Background The pharmacokinetic basis of magnesium sulphate
(MgSO4) dosing regimens for eclampsia prophylaxis and
treatment is not clearly established.
Objectives To review available data on clinical pharmacokinetic
properties of MgSO4 when used for women with pre-eclampsia
and/or eclampsia.
Search strategy MEDLINE, EMBASE, CINAHL, POPLINE, Global
Health Library and reference lists of eligible studies.
Selection criteria All study types investigating pharmacokinetic
properties of MgSO4 in women with pre-eclampsia and/or eclampsia.
Data collection and analysis Two authors extracted data on basic
pharmacokinetic parameters reflecting the different aspects of
absorption, bioavailability, distribution and excretion of MgSO4
according to identified dosing regimens.
Main results Twenty-eight studies investigating pharmacokinetic
properties of 17 MgSO4 regimens met our inclusion criteria.
Most women (91.5%) in the studies had pre-eclampsia. Baseline
serum magnesium concentrations were consistently <1 mmol/l
across studies. Intravenous loading dose between 4 and 6 g was
associated with a doubling of this baseline concentration half an
hour after injection. Maintenance infusion of 1 g/hour
consistently produced concentrations well below 2 mmol/l,
whereas maintenance infusion at 2 g/hour and the Pritchard
intramuscular regimen had higher but inconsistent probability of
producing concentrations between 2 and 3 mmol/l. Volume of
distribution of magnesium varied (13.65–49.00 l) but the plasma
clearance was fairly similar (4.28–5.00 l/hour) across populations.
Conclusion The profiles of Zuspan and Pritchard regimens
indicate that the minimum effective serum magnesium
concentration for eclampsia prophylaxis is lower than the
generally accepted level. Exposure–response studies to identify
effective alternative dosing regimens should target concentrations
achievable by these standard regimens.

Intravenous regimens
4-g loading dose and 1 g/hour continuous maintenance
infusion (Zuspan regimen)
Seven studies reported pharmacokinetic data based on the
Zuspan regimen. The reported baseline levels were all <1.00 mmol/l with mean values ranging between
0.74 and 0.85 mmol/l (Table 2). Following the loading
dose, serum magnesium levels rose sharply to about twice
the baseline levels at ½ hour (1.48–1.70 mmol/l). At 1, 2,
and 4 hours of the maintenance dose, the mean serum
levels remained at a fairly constant level that was consistent
with the values attained at ½ hour. The serum levels at 8,
12, and 24 hours also remained within the same range and
at no point did the mean serum concentration level
reached 2.00 mmol/l. The described serum magnesium
levels between ½ hour and 24 hours following initiation of
treatment was consistent with the steady-state level of
1.64 mmol/l and an ‘average concentration’ of 1.70 mmol/l
reported by two studies.5,17 One study showed that the
peak serum concentration was achieved within half an hour
of treatment.24 The apparent volume of distribution estimated
by two model-based studies varied considerably
between the populations studied—15.60 l in a population
of Indian women compared with 32.20 l in Australian
women.17,27 However, the estimated plasma clearance was
fairly consistent across the two populations—4.81 and
4.28 l/hour, respectively. One study estimated the half-life
of MgSO4 to be 5.2 hours.17 No other pharmacokinetic
parameters were reported in the included studies.
4-g loading dose and 2 g/hour continuous maintenance
Five studies reported pharmacokinetic data based on the
use of this regimen.5,8,22,32,33 Following administration of
the loading dose, serum magnesium concentration rose
rapidly to double the baseline values by ½ hour (1.73–
2.25 mmol/l). Data on mean serum concentrations at 1, 2,
4, 8, and 12 hours from the start of infusion showed a
gradual rise in serum magnesium to a plateau level, with
mean levels slightly above 2.00 mmol/l being more consistent
after 4 hours. The fluctuations in the serum magnesium
levels were minimal and the described pattern was
consistent with the steady state concentrations of 1.84 and
1.99 mmol/l, as reported by two studies.5,32
Based on this regimen, one study estimated the volume
of distribution of magnesium to be 16.40 l, plasma clearance
to be 1.21 l/hour and elimination half-life to be
20.2 hours.32 Three studies reported considerable variations
in the ionised (free) magnesium fraction at baseline and
during maintenance infusion.8,22,32 At baseline, the ionised
fraction was between 50.0 and 64.9% of the total serum
magnesium but these fractions appeared to decrease as the
serum level approached steady-state levels. Two of these
studies demonstrated no correlation between ionised and
total magnesium,8,32 whereas the third study reported a
strong correlation between ionised and total magnesium.22
This finding was supported by a report of a positive correlation
of ionised and total magnesium in pre-eclamptic
women in another study.36 No other pharmacokinetic
parameters were reported in the included studies.
