Table 3. Summaries of magnesium studies reporting anxiety-related outcomes.
GAD, generalized anxiety disorder; MHQ, menstrual health questionnaire; HAM-A, Hamilton Anxiety Rating Scale; DASS, Depression, Anxiety, Stress Scales; OCPs, oral contraceptive pills; RCADS, Revised Child Anxiety and Depression Scale; HADS, Hospital Anxiety and Depression Scale; CGI, clinical global impression; STAI, Spielberger State-Trait Anxiety Inventory
| Authors | Country | Study design | Participants | Inclusion/exclusion criteria | Intervention | Outcome measures | Relevant results | Positive/negative | ||||||
| N | Population | Inclusion | Exclusion | Magnesium form | Dose | Additional therapies | Duration | Description | ||||||
| Saba et al. (2022) [20] | Iran | Single-blind controlled trial | 60; control group N=30, tx group N=30 | Hospitalized adults undergoing open heart surgery | Age <70 years old; candidate for elective CABG surgery | Hx of atrial fibrillation prior to CABG surgery; hx of liver or renal failure; hx of stroke or recent TIA; postoperative respiratory failure; liver or kidney failure; emergency surgery during the study period; chronic diarrhea; allergy to study drug; hx of sleep disorders; hx of anxiety or depression; parenteral MgSO4 tx | MgO | 500 mg | No | Five days | 500 mg MgO for five days (two pills of 250 mg MgO each) | Anxiety and depressive sx as assessed by the HADS | The mean HADS score was significantly lower in the treatment group versus the control group (P=0.007) | P |
| Kovacevic et al. (2017) [32] | Serbia | Prospective cohort study | 32 | Pediatric migraine patients without comorbidities | Ages 7-17; dx of migraine at least one month prior; no previous migraine prophylaxis | Presence of any other medical, neurological, or psychiatric disorder | MgO; magnesium glycinate | 4-6 mg/kg/day | NSAIDs for acute migraine attacks during the study period | Six months | MgO or magnesium glycinate, dosed at 4-6 mg/kg/day for six months | Anxiety and depressive sx as assessed by the RCADS | Self-reported anxiety scores decreased significantly between baseline and six months (P=0.001); scores between baseline and three months trended lower but did not achieve significance (P=0.115) | P |
| De Souza et al. (2000) [33] | United Kingdom | Double-blind randomized controlled crossover trial | 44 (crossover trial; each served as own control) | Women with premenstrual anxiety sx | Premenstrual scores >30% higher than post-menstrual sx as measured by the MHQ | Patients already taking any vitamin or mineral supplement; any medication aside from OCPs | MgO | 200 mg | Pyridoxine 50 mg | Five menstrual cycles | Crossover to alternate treatment during each subsequent menstrual period (1) MgO 200 mg, (2) MgO + pyridoxine 50 mg, (3) pyridoxine 50 mg, (4) placebo | Self-reported, subjective anxiety sx reported on a five-point ordinal scale; 24-hour magnesium and creatinine output | Significantly lower anxiety scores during treatment with a combination of magnesium and B6 compared with other treatments (P=0.04); urinary magnesium did not differ significantly between groups | Mixed |
| Oddoux et al. (2022) [34] | France | Prospective cohort study | 93 | Adults aged 18-70 with adjustment disorder with anxiety, at least mild-moderate in severity | Adults aged 18-70 years; HAM-A score >20 | Previous treatment for anxiety within the prior three months (including psychotherapy); the presence of any other mental illness; anxiety lasting >3 months; history of substance abuse; >10 mg/kg of caffeine/day; >1 pack of cigarettes/day; pregnancy; breastfeeding; severe medical comorbidity | MgO and magnesium bisglycinate | 300 mg (270 mg as MgO and 30 mg as magnesium bisglycinate) | 200 mg of fish protein hydrolysate ("Gabolysat®") and 1.4 mg of vitamin B6 (as pyridoxal chlorate) | Four weeks | Four weeks of 300 mg magnesium, 200 mg fish protein hydrolysate, and 1.4 mg B6 combo, taken with breakfast | % of patients with ≥50% decrease in HAM-A score; change in HAM-A score; change in Clinical Global Index Scale | 41.9% (39/93) experienced a ≥50% decrease in HAM-A score (primary endpoint); mean HAM-A score decreased by 12.1 +/- 5.7 points (P<0.001); 75.3% improved significantly or very significantly on CGI scale | P |
