors marginally associated with hypermagnesemia were magnesium oxide dosage >1,000 mg/day (p = 0.053) and age >75 years (p = 0.098)

What the evidence supports

In a randomised, placebo-controlled, double-blind crossover trial in volunteers over 55, magnesium hydroxide produced dose-dependent increases in bowel movements, stool water, stool volume and stool magnesium. (Source 1)

• Randomized trial, Low certainty.
• Size: 6 volunteers, four 5-day inpatient periods each.
• Who: Volunteers more than 55 years old with normal bowel habits.
• How long: Four inpatient periods of 5 days, with 9 days off medication between.
• Result: Stool prostaglandin E2 rose from 95 +/- 18 pg/24 h on placebo to 260 +/- 100, 357 +/- 117 and 525 +/- 196 at the three ascending doses; bowel movements, stool water percentage, stool volume and stool magnesium all rose dose-dependently.
• Funding: Research Support, U.S. Gov't, P.H.S. and Research Support, Non-U.S. Gov't per the record.

Compared to placebo, Mg(OH)2 caused the following dose-dependent results: (a) increased number of bowel movements; (b) increased percentage of stool water; (c) increased stool volume; (d) increased stool Mg2+

The same trial recorded no difference in adverse events between magnesium hydroxide and polyethylene glycol and no severe clinical or biochemical adverse events over 12 months. (Source 14)

• Randomized trial, Low certainty.
• Size: 83 children.
• Who: Children 6 months to 18 years with functional constipation, in an open-label trial.
• How long: 12 months.
• Result: No between-group difference in adverse events; no severe clinical or biochemical adverse events registered in either arm.
• Funding: not stated.

Limit of this finding: Adverse events were collected in an open-label trial of 83 children over 12 months. That is too small and too short to detect an uncommon harm such as hypermagnesaemia, so "no severe clinical or biochemical adverse events were registered" means none were seen here, not that none occur.

There were no differences regarding adverse events between the two groups and no severe clinical or biochemical adverse events were registered.

In a single-blind crossover trial in people with heartburn, an antacid containing aluminium and magnesium hydroxide raised oesophageal pH more than placebo and was the only one of the two antacids tested to raise gastric pH above placebo, for a short time in each case. (Source 2)

• Blood level study, Low certainty.
• Size: 83 subjects.
• Who: Subjects with heartburn, given a refluxogenic meal. The source calls it a single-blind crossover trial and never says it was randomised; NCBI indexes it as Journal Article only, with no trial publication type.
• How long: Single doses; pH followed from 1 h before to 4 h after the meal.
• Result: Both antacids raised oesophageal pH versus placebo; oesophageal area under the pH-time curve was greater for aluminium/magnesium hydroxide than calcium carbonate (p < 0.05); duration of action 82 min in the oesophagus and 26 min in the stomach.
• Funding: not stated.

Limit of this finding: The authors themselves describe the gastric effect as minimal and say the action is mainly in the oesophagus, not the stomach: "the magnitude and duration of action of both antacids on esophageal pH, in contrast to minimal effects on gastric pH, suggest strongly that the lower esophagus is the primary site of antacid activity in relief of heartburn." Also, the 82 minutes of oesophageal action is reported against calcium carbonate's 60 minutes, not against placebo. This is a single-blind crossover trial measuring pH, not a randomised trial measuring symptoms.

Both antacid formulations significantly increased esophageal pH, as compared with placebo.

What the evidence does not support

In a 12-month open-label controlled trial in children with functional constipation, no difference in success was detected between magnesium hydroxide and polyethylene glycol 3350, and no overall difference in acceptability either, although significantly more children rejected magnesium hydroxide in the two older age subgroups. (Source 14)

• Randomized trial, Low certainty.
• Size: 83 children.
• Who: Children 6 months to 18 years with functional constipation; the trial was open-label, so neither children nor assessors were blinded.
• How long: 12 months.
• Result: Success 40/41 (97.6%) with polyethylene glycol versus 40/42 (95.2%) with magnesium hydroxide, p = 0.616 - a non-significant difference, not an identical rate; no overall difference in acceptability, with significantly more rejections of magnesium hydroxide only in the subgroups over 4 to 12 years (P = .037) and over 12 to 18 years (P = .020)
• Funding: not stated.

Limit of this finding: Non-significance is not equivalence. With 83 children and no non-inferiority margin reported, this trial could not have detected a modest real difference. Read it as "no difference was detected", not as "the two work equally well". The authors' own Conclusions do say the two were "equally effective and safe", which is their wording, not a demonstrated equivalence.

