Toggle Health Problems and D

ICU patients with low Magnesium were 2X more likely to die – June 2014

Significance of serum magnesium levels in critically ill-patients

Int J Appl Basic Med Res. 2014 Jan-Jun; 4(1): 34–37; doi: 10.4103/2229-516X.125690
PMCID: PMC3931212
Mir Sadaqat Hassan Zafar, Javaid Iqbal Wani, Raiesa Karim,1 Mohammad Muzaffer Mir,2 and Parvaiz Ahmad Koul
Department of Internal and Pulmonary Medicine, Sher-i-Kashmir Institute of Medical Sciences, Srinagar, Jammu and Kashmir, India
1Department of Social and Preventive Medicine, Government Medical College, Srinagar, Jammu and Kashmir, India
2Department of Clinical Biochemistry, Sher-i-Kashmir Institute of Medical Sciences, Srinagar, Jammu and Kashmir, India
Address for correspondence: Dr. Mir Sadaqat Hassan Zafar, Department of Hematology, Room No. 3, First Floor, SSRB, Sir Ganga Ram Hospital, New Delhi, India. E-mail: moc.liamg at taqadas.rim


ICU death 2X more likely if low Magnesium – March 2015 very similar results
Unfortunately Mg studies do not report on Vitamin D levels and visa versa

Image Image

Magnesium is one of the major electrolytes, deficiency of which is frequently overlooked in critical illness, leading to an adverse clinical outcome if not monitored regularly.

Settings and Design:
Single center prospective observational study of 2 years duration.

Materials and Methods:
The subjects studied were monitored for serum magnesium levels 2 times: Day 1 and day 4 of intensive care unit (ICU) admission. Patients were divided into normomagnesemic and hypomagnesemic groups and compared for various parameters.

Out of 70 critically ill-patients, 50 patients (71.43%) were normomagnesemic, 17 patients (24.29%) were hypomagnesemic and three patients were hypermagnesemic. The stay of the patients in ICU (P > 0.05), Acute Physiology and Chronic Health Evaluation-II (APACHE-II) scoring (P = 0.34) and co-morbidity (P = 0.360) showed an insignificant variation between the two groups. Associated electrolyte abnormalities in hypomagnesemic patients were hypokalemia (58.82%), hyponatremia (47.05%), hypocalcemia (70.58%) and hypophosphatemia (29.41%). About 76.47% of hypomagnesemic population was on magnesium lowering drugs while as 46% of normomagnesemic population was on magnesium lowering drugs (P = 0.030). Mortality of hypomagnesemic group was 74.47% while that of normomagnesemic group was 36% (P = 0.004).

Hypomagnesemia is a significant electrolyte abnormality in critically ill-patients. Critically ill hypomagnesemic patients have higher mortality than the normomagnesemic patients.

Keywords: Critically ill-patients, hypomagnesemia, normomagnesemia, outcome


Magnesium is fourth most common cation in the body and second most common intracellular cation after potassium, yet its deficiency in critically ill-patients is frequently overlooked.[1,2,3] Various studies have reported the incidence of hypomagnesemia up to 65% in critically ill-patients.[4] Although many paradigms have been explored to minimize the mortality in critical care units, magnesium loss has been scarcely addressed; in this respect leading to inconclusive results. Serum magnesium monitoring may have prognostic and perhaps therapeutic implications because critically ill-patients are predisposed to both symptomatic or asymptomatic magnesium deficiency that can lead to some important clinical consequences (such as hypokalemia, cardiac arrhythmias, hypocalcemia, neurotoxicity and psychiatric problems), ultimately increasing the morbidity and mortality.[5] There is a paucity of data in Indian literature, addressing this common, but underdiagnosed electrolyte deficiency. Present study was undertaken against this backdrop at a tertiary care teaching hospital to assess the magnitude of magnesium deficiency and its influence on the outcome of critically ill-patients so that baseline reference data for insight and management of the problem in routine cases in our intensive care units (ICUs) is formed.