| Fard et al. (2017) [35] | Iran | Triple-blind RCT | 99; control group N=33, each tx group N=33 | Postpartum women having given birth within 48 hours | ≥18 years old; living in Tabriz, Iran; low-risk pregnancy | Depression (12 or greater on the Edinburgh Scale); any psychiatric history; complications with childbirth; NICU admission or death in the infant; significant stressful event (divorce, hospitalization, or death of relative); chronic illness; liver or kidney disease; hx of infertility or previous miscarriage | MgSO4 | 320 mg (64.6 elemental magnesium) | None | Eight weeks | Eight weeks of either placebo, 27 mg of ZnSO4, or 320 mg of MgSO4 | STAI; Edinburgh Postnatal Depression Scale | No significant differences were found in depression and anxiety scores between the placebo, zinc, and magnesium groups | N |
| Noah et al. (2020) [36] | France | Single-blind RCT | 264 control group N=132, tx group N=132 | Adults aged 18-50 experiencing moderate-severe stress | Ages 18-50 years old; moderate to severe stress as measured by >18 on the DASS stress subscale; low magnesium status as defined by serum values of 0.66-0.84 mmol/L | Use of levodopa, quinidine, and proton-pump inhibitors within the three months prior to screening; alcohol intake of >3 drinks per day; hx of substance abuse; type 1 and type 2 diabetes mellitus; moderate or severe kidney disease; severe hypomagnesemia (serum magnesium of <0.45 mmol/L) | Magnesium lactate dihydrate | 300 mg elemental magnesium total (across six tablets) | 30 mg of pyridoxine (in tx group; just magnesium lactate alone as control) | Eight weeks | Eight weeks of either magnesium lactate 100 mg AM, 100 mg at noon, and 100 mg PM (control group) or eight weeks of magnesium lactate-B6 100 mg-10 mg AM, 100 mg-10 mg at noon, and 100 mg-10 mg PM (tx group) | DASS depression and anxiety subscales | DASS anxiety subscale scores decreased significantly from baseline in both the magnesium and magnesium+B6 groups (P<0.05); no significant difference between groups was found | P |
| Hanus et al. (2003) [37] | France | Double-blind RCT | 264 control group N=134, tx group N=130 | Patients with mild-moderate GAD | Adults >18 years old with mild-moderate GAD | Other axis 1 disorders; treatment with psychotropic medication, sedatives, or magnesium within one month of study | MgO | 248.7 mg MgO (150 mg elemental magnesium) | Extracts of the plants: Crataegus oxyacantha and Eschscholtzia californica | 90 days | Two tablets of 124.35 mg MgO and plant extracts of Crataegus oxyacantha and Eschscholtzia californica (75 mg elemental magnesium per tablet) before AM and PM meal. 300 mg MgO total daily | Anxiety as assessed by the HAM-A; Patient self-reported anxiety scores (VAS 0-100 scale, 100=very anxious); % responsive (>50% reduction in sx); physician’s CGI | Total HAM-A score decreased significantly more in the tx group versus placebo (-10.6±1.2 versus -8.9±1.2, P=0.005); VAS subjective anxiety score decreased significantly more in tx group versus placebo (-38.5 versus -29.2, P=0.005); significantly greater response rate in tx group versus placebo for both the HAM-A and the VAS scores: 45% versus 32%, P=0.017 and 58% versus 43%, P=0.008, respectively | P |