Magnesium: glycinate vs citrate vs malate vs threonate?

How it works

Bioavailability comparisons (Saris 2000; Blancquaert 2019; Schiopu 2022 SR; Rylander 2025) inform form choice. Glycinate absorbs via dipeptide transporter pathways, partially independent of stomach acid (relevant for older adults and PPI users). Citrate is moderately well-absorbed and has osmotic effect at higher doses. L-threonate uniquely facilitates blood-brain barrier transport via glucose transporters (Sun 2016, Neuropharmacology 108:426-439). Oxide bioavailability is lower than organic forms but the often-cited 4% figure may overstate the difference; Rylander 2025 reported around 60% in fed-condition in-vitro models.

Effective dose

Dose tier framework (see therapeutic dosing entry a768f64b): Tier 1 dietary RDA, Tier 2 standard supplementation 200-300 mg/day, Tier 3 higher-dose above 350 mg/day for specific clinical applications. Doses above 350 mg/day commonly produce GI side effects and form selection matters more at higher elemental dose; glycinate and malate have the best tolerability profiles at higher doses.

Forms compared

Direct head-to-head RCTs comparing forms for specific clinical outcomes are scarce; most form recommendations in the literature derive from bioavailability data plus mechanistic rationale rather than comparative outcome trials. The strongest comparative data is on bioavailability; comparative outcome data is limited. Glycinate and bisglycinate are essentially the same product (the bis- prefix denotes two glycine molecules per magnesium). Form-specific evidence: glycinate (Schuette 1994 in ileal resection; Walker 2003); citrate (Walker 2003; Coudray 2005; multiple BP RCTs); malate (Russell 1995 fibromyalgia, blinded portion negative; Schipane 2019 SR limited benefit); taurate (McCarty 1996 Med Hypotheses, Shrivastava 2019 rat model); L-threonate (Liu 2016 cognitive RCT, see entry b7b5a23e). Oxide bioavailability per Rylander 2025 around 60% in fed-condition in-vitro models, lower than chloride (68%) and bisglycinate (67%).

Timing

Trial durations to assess form-specific clinical response vary by application (4-8 weeks for sleep and anxiety; 12 weeks for BP and migraine prophylaxis; 12 weeks for cognitive endpoints in L-threonate trials). Most positive trials measured outcomes at the trial-defined endpoint; direct form-vs-form comparison at matched timepoints is rare.

Safety profile

All magnesium forms carry interaction considerations with bisphosphonates, fluoroquinolone and tetracycline antibiotics (timing separation required) and with magnesium-affecting diuretics. Oxide and citrate at higher doses produce more frequent osmotic diarrhoea than glycinate or malate. Hypermagnesaemia symptoms include nausea, vomiting, hypotension, and (in severe cases) cardiac and neuromuscular effects. CKD eGFR below 30 or concurrent K-sparing diuretic therapy: hypermagnesaemia risk warrants monitoring at any dose.

Special populations

Older adults and PPI users: glycinate absorption is partially independent of stomach acid, which can be advantageous in these populations. Adults with sensitive digestion: glycinate is the form with the best tolerability profile. Insulin-resistant populations: form-specific outcome data is limited; magnesium effects on insulin sensitivity have been studied across forms with no clear form-superiority finding. Athletes: form selection is less critical than total elemental dose for general repletion.

Interactions

Form-specific interaction nuances are limited. Glycinate may be marginally less affected by acid-suppression than oxide and carbonate (which require stomach acid for dissolution). Calcium at high single doses (above 500 mg) competes with magnesium absorption regardless of form. Iron and magnesium at supplemental doses can compete at high single doses.

Guideline positions

Bioavailability anchors: Saris 2000; Blancquaert 2019; Walker 2003; Coudray 2005; Schiopu 2022 SR; Rylander 2025 in-vitro. Form-specific clinical anchors: glycinate (Schuette 1994 in ileal resection); citrate in BP RCTs (Zhang 2016 Hypertension 68(2):324-333 PMID 27402922 and 2025 AHA meta-analysis update); malate in fibromyalgia (Russell 1995 J Rheumatol n=24, Schipane 2019 SR); taurate (McCarty 1996 Med Hypotheses hypothesis paper, Shrivastava 2019 rat model); L-threonate (Liu 2016 J Alzheimer's Dis 49(4):971-990 PMID 26519439). Direct head-to-head comparative outcome RCTs across forms are scarce; most recommendations derive from bioavailability data plus mechanistic rationale.

Practical framework

Consider GI tolerability: glycinate and malate at the well-tolerated end; oxide and citrate at higher doses produce more frequent osmotic effects. Consider cost: oxide cheapest, L-threonate most expensive per unit elemental magnesium. Consider acid-suppression context: glycinate works in PPI users where oxide and carbonate may underperform. This is a summary of published research, not personal health advice. Discuss any health or supplement decisions with a qualified healthcare professional, particularly during ongoing care, pregnancy, or with chronic conditions.

Common misconceptions

Claim: glycinate is dramatically superior to other forms for all applications. Direct head-to-head RCTs comparing forms for specific outcomes are scarce; the comparative bioavailability data is the strongest comparative evidence. Glycinate is well-tolerated and a reasonable default but not necessarily superior for outcomes where other forms have specific trial evidence (e.g., citrate in BP, oxide in migraine prophylaxis per Karimi 2019).

Claim: malate has strong evidence for fibromyalgia. The Russell 1995 blinded portion did not show significant benefit; Schipane 2019 SR concluded little or no difference.