Materials and Methods

This study was undertaken in a tertiary care teaching hospital of north India, over a period of 2 years. This hospital has independently functioning ICUs, both medical as well as surgical. The study was started after getting clearance from the Ethical Committee, which is an independently functioning body of this institute. A detailed protocol was presented and discussed with the committee before the start of the study. It was a prospective observational study, which involved patients from both the medical and surgical ICUs. Patients were enrolled at random and the selection bias was avoided by registering those patients who fulfilled the criteria for critical illness and who were newly admitted to ICU. Thus, the condition of subjects was first demarcated as critical who fulfilled the criterion of minimal physiologic reserve, requiring life support until abnormalities have been reversed, undergone massive disruption of physiology, requiring intensive monitoring in anticipation of possible aggressive intervention or post-operative cares. Those patients were included who had ≥1 organ failure (acute respiratory failure, acute liver failure, acute renal failure, septicemia, acute myocardial infarction, cerebrovascular accidents or multiorgan failure). Post-operative (especially major thoracoabdoninal surgery) and polytrauma patients were also included. Only those patients were included in which it was possible to take day 1 sample for serum magnesium. Critically ill-patients in which day 1 of ICU stay was missed were excluded. Patients who were stable and did not fulfill the criteria for ICU admission or who were in the general wards of the hospital were excluded. Furthermore, the patients who had documented hypomagnesemia before the ICU admission or who were on magnesium supplementation were excluded. Patients were enrolled for the study after informed consent was given by first-degree relatives of critically ill-patients in the presence of two witnesses. Demographic data (age and sex), medical history, surgical history, medications administrated and length of ICU stay were recorded for each patient. The severity scoring system used was Acute Physiology and Chronic Health Evaluation-II (APACHE-II) which includes acute physiology score, Glasgow coma scale, points assigned to age with chronic disease and chronic health condition score. Other investigations such as arterial blood gases, complete blood count, kidney function tests, liver function tests, serum electrolytes, electrocardiography, chest X-ray and ultrasonography were also done. Blood, urine and endotracheal secretions were sent for culture screen in every febrile patient. Computed tomography and magnetic resonance imaging scanning was performed wherever it was required.

About 1 ml of venous blood was taken after informed consent from the patient's first degree relative on day 1 and day 4 of ICU admission and immediately used for sera separation. The blood was collected in green top vacutainers and made to stand for 30 min until centrifugation on 3000 rpm in a centrifuge (Sigma USA). The clear upper layer was isolated and freezed at – 20°C until assayed. The concentration of serum magnesium in the samples was determined by magnesium fluid monoreagent (Centronic GmbH-Germany), a colorimetric endpoint method. The sensitivity of this method has shown a correlation with atomic absorption spectroscopy. Two ml of monoreagent was added to 20 μl of sample. The contents were vortex shaked and incubated at 25°C for 5 min. The colored complex formed by magnesium with the monoreagent was read on a spectrophotometer (Shimadzu, Japan) in 1 ml cuvettes at 545 nm against reagent blank. The results were expressed as mg/dl. Normal range for magnesium concentration in our laboratory was 1.70-2.55 mg/dl. Subjects were divided into normomagnesemic and hypomagnesemic groups and compared for various parameters. Serum magnesium level of <1.70 mg/dl was regarded as hypomagnesemia and ≥2.56 mg/dl as hypermagnesemia.

Statistical analysis

All the data were expressed in Mean ± standard deviation (SD). Analysis was performed by using test statistic Student's t-test for the difference of means, Chi-square test, Fisher's exact test and correlation. These tests were further referenced for P values for their significance and P < 0.05 was considered to be significant.


We studied 70 critically ill-patients (age > 16 years) and followed their clinical as well as biochemical parameters from the day of admission to ICU to the day of discharge from ICU or death. Septicemia, multiorgan dysfunction, respiratory failure and patients with major thoraco-abdominal surgery formed more than 50% of our study population [Table 1]. Of 70 critically ill-patients, 71.43% were normomagnesemic, 24.29% were hypomagnesemic and 4.28% were hypermagnesemic. Mean ± SD of serum magnesium of hypomagnesemic patients was 1.63 ± 0.06 mg/dl while as mean ± SD of serum magnesium of normomagnesemic patients was 2.03 ± 0.22 mg/dl. Out of 17 critically ill hypomagnesemic patients, 13 patients had hypomagnesemia on day 1 and day 4 of ICU admission while as four patients had hypomagnesemia on day 1 and normomagnesemia on day 4 of ICU admission. There was no patient who had normomagnesemia on day 1 and hypomagnesemia on day 4 of ICU admission. Hypomagnesemic patients mostly comprised of multiorgan dysfunction (41.17%), respiratory failure (17.64%) and septicemia (11.76%) while as normomagnesemic patients had mostly septicemia (20%), post-operative course (16%), respiratory failure (14%), renal failure (10%) and acute myocardial infarction (10%). Out of three hypermagnesemic patients, two had renal failure and one had respiratory failure. Age of hypomagnesemic patients ranged from 21 to 65 years (males 10 and females 7) and that of normomagnesemic patients ranged from 16 to 80 years (males 37 and females 13). Underlying chronic illnesses had no statistical significance between hypomagnesemic and normomagnesemic patients. About 52.94% of hypomagnesemic population had a history of underlying chronic illness while as 44% of normomagnesemic population had such history (P = 0.360).