There were no differences in success between groups (40/41 PEG vs 40/42 MH, p = 0.616).

In a 6-month randomised trial in children, magnesium hydroxide and polyethylene glycol 4000 improved every measure with no statistically significant difference, but 42.9% of children refused magnesium hydroxide. (Source 15)

• Randomized trial, Very low certainty.
• Size: 38 children (17 polyethylene glycol, 21 magnesium hydroxide)
• Who: Children with chronic functional constipation.
• How long: 6 months.
• Result: All variables improved in both arms with no statistically significant differences; refusal 0% for polyethylene glycol versus 42.9% for magnesium hydroxide.
• Funding: not stated.

All variables analyzed improved for both groups, with no statistically significant differences. All children accepted polyethylene glycol, while 42.9% refused magnesium hydroxide.

In a randomised crossover trial in 15 patients with surgically corrected anorectal malformation, no significant difference was detected between magnesium hydroxide, senna and polyethylene glycol on radiographic faecal clearance. (Source 3)

• Randomized trial, Very low certainty.
• Size: 15 patients, each receiving all three laxatives.
• Who: Patients with surgically corrected anorectal malformation and diagnosed constipation. The abstract says only "Fifteen patients" and never states their age; the journal is Pediatric Surgery International, so they were very probably children, but the source does not print it.
• How long: Three 21-day treatment periods separated by washouts.
• Result: Clean faecal loading in 40% with senna, 46.67% with magnesium hydroxide, 60% with polyethylene glycol, p = 0.655; mean post-treatment Leech scores 6.67 +/- 2.09, 6.80 +/- 2.37 and 5.80 +/- 2.04 respectively, p = 0.841.
• Funding: not stated.

Limit of this finding: With 15 patients this trial had almost no power to separate three laxatives, so the absence of a significant difference is an unpowered null and not a finding that the three work equally well.

Clean fecal loading was achieved in 40% of cases with Sennosides, 46.67% with Mg(OH)2, and 60% with PEG(p = 0.655).

In a crossover comparative trial in constipated elderly long-stay patients, magnesium hydroxide did not change plasma lipids, whole blood minerals or vitamins A and E compared with a bulk laxative. (Source 16)

• Blood level study, Very low certainty.
• Size: 64 constipated elderly long-stay patients, 55 on diuretics.
• Who: Constipated elderly long-stay hospital patients, 55 of 64 on diuretics. The abstract says only "In a crossover study" - it never says the allocation was randomised, and PubMed indexes the paper as Clinical Trial / Comparative Study / Controlled Clinical Trial, NOT Randomized Controlled Trial. What it measured was blood chemistry (serum lipids, carbohydrates, uric acid, vitamins A and E, whole-blood minerals), not a clinical outcome.
• How long: Crossover, duration not stated in the abstract.
• Result: No significant differences in plasma lipids, whole blood minerals or vitamins A and E; hypomagnesaemia in 11 (17%) after bulk laxative versus 2 (2%) after magnesium hydroxide.
• Funding: not stated.

Limit of this finding: Two things about this entry, one about the study and one about how this page files it. First, the study: it is a CONTROLLED, NON-RANDOMISED crossover trial - the abstract never says anyone was randomised, and PubMed types it Controlled Clinical Trial rather than Randomized Controlled Trial - so it is weaker than a randomised trial, and its "no significant difference" comes from a small sample that was not powered to rule out a difference. Second, the filing: this page's design list has NO value for a controlled non-randomised interventional trial. We have recorded it as "pharmacokinetic" because that is defined for this project as a small human study of blood levels rather than outcomes, which is exactly what this is - but it is not a pharmacokinetic study in the pharmacological sense, because it measured no drug concentration and no absorption, distribution or elimination parameter. Do not read the design label as meaning more than that. Third, the source's own arithmetic does not quite work: it prints hypomagnesaemia in 2 of 64 patients as 2%, where 2/64 is 3.1%. We have left the source's figure as printed.

There were no significant differences in plasma lipids, whole blood minerals or vitamins A and E using either laxative.