5-g loading dose and 1 g/hour continuous maintenance
Based on this regimen, one study reported that serum magnesium
rose rapidly from baseline level of 0.95 mmol/l to
1.97 mmol/l by ½ hour, followed by a gradual decline by
1 hour before rising slowly again to steady between 2.20
and 2.42 mmol/l between 12 and 24 hours of maintenance
6-g loading dose and 2 g/hour continuous maintenance
infusion (Sibai regimen)
Six studies provided sparse pharmacokinetic data based
on this regimen.13,16,21,25,30,34 The reported baseline
serum magnesium values were between 0.58 and 0.80
mmol/l.16,21,25,30 Following initiation of MgSO4, one study
showed that this level doubled in ½ hour before declining
slightly to plateau between 1.70 and 1.80 mmol/l between
1 hour and 12 hours of maintenance infusion.16 At no
point during the treatment did the serum level reach
2.00 mmol/l (peak concentration of 1.96 mmol/l was
attained at 0.90 hour). Another study recorded a similar
pattern of rapid rise and fall in serum magnesium following
the loading dose but the mean levels gradually
increased to achieve levels above 2.00 mmol/l between 12
and 24 hours.30 Two other studies also reported ‘average’
values of 2.05 and 2.27 mmol/l during administration of
this regimen.13,34
Two studies reported cerebrospinal fluid (CSF) magnesium
levels of 1.23–1.34 mmol/l.13,34 One study reported
no significant alteration in the baseline ionised fraction of
53.6% in the CSF despite considerable increase in serum
magnesium during administration of MgSO4 for periods of
up to 48 hours.13 Baseline ionised magnesium in the serum
was reported to be 69.8% in another study.25
Other intravenous regimens
Table 2 also shows the available data for seven less popular
intravenous regimens. With 4-g loading dose followed by
2 g/hour intermittent IV bolus injections, one study
showed a rapid rise of serum magnesium from baseline
value to a peak concentration of 1.64 mmol/l by 15 minutes,
after which it fell very rapidly to 1.23 mmol/l by
2 hours.12 The first maintenance bolus dose at 2 hours was
accompanied by another peak (1.69 mmol/l) and rapid fall
to 1.07 mmol/l at 4 hours. For the most of the 4-hour follow
up of post-dose serum concentration, the mean levels
of magnesium remained around 1.00 mmol/l.
A study that administered a 4.5-g loading dose and
1.8 g/hour maintenance dose reported a gradual serum
magnesium rise from 2.01 mmo/l at 2 hours to a peak of 2.52 mmol/l at 12 hours before it declined to 2.43 mmol/l
at 24 hours.19Another study evaluating the pharmacokinetics
of lower intravenous regimen (2 g loading plus 1.5
g/hour continuous infusion) to women with eclampsia
showed a very slow rise in mean serum magnesium levels
from 1.27 at 1 hour to 1.56 mmol/l at 4 hours.3 Data (not
shown) showed that the level remained sustained at
<2 mmol/l by 6 hours of maintenance. The same study
reported serum concentration-time data for a 2-g loading
dose only for three eclamptic women.3 The study showed a
transitory and trivial effect on the baseline magnesium—
peaking within 10 minutes of injection and falling rapidly
to 0.90 mmol by 2 hours.