Mean APACHE-II score of hypomagnesemic patients was 23.82 ± 5.90 while that of normomagnesemic patients was 22.26 ± 4.37 (P = 0.34). In addition, mean ICU stay [Table 2] in our study was 5.53 ± 4.32 and 5.66 ± 3.99 days for hypomagnesemic and normomagnesemic groups respectively (P > 0.05). About 2/3rd (76.47%) of hypomagnesemic population was on magnesium lowering drugs (mannitol, aminoglycosides and furosemide) while as 46% of normomagnesemic population was on magnesium lowering drugs (P = 0.030). Other associated findings were the electrolyte abnormalities such as hypokalemia (58.82% of hypomagnesemic patients), hyponatremia (47.05% of hypomagnesemic patients), hypocalcemia (70.58% of hypomagnesemic patients) and hypophosphatemia (29.41% of hypomagnesemic patients). There was a highly significant difference in mortality between hypomagnesemic and normomagnesemic patients [Table 3]. In hypomagnesemic patients, 76.47% (13/17) expired, while as out of normomagnesemic patients, 36% (18/50) expired (P < 0.05).


Magnesium is mostly located in bone or within the cells. Assessment of magnesium status in either of these compartments in critical illness is impractical. The physician must therefore rely on determination of serum magnesium to determine if a patient is magnesium deficient.[6] Occurrence of magnesium deficiency in critical illness correlates with higher morbidity and mortality. The primary end points of this study were serum magnesium level and outcome of critically ill-patients. Secondary outcome measures were the associated electrolyte disturbances and length of ICU stay. In our study of 70 critically ill-patients, 17 (24.29%) patients were hypomagnesemic, 50 (71.23%) were normomagnesemic and 3 (4.28%) were hypermagnesemic, signifying remarkable occurrence of hypomagnesemia in ICU patients. Magnesium deficiency has been found to co-exist in up to 40% of patients with other electrolyte abnormalities. Magnesium has a marked effect on the regulation of transmembrane sodium and potassium movement. There are multiple reasons for magnesium deficiency in critical care settings e.g., decreased absorption caused by impaired gastrointestinal activity, malnutrition, renal wasting of various drugs (e.g., digoxin, gentamicin, loop diuretics etc.), diabetes mellitus, hypokalemia and hypocalcemia.[7] Genetic differences in magnesium utilization may account for differences in vulnerability to magnesium deficiency and differences in body responses to stress.[8,9] In our study, we found 76.47% patients of hypomagnesemia were on magnesium lowering drugs while as 46% of normomagnesemic patients were on magnesium lowering drugs. The difference was statistically significant (P = 0.030). Diuretics acting at loop of Henle such as furosemide have been shown by micropuncture studies to result in marked magnesium deficiency. Aminoglycosides have been shown to cause reversible renal lesion that results in hypermagnesuria and hypomagnesemia.[10] In a retrospective analysis of 179 children of a pediatric ICU, it was found that hypomagnesemia was more often seen in patients receiving diuretics and aminoglycosides.[11] Proton pump inhibitor associated hypomagnesemia is a rare, but potentially life-threatening side-effect that has emerged only in the era of mass use of these agents.[12]

The relationship between hypomagnesemia and mortality rate varies from study to study. A higher mortality rate was detected in hypomagnesemia patients when compared with normomagnesemic patients by Chernow et al.[13] (41% vs. 13%), Rubeiz et al.[14] (46% vs. 25%) and Safavi and Honarmand[15] (55% vs. 35%). Guérin et al.[16] had found no difference in ICU mortality between hypomagnesemic and normomagnesemic groups (18% vs. 17%); but noted a higher mortality rate among hypermagnesemic patients. Soliman et al.[17] observed that patients who develop ionized hypomagnesemia during their ICU stay have higher mortality rates (2-3 times higher). Dabbagh et al.[18] observed higher mortality rates in critically ill-patients with daily magnesium supplementation index (DMSI) <1 g/day in comparison to DMSI > 1 g/day (43.5% vs. 17%). Limaye et al.[19] observed that mortality rate in hypomagnesemic group was 57% when compared with 31% in the normomagnesemic group. Present study showed the mortality rate of hypomagnesemic patients 76.47% as compared to normomagnesemic patients (36%), almost similar results as previously reported. The higher mortality in our study can be ascribed to higher incidences of other electrolyte deficiencies and multiorgan dysfunction in the hypomagnesemic group when compared with normomagnesemic group.


Hypomagnesemia is a frequent electrolyte deficiency in critically ill-patients. Vital predictors of this deficiency are hypokalemia or hypocalcemia and considerable predisposing factor is the use of magnesium lowering drugs. There is an unfavorable outcome of critically ill hypomagnesemic patients irrespective of insignificant ICU stay and APACHE-II score. This study highlights the role of magnesium monitoring in critical illness and its value for favorable outcome. It adds to the scarce Indian data regarding magnesium homeostasis in ICUs. Our study has a limitation of smaller size of study population to comment on similar ICU stay, APACHE-II score and co-morbidity; for that a prospective study with larger sample size is needed.