A randomised but unblinded trial testing magnesium hydroxide as a way to block iron absorption after a supratherapeutic iron dose detected no effect on serum iron at a magnesium hydroxide to elemental iron ratio of 5 to 1. (Source 6)

• Randomized trial, Very low certainty.
• Size: 16 healthy fasting men.
• Who: Healthy fasting male volunteers given 10 mg/kg elemental iron as ferrous sulfate, with magnesium hydroxide at 5 mg for every 1 mg of ingested elemental iron given half an hour later; 8 per group, unblinded.
• How long: Serum iron followed for 6 or 7 hours.
• Result: Mean peak serum iron 300.8 mcg/dL in controls versus 272.5 mcg/dL (the paper prints these as micrograms/dL) with magnesium hydroxide; no statistical difference between groups (P = .20)
• Funding: Research Support, U.S. Gov't, P.H.S. per the record.

Limit of this finding: This is an 8-versus-8 null in an unblinded trial, and the authors attach a condition to it: their conclusion holds "when the ratio of MgOH to elemental iron is 5:1". It does not show that magnesium hydroxide never affects iron absorption, and with 16 subjects it could not have detected a modest effect.

Mean serum iron levels at each time point and peak serum iron levels did not differ significantly between groups.

A Cochrane systematic review of laxatives for childhood constipation pooled three trials in 211 children and found more stools per week on polyethylene glycol than on milk of magnesia, while rating its own evidence low to very low certainty. (Source 17)

• Systematic review, Low certainty.
• Size: 3 trials, 211 participants, within a review of 25 randomised trials and 2,310 participants.
• Who: Children aged 0 to 18 years with functional constipation.
• How long: Searches from inception to 10 March 2016; the review notes follow-up in the included trials was short.
• Result: Mean difference 0.69 more stools per week with polyethylene glycol than with milk of magnesia (95% CI 0.48 to 0.89), which the reviewers themselves call "quite small and may not be clinically significant". One child was allergic to polyethylene glycol; no other serious adverse events were reported. Fourteen of the 25 included trials were judged at high risk of bias for lack of blinding, incomplete outcome data and selective reporting.
• Funding: not stated in the abstract.

Limit of this finding: The review's own GRADE rating for this outcome is low or very low certainty, because of sparse data, heterogeneity and high risk of bias in the pooled trials, and its authors say the results "should be interpreted with caution". One line in the same passage cannot be read on its own: the review reports a single study as "favouring milk of magnesia over lactulose (MD -1.51, 95% CI -2.63 to -0.39, 50 patients)", and the abstract never states which arm is the reference, so the negative sign cannot be reconciled with the stated direction from the abstract alone. We therefore do not rely on that line.

Meta-analysis of 3 studies with 211 participants comparing PEG with milk of magnesia showed that the stools per week were significantly greater with PEG (MD 0.69, 95% CI 0.48 to 0.89). However, the magnitude of this difference was quite small and may not be clinically significant.

The same Cochrane review's conclusion is that polyethylene glycol may be superior to milk of magnesia for childhood constipation, on low to very low certainty evidence. (Source 18)

• Systematic review, Low certainty.
• Size: 25 randomised trials, 2,310 participants.
• Who: Children aged 0 to 18 years with functional constipation.
• How long: Searches to 10 March 2016.
• Result: The reviewers rate the certainty of the evidence for their primary outcome, stools per week, as low or very low, and attribute that to sparse data, inconsistency and high risk of bias; they ask for the pooled analyses to be read with caution.
• Funding: not stated in the abstract.

The pooled analyses suggest that PEG preparations may be superior to placebo, lactulose and milk of magnesia for childhood constipation. GRADE analyses indicated that the overall quality of the evidence for the primary outcome (number of stools per week) was low or very low due to sparse data, inconsistency (heterogeneity), and high risk of bias in the studies in the pooled analyses.

Where the research disagrees

Whether magnesium hydroxide or polyethylene glycol should be preferred for childhood constipation

• Authors of the 12-month controlled trial in children from 6 months to 18 years, Open-label parallel-group controlled clinical trial, 83 children, 12 months. "Equally effective and safe" is the authors' own wording; the trial rested on 40/41 against 40/42 successes (p = 0.616) with no non-inferiority margin reported, so it is not a demonstrated equivalence: The two laxatives were equally effective and safe for treating FC in children from 0.5 to 18 years of age. (Source 19)
• Authors of the 6-month randomised trial of polyethylene glycol 4000 versus magnesium hydroxide, Randomised comparative trial, 38 children, 6 months: The two laxatives showed no difference in effectiveness for the treatment of constipation. However, due to its better acceptance, because it is odorless and tasteless, polyethylene glycol proved to be a better option for treating chronic functional constipation. (Source 20)

