Intravenous Magnesium Sulfate Does Not Increase Ventricular
CSF Ionized Magnesium Concentration of Patients with
Intracranial Hypertension
Randall P. Brewer, Augusto Parra, Cecil O. Borel, Michael B. Hopkins, and James D. Reynolds
Department of Anesthesiology, Duke University Medical Center, Durham, North Carolina, USA
Summary: Magnesium sulfate has attracted interest as a potential neuroprotectant but
passage of magnesium ion into the central nervous system has not been well docu-
mented. For this study, we quantified plasma and cerebrospinal fluid (CSF) ionized
magnesium concentration after systemic magnesium sulfate infusion in patients with
intracranial hypertension. Patients (N = 9) received an intravenous infusion of 5 g/20
mmol magnesium sulfate (125 mL of a 4% wt/vol solution) over 30 minutes. Before
and after dosing, CSF (from an indwelling ventricular catheter) and blood samples were
collected at hourly intervals. Ionized magnesium concentration in all samples was de-
termined using an electrolyte analyzer. Baseline plasma and CSF ionized magnesium
concentrations were 0.58 ± 0.05 and 0.82 ± 0.06 mmol/L, respectively. Intravenous
magnesium sulfate infusion significantly increased plasma ionized magnesium concen-
tration (peak, 0.89 ± 0.11 mmol/L), but CSF magnesium levels did not change during the
4-hour study. Systemic administration of magnesium sulfate failed to increase CSF ion-
ized magnesium concentration in patients with intracranial hypertension despite increas-
ing plasma magnesium levels by >50%. Key Words: Cerebrospinal fluid—Intracranial
hypertension—Magnesium
Magnesium ion (Mg) is an essential cofactor for
many enzymatic reactions and an important regulator
of excitatory neurotransmission in the central nervous
system (CNS). It is the latter role that has produced
considerable interest in the use of Mg as a pharmaco-
logic neuroprotective agent (1). To that end, assessing
the human therapeutic potential of Mg in cases of
stroke and traumatic brain injury is the primary goal of
several randomized clinical trials (e.g., [2]).
One of the key tenets for an effective neuroprotectant
is the ability to enter the CNS. Despite this necessity,
few studies have been conducted to determine whether
central Mg levels actually increase after dosing. After
systemic administration, usually in the form of magne-
sium sulfate (MgSO
4
), blood Mg is present in three
forms: 27%–34% is protein-bound; 8%–12% is com-
plexed to small anions; and the remainder is in the free,
hydrated ion form, Mg
2+
(3). While clinical laborato-
ries generally measure total Mg concentration, it is im-
portant to realize that only the ionized or “free” fraction
of Mg in the blood is physiologically active (4). By ex-
tension, only the ionized form of Mg has the potential
to enter the CNS. However, its divalent state does sug-
gest that such passage could be hindered.
Previous investigations that quantitated central Mg
concentration after systemic administration of a Mg
salt have yielded equivocal results. Hallak et al. (5) re-
ported increased Mg in both rat cerebrospinal fluid
(CSF) and brain 2 hours after MgSO
4
injection,
whereas Kim et al. (6) reported no change in rat brain
parenchymal Mg concentration after 5 days of MgSO
4
infusion, this despite observing almost a threefold in-
crease in plasma Mg. In humans, preeclamptic patients
receiving MgSO
4
therapy had higher CSF total Mg lev-
els compared to control parturients (7), although infu-
sion of MgSO
4
during craniotomy, which increased
plasma Mg concentration by 380%, only produced a
15% increase in CSF Mg (8).
Address correspondence and reprint requests to James D. Reynolds,
Division of Women’s Anesthesia, Departments of Anesthesiology and
Surgery, Box 3094, Duke University Medical Center, Durham, NC
27710, USA.
Clinical Neuropharmacology
Vol. 24, No. 6, pp. 341–345
© 2001 Lippincott Williams & Wilkins, Inc., Philadelphia
341