RESUMEN
INTRODUCTION: Recent evidence suggests that magnesium may be neuroprotective in the setting of cerebral ischemia, and therapeutic magnesium infusion has been proposed for prophylaxis and treatment of delayed ischemic neurological deficit (DIND) resulting from vasospasm in patients with aneurysmal subarachnoid hemorrhage (SAH). We studied the association between serum magnesium levels, the development of DIND, and the outcomes of patients with SAH. METHODS: We studied 128 consecutive patients with aneurysmal SAH treated at our institution between 1990 and 1997 who had a serum magnesium level measured at least once during the acute phase of their hospitalization. Delayed ischemic neurological deficit was defined as severe (major focal deficit or coma), moderate (incomplete focal deficit or decreased sensorium without coma), or none. RESULTS: There was no significant difference in mean, minimum, or maximum serum magnesium levels between patients with and without DIND (1.93, 1.83, 2.02 versus 1.91, 1.84, 1.97 mg/dL, respectively). Similarly, no difference was found in mean serum magnesium levels among patients with severe (1.94 mg/dL), moderate (1.92 mg/dL), or no DIND (1.91 mg/dL). Analyses of serum magnesium levels before (0-4 days following SAH), during (4-14 days following SAH), and after (greater than 14 days following SAH) the period of highest risk for vasospasm revealed no association with the development or severity of DIND. Permanent deficit or death resulting from vasospasm and Glasgow Outcome Scale score at longest follow-up were similarly unaffected by serum magnesium levels overall or during any time interval. Forty (31.5%) patients were hypomagnesemic (less than 1.7 mg/dL) during hospitalization, but no difference in outcome (p = 0.185) or development of DIND (p = 0.785) was found when compared to patients with normal (1.7-2.1 mg/dL) or high (greater than 2.1 mg/dL) magnesium serum levels. CONCLUSION: We identified no relationship between serum magnesium levels and the development of DIND or outcome following aneurysmal SAH. Based on these data, magnesium supplementation to normal or high-normal physiological ranges seems unlikely to be beneficial for DIND resulting from vasospasm. However, no inference can be made regarding the value of therapeutic infusion of magnesium to supraphysiological levels.