How much

• Reference intake: No dietary reference intake applies: this is a medicine, and the amount taken is set by the product's Drug Facts directions or by a clinician. The US label published 20 August 2026 for a magnesium hydroxide oral suspension states a 30 mL dose contains magnesium hydroxide 2400 mg and directs adults and children 12 years and over to 30 mL up to twice a day as an antacid. (Source 4)
• Upper limit: The label carries a separate ceiling and a separate duration for each of the product's two uses, and they are not the same. For the ANTACID use the label's "When using this product" block says: do not take more than 60 mL (2 doses) in a 24-hour period, and do not use the maximum dosage for more than 2 weeks except under the advice and supervision of a doctor (the label prints "superision"). For the SALINE LAXATIVE use the "Directions for Saline Laxative" block sets the same 60 mL daily maximum, but the laxative duration limit is ONE week, not two: the laxative "Stop use and ask a doctor if" block says to stop if you need to use a laxative for more than one week (the label prints "more that 1 week"). These are labelling positions dated 20 August 2026, not toxicological upper limits. (Source 4)
• Studied: The mechanism trial gave volunteers 45 ml containing placebo or 1,200, 2,400 or 3,600 mg of magnesium hydroxide with 240 ml of water at 8 p.m. on each study day. (Source 1)
• Studied: In the two-case report, a dose is given for one case only. The 82-year-old woman with end-stage renal disease who developed metabolic encephalopathy, reached a serum magnesium of 9.9 mg/dL and RECOVERED had been taking 3,000 mg of magnesium hydroxide daily for constipation. No dose is reported anywhere in the source for the second, FATAL case - a 50-year-old woman with normal renal function whose magnesium reached 11.0 mg/dL. The 3,000 mg figure must not be read as a fatal dose. (Source 9)
• Studied: This threshold is MAGNESIUM OXIDE, not magnesium hydroxide. In the hospital cohort of patients prescribed magnesium oxide, the independent dose threshold for hypermagnesaemia was a magnesium oxide dose of 1650 mg/day or more and a duration of 36 days or more. Magnesium oxide carries about 1.5 times as much elemental magnesium per mg as magnesium hydroxide, so 1650 mg of magnesium oxide is roughly 995 mg of elemental magnesium, which would take about 2,400 mg of magnesium hydroxide to match. The figure is not a magnesium hydroxide dose and must not be read as one. (Source 12)

A common belief, and what the research shows

The belief: Milk of magnesia is just a gentle, old-fashioned remedy that cannot do real harm.

What the research shows: The magnesium it delivers can accumulate to dangerous levels. Published cases include an 82-year-old woman with end-stage kidney disease who developed encephalopathy with a serum magnesium of 9.9 mg/dL while taking 3,000 mg daily, and who recovered when the drug was stopped and dialysis continued; and a 50-year-old woman with NORMAL kidney function whose serum magnesium reached 11.0 mg/dL and who had a cardiac arrest and died. The source gives no dose for the fatal case, so the 3,000 mg figure belongs to the case that recovered and is not a fatal dose. A separate case report with a literature review concluded that severe hypermagnesaemia frequently results in death even when kidney function is normal.

Questions and answers

What is it?

Magnesium hydroxide is a simple magnesium salt, Mg(OH)2, sold as the white suspension known as milk of magnesia and in chewable antacid tablets. It is a medicine, not a nutrient supplement, and the same product is marketed for two different jobs. A US Drug Facts label states a 30 mL dose contains magnesium hydroxide 2400 mg and describes the product's purpose as antacid and saline laxative. (Source 4)

What does it do in the body?

In the stomach it neutralises acid. Further down, the magnesium that is not absorbed holds water in the bowel, so stools become wetter, bulkier and more frequent. A randomised, placebo-controlled, double-blind crossover trial in six volunteers over 55 showed all of those effects rising with the dose. Those volunteers had normal bowel habits, so the trial shows the laxative mechanism working, not treatment success in constipation. (Source 1)

Is it good or bad for you?

It depends almost entirely on kidney function and on how long it is used. In a 12-month open-label trial in 83 children no severe clinical or biochemical adverse events were recorded in either arm. That is reassuring but it is a small, short trial and cannot rule out an uncommon harm. Taken daily, or taken by someone whose kidneys clear magnesium poorly, it can cause hypermagnesaemia, which has been fatal - including in a reported case with normal kidney function. (Source 14)

How do you get more of it?