Intramuscular MgSO4 regimens
4-g IV and 10-g IM loading dose, and 5-g IM maintenance
dose every 4 hours (Pritchard regimen)
Six studies provided serum magnesium concentration-time
data based on this regimen.5,14,18,27–29 Reported baseline
magnesium levels were <1.00 mmol/l (Table 3). Following
the loading dose, serum magnesium level rose sharply from
the baseline to at least two-fold by ½ hour (1.90–
2.79 mmol/l). After the initial rise, Chissell et al. reported a
slight decline in serum magnesium at 1 hour but relatively
steady levels between 1.60 and 1.75 mmol/l until 12 hours
of the maintenance injection.16 In the same study, serum
level peaked at 2.07 mmol/l at 1½ hours following the initiation
of treatment.
Sibai et al.5 showed a rapid rise in the first 4 hours with
values all above 2.00 mmol/l and a slight decline by 8 and
12 hours. Singh et al. reported a similar pattern with a
gradual increase demonstrated between 4 and 24 hours following
initiation of treatment.30 Ekele and Badung also
showed a more than two-fold rise in serum magnesium
compared with baseline and mean levels slightly above
2.00 mmol/l at 8 and 12 hours of maintenance injections.20
Shreya et al. 29 also reported a similar increase in serum
magnesium level in the first ½ hour and values slightly
lower than 2.00 mmol/l at 4 hours. Overall, the serum-concentration
data fluctuated much more with this regimen
than with continuous intravenous regimens described
above, and serum level versus time data were less consistent
across studies. However, for every time point reported,
there were mean values reaching ≥2.00 mmol/l, but none
reached 3.00 mmol/l.
Only one study reported on magnesium sulphate toxicity
using this regimen. The study reported respiratory depression
and death in a woman with a serum magnesium level
of 9.90 mmol/l.31
10-g IM loading dose and 5-g IM maintenance dose every
4 hours
Two studies reported serum-concentration data based on
this regimen.3,5 In one of the studies, the mean levels of
serum magnesium at 1, 2, and 4 hours were observed to be
1.36, 1.56, and 1.48 mmol/l, respectively.3 The other study
only reported a steady state level of 1.83 mmol/l.5
3-g IV and 10-g IM (13 g) loading dose only
With this regimen, one study reported a baseline serum
magnesium of 2.10 mmol/l.15 The mean magnesium levels
rose to 2.25 and 2.30 mmol/l at 1 and 2 hours following
treatment, respectively, and gradually declined at 4 hours
to 1.90 mmol/l.
Other intramuscular regimens
One study reported serum magnesium levels following the
administration of a single 10-g IM loading dose.3 Mean
serum magnesium concentration rose slowly to attain its
average peak of 1.81 mmol/l between 1½ and 2 hours
before it declined to 1.44 and 1.34 mmol/l at 4 and
6 hours, respectively. Another study reported a steady-state
level of 1.50 mmol/l following a single 12-g loading dose
regimen.28 Using a single dose of 4 g MgSO4 intravenously
and 4 g intramuscularly, another study reported that the
mean serum level increased notably in the first ½ hour and
then gradually decreased to levels similar to those achieved
by the 10-g IM loading dose-only regimen.

It would be useful to consider what is known about the
basic pharmacokinetic profile of MgSO4 when selecting
dosage regimens of unproven efficacy for clinical use. As
evident from the serum levels attained during treatment
with the Zuspan regimen, it appears that MgSO4 can be
protective even with serum concentrations of <2 mmol/l.
Therefore, titrating MgSO4 injections to achieve a pre-set
therapeutic range of 2–3.5 mmol may risk toxic levels
without necessarily improving clinical protection against
seizures. Regardless of the slight differences in the pharmacokinetic
profiles of the two currently recommended
regimens, the comparability of their clinical efficacy is
reassuring and does not justify a further increase in the
total dose of MgSO4 for prophylaxis and treatment of
eclampsia. The single or intermittent use of intravenous
bolus injection of 2 g MgSO4 does not produce a sustained increase in serum magnesium levels to provide a
clinically meaningful protection against seizures and is
best avoided.