The authors would like to thank patients for their participation in the study. Authors are grateful to the technical staff of clinical biochemistry and resident staff of the critical care unit.


  • 1. Cortés YE, Moses L. Magnesium disturbances in critically ill patients. Compend Contin Educ Vet. 2007;29:420–7. [PubMed: 17727050]
  • 2. Quamme GA, de Rouffignac C. Epithelial magnesium transport and regulation by the kidney. Front Biosci. 2000;5:D694–711. [PubMed: 10922295]
  • 3. Fawcet WJ, Haxby EJ, Male DA. Magnesium physiology and pharmacology. Br J Anesth. 1999;83:302–20.
  • 4. Deheinzelin D, Negri EM, Tucci MR, Salem MZ, da Cruz VM, Oliveira RM, et al. Hypomagnesemia in critically ill cancer patients: A prospective study of predictive factors. Braz J Med Biol Res. 2000;33:1443–8. [PubMed: 11105096]
  • 5. Tong GM, Rude RK. Magnesium deficiency in critical illness. J Intensive Care Med. 2005;20:3–17. [PubMed: 15665255]
  • 6. Herroeder S, Schönherr ME, De Hert SG, Hollmann MW. Magnesium – Essentials for anesthesiologists. Anesthesiology. 2011;114:971–93. [PubMed: 21364460]
  • 7. Lee JW. Fluid and electrolyte disturbances in critically ill patients. Electrolyte Blood Press. 2010;8:72–81. [PMCID: PMC3043756] [PubMed: 21468200]
  • 8. Whitted AD, Stanifer JW, Dube P, Borkowski BJ, Yusuf J, Komolafe BO, et al. A dyshomeostasis of electrolytes and trace elements in acute stressor states: Impact on the heart. Am J Med Sci. 2010;340:48–53. [PubMed: 20610973]
  • 9. Weber KT, Bhattacharya SK, Newman KP, Soberman JE, Ramanathan KB, McGee JE, et al. Stressor states and the cation crossroads. J Am Coll Nutr. 2010;29:563–74. [PubMed: 21677120]
  • 10. Buckley MS, Leblanc JM, Cawley MJ. Electrolyte disturbances associated with commonly prescribed medications in the intensive care unit. Crit Care Med. 2010;38:S253–64. [PubMed: 20502178]
  • 11. Haque A, Saleem AF. On admission hypomagnesemia in critically ill children: Risk factors and outcome. Indian J Pediatr. 2009;76:1227–30. [PubMed: 19936657]
  • 12. Cundy T, Mackay J. Proton pump inhibitors and severe hypomagnesaemia. Curr Opin Gastroenterol. 2011;27:180–5. [PubMed: 20856115]
  • 13. Chernow B, Bamberger S, Stoiko M, Vadnais M, Mills S, Hoellerich V, et al. Hypomagnesemia in patients in postoperative intensive care. Chest. 1989;95:391–7. [PubMed: 2914492]
  • 14. Rubeiz GJ, Thill-Baharozian M, Hardie D, Carlson RW. Association of hypomagnesemia and mortality in acutely ill medical patients. Crit Care Med. 1993;21:203–9. [PubMed: 8428470]
  • 15. Safavi M, Honarmand A. Admission hypomagnesemia - Impact on mortality or morbidity in critically ill patients. Middle East J Anesthesiol. 2007;19:645–60. [PubMed: 18044292]
  • 16. Guérin C, Cousin C, Mignot F, Manchon M, Fournier G. Serum and erythrocyte magnesium in critically ill patients. Intensive Care Med. 1996;22:724–7. [PubMed: 8880238]
  • 17. Soliman HM, Mercan D, Lobo SS, Mélot C, Vincent JL. Development of ionized hypomagnesemia is associated with higher mortality rates. Crit Care Med. 2003;31:1082–7. [PubMed: 12682476]
  • 18. Dabbagh OC, Aldawood AS, Arabi YM, Lone NA, Brits R, Pillay M. Magnesium supplementation and the potential association with mortality rates among critically ill non-cardiac patients. Saudi Med J. 2006;27:821–5. [PubMed: 16758043]
  • 19. Limaye CS, Londhey VA, Nadkart MY, Borges NE. Hypomagnesemia in critically ill medical patients. J Assoc Physicians India. 2011;59:19–22. [PubMed: 21751660]

Attached files

ID Name Comment Uploaded Size Downloads
5540 ICU Mg.jpg admin 30 May, 2015 70.62 Kb 2918
5539 ICU Mg T1.jpg admin 30 May, 2015 60.58 Kb 1360