It is bought over the counter as a suspension or chewable tablet, and the label sets how much and for how long for each of its two uses separately. For the antacid use the label directs adults and children 12 years and over to 30 mL up to twice a day, with a 60 mL daily ceiling and a two-week limit at maximum dosage. For the laxative use the ceiling is the same 60 mL a day, but the label's stop-use rule for a laxative is one week, not two. These are the label's directions, not advice for any individual. (Source 4)

If it is harmful, what reduces it?

When magnesium has built up, the documented step is stopping the drug, with dialysis where kidney function cannot clear it. In the reported case of an 82-year-old with end-stage renal disease and metabolic encephalopathy, stopping magnesium hydroxide together with continued haemodialysis led to recovery. (Source 9)

Why might someone be low in it or missing it?

Does not apply in the way it would for a nutrient. Magnesium hydroxide is a manufactured medicine taken at a stated strength, not something the body makes or stores, so nobody is deficient in it. The relevant question for this drug is the opposite one, too much magnesium rather than too little, and the label's laxative warnings flag kidney disease and a magnesium-restricted diet as two of four reasons to check with a doctor first. (Source 4)

Which whole foods contain it or feed it?

No whole food contains magnesium hydroxide as such; it is a manufactured salt. The magnesium it delivers is the same element found in foods, and a 30 mL dose of the labelled product supplies 1 g of magnesium, but we found no study in this search comparing dietary magnesium with magnesium hydroxide for either the antacid or the laxative effect. (Source 4)

We searched: Europe PMC searches for magnesium hydroxide combined with bioavailability, absorption, gastric pH and acid-neutralising capacity, and for magnesium hydroxide with constipation or laxative; no human study comparing foods with the drug was found.

What happens if you do not have it?

Nothing is lost by not taking it; the symptoms it treats simply go untreated. In the paediatric trials the comparison was against another laxative rather than against nothing, and no difference in success was detected: 40 of 41 children on polyethylene glycol and 40 of 42 on magnesium hydroxide met the success definition (p = 0.616). A Cochrane review pooling three such trials found slightly more stools per week on polyethylene glycol, on low to very low certainty evidence. (Source 14)

How can you test for it?

The relevant test is a serum magnesium measurement, used to detect the drug's main harm rather than to measure a deficiency. Its real-world weakness is that it is rarely done. In a records review of 2,176 patients given magnesium oxide as a laxative, 1,499 of them had their creatinine measured but not their magnesium, and only 193 had a magnesium assay at all. Among those 193, 16.6% had a high level. A normal result also does not rule out a later rise, because retained drug in the gut can keep releasing magnesium. (Source 21)

References

1. Journal of clinical gastroenterology. Magnesium hydroxide: new insights into the mechanism of its laxative effect and the potential involvement of prostaglandin E2 — abstract, latter part of the abstract (dosing, stool results and summary). 1992. PMID 1556404. Read the source
2. American journal of therapeutics. Effects of Aluminum/Magnesium Hydroxide and Calcium Carbonate on Esophageal and Gastric pH in Subjects with Heartburn — abstract, first part of the abstract (design and pH results). 1995. PMID 11854825, DOI 10.1097/00045391-199508000-00006. Read the source
3. Pediatric surgery international. Sennosides vs magnesium hydroxide vs polyethylene glycol as a treatment for constipation in anorectal malformation: a randomized crossover trial — abstract, Results section. 2025. PMID 40856829, DOI 10.1007/s00383-025-06174-9. Read the source
4. DailyMed (US National Library of Medicine), SPL from PAI Pharma (Greenville, SC, USA). Milk of Magnesia (magnesium hydroxide) oral suspension - US OTC Drug Facts label, complete content from "Drug Facts" through both "Other information" sections. Aug 20, 2026. Read the source
5. British journal of clinical pharmacology. The effect of magnesium hydroxide on the oral absorption of ibuprofen, ketoprofen and diclofenac — abstract. 1991. PMID 2054265, DOI 10.1111/j.1365-2125.1991.tb05527.x. Read the source
6. Annals of emergency medicine. Effect of magnesium hydroxide administration on iron absorption after a supratherapeutic dose of ferrous sulfate in human volunteers: a randomized controlled trial — abstract, Results section. 1999. PMID 10092717, DOI 10.1016/s0196-0644(99)70303-3. Read the source
7. Antimicrobial agents and chemotherapy. Effect of an Antacid (Aluminum Hydroxide/Magnesium Hydroxide/Simethicone) or a Proton Pump Inhibitor (Omeprazole) on the Pharmacokinetics of Tebipenem Pivoxil Hydrobromide (TBP-PI-HBr) in Healthy Adult Subjects — abstract, first part of the abstract (design and pharmacokinetic results). 2023. PMID 36943038, DOI 10.1128/aac.01495-22. Read the source
8. Pulmonary circulation. Effects of omeprazole and aluminum hydroxide/magnesium hydroxide on riociguat absorption — abstract, complete abstract (the abstract is unstructured and carries no section headings). 2016. PMID 27162626, DOI 10.1086/682228. Read the source
9. Electrolyte & blood pressure : E & BP. Fatal Hypermagnesemia in Patients Taking Magnesium Hydroxide — abstract, latter part of the abstract (the two reported cases). 2023. PMID 38152602, DOI 10.5049/ebp.2023.21.2.66. Read the source
10. Translational pediatrics. Weaning strategies for osmotic laxatives in children with functional constipation: a pilot multicenter randomized controlled trial — abstract, Conclusions section. 2025. PMID 40115451, DOI 10.21037/tp-24-436. Read the source
11. The American journal of the medical sciences. Fatal Hypermagnesemia Due to Laxative Use — abstract. 2018. PMID 29661354, DOI 10.1016/j.amjms.2017.08.013. Read the source
12. Journal of pharmaceutical health care and sciences. Risk factors for the development of hypermagnesemia in patients prescribed magnesium oxide: a retrospective cohort study — abstract, Results section. 2019. PMID 30805197, DOI 10.1186/s40780-019-0133-7. Read the source
13. Journal of clinical biochemistry and nutrition. Clinical features of hypermagnesemia in patients with functional constipation taking daily magnesium oxide — open-access full text, the Results paragraph on factors associated with hypermagnesaemia (Tables 2 and 3). 2019. PMID 31379418, DOI 10.3164/jcbn.18-117. Read the source
14. Revista de gastroenterologia de Mexico (English). Efficacy, safety, and acceptability of polyethylene glycol 3350 without electrolytes vs magnesium hydroxide in functional constipation in children from six months to eighteen years of age: A controlled clinical trial — abstract, Results section. 2023. PMID 34961695, DOI 10.1016/j.rgmxen.2021.12.005. Read the source
15. Jornal de pediatria. Comparison of the effectiveness of polyethylene glycol 4000 without electrolytes and magnesium hydroxide in the treatment of chronic functional constipation in children — abstract, Results section. 2011. PMID 21116598, DOI 10.2223/jped.2051. Read the source
16. The Journal of international medical research. Comparison of the effects of magnesium hydroxide and a bulk laxative on lipids, carbohydrates, vitamins A and E, and minerals in geriatric hospital patients in the treatment of constipation — abstract. 1989. PMID 2553511, DOI 10.1177/030006058901700506. Read the source
17. The Cochrane database of systematic reviews. Osmotic and stimulant laxatives for the management of childhood constipation (Cochrane review CD009118.pub3) — abstract, the part of the Main results section covering milk of magnesia. 2016. PMID 27531591, DOI 10.1002/14651858.CD009118.pub3. Read the source
18. The Cochrane database of systematic reviews. Osmotic and stimulant laxatives for the management of childhood constipation (Cochrane review CD009118.pub3) — abstract, Authors' conclusions section. 2016. PMID 27531591, DOI 10.1002/14651858.CD009118.pub3. Read the source
19. Revista de gastroenterologia de Mexico (English). Efficacy, safety, and acceptability of polyethylene glycol 3350 without electrolytes vs magnesium hydroxide in functional constipation in children from six months to eighteen years of age: A controlled clinical trial — abstract, Conclusions section. 2023. PMID 34961695, DOI 10.1016/j.rgmxen.2021.12.005. Read the source
20. Jornal de pediatria. Comparison of the effectiveness of polyethylene glycol 4000 without electrolytes and magnesium hydroxide in the treatment of chronic functional constipation in children — abstract, Conclusion section. 2011. PMID 21116598, DOI 10.2223/jped.2051. Read the source
21. Journal of clinical biochemistry and nutrition. Clinical features of hypermagnesemia in patients with functional constipation taking daily magnesium oxide — open-access full text, the OPENING of the single "Patient characteristics" paragraph of the Results section: patient flow and the age comparison only. It holds no serum-magnesium result; the later sentences of the same paragraph are recorded separately under mgo-crosssectional-fulltext-magnesium. 2019. PMID 31379418, DOI 10.3164/jcbn.18-117. Read the source