American Journal of Obstetrics and Gynecology (2006) 195, 186–91

www.ajog.org

Neuropsychological performance in normal pregnancy

and preeclampsia
Sarosh Rana, MD,a Marshall Lindheimer, MD,a,b Judith Hibbard, MD,c
Neil Pliskin, PhDd

Departments of Obstetrics and Gynecologya and Medicine,b University of Chicago; Departments of Obstetrics and
Gynecologyc and Psychiatry,d University of Illinois at Chicago, Chicago, IL

Received for publication November 4, 2005; revised December 19, 2005; accepted December 22, 2005

KEY WORDS Objective: The objective of the study was to evaluate neurocognitive function in preeclampsia
Neurocognitive and normal pregnancy.
performance Study design: Three groups (each n = 15) were studied before and after delivery using standard
Preeclampsia neurocognitive test techniques. Group A consisted of normal laboring patients; group B,
Magnesium sulphate preeclamptics receiving magnesium; and group C, women in preterm labor receiving MgSO4
tocolysis (Mg control). The tests, examining attention, working memory, explicit memory, audi-
tory comprehension, and measures for emotional distress, pain, and fatigue were analyzed via
4-way multivariate analysis of variance and multiple t tests.
Results: Preeclamptics receiving MgSO4 had better attention and working memory (P = .05),
compared with normal laboring women and the preterm patients. Explicit memory was impaired
in all groups, and this could not be accounted for by pain or emotional distress.
Conclusion: We could detect no cognitive defects in preeclamptics, compared with normotensive
gravidas, at least while the cerebral vasodilator MgSO4 is being infused.
Ó 2006 Mosby, Inc. All rights reserved.

Mental status changes that range from mood swings
to psychosis have been described in preeclamptic and
eclamptic patients, but, surprisingly, we could locate no
studies evaluating cognitive performance in women with
this disorder. Indeed, there is a marked paucity of
reports describing neurocognitive function in gestation
or textbook discussions of behavior and memory func-
tion during pregnancy, in general. There are, however,
limited data suggesting that pregnant women rate their
memories as poorer than before conception and test
poorer in this function than nonpregnant controls.1-5 In
Reprints not available from the authors.
this respect, some attribute the decrease in functional
memory to decrements in plasma epinephrine, seroto-
nin, and dopamine levels during gestation.6
Obstetricians performing interventional procedures
have an ethical and legal responsibility to provide
patients with relevant information when obtaining in-
formed consent prior to an intervention, and an impor-
tant consideration is the patient’s status at the time of
the consent process. A study was therefore designed to
evaluate neurocognitive function in both normotensive
and preeclamptic subjects, using a battery of 7 specific
neurocognitive tests administered to each subject during
and after gestation.

0002-9378/$ - see front matter Ó 2006 Mosby, Inc. All rights reserved.
doi:10.1016/j.ajog.2005.12.051
Rana et al
Material and methods
Three groups were studied: normotensive women in
labor at term (group A), preeclamptics (group B), and
patients admitted with preterm labor receiving magne-
sium tocolysis (group C). There were 15 patients in each
group. The diagnosis of preeclampsia was made by the
attending physician and required de novo hypertension
and abnormal proteinuria after midpregnancy. Three of
the 15 were designated as severe preeclamptics, based on
blood pressures greater than 160/ 110 mm Hg and/or
elements of hemolysis, elevated liver enzymes, and low
platelet count syndrome.
For group A, the initial cognitive evaluation was
performed after their admission to the labor and
delivery suite and repeated 36 hours postpartum. In
preeclamptics (group B), cognitive testing was per-
formed more than 2 hours after the commencement of
intravenous MgSO4 therapy, magnesium levels obtained
at the time of interview. The level in all patients was be-
tween 4.7 and 7.0 mg/dL. Repeat testing was performed
more than 36 h postpartum, more than 12 hours after
magnesium therapy had been discontinued. Preeclampsia
was diagnosed by the attending physician as de novo
hypertension (blood pressures greater than 140/90 mm
Hg twice 6 hours apart or greater than 160/110 mm
Hg on a single measurement) and new-onset proteinuria
after gestational week 20. Severe preeclampsia was de-
fined as blood pressure greater than 160/110 mm Hg;
3-4C proteinuria; and presence of marked multiorgan
involvement including HEELP, intrauterine growth
retardation, and neurological symptoms. Preterm labor
was defined as regular uterine contractions concomitant
with cervical change less than 34 weeks’ gestation.
Patients in preterm labor (group C) were considered
the magnesium control and tested more than 2 hours
after initiation of MgSO4 and more than 36 hours post-
partum as was described for group B. Exclusion criteria
included imminent delivery, nonreassuring tests of fetal
well-being, positive urine toxicology screen, preexisting
history or signs of central nervous system disease or
mental illness, and medications that affect cognitive
function (ie, sedatives, narcotics). The protocol was
approved by the Institutional Review Board at the Uni-
versity of Chicago, and all participants gave informed
consent in writing.
Neurocognitive testing
The neuropsychological evaluation consisted of a battery
of tests administered over a 30- to 35-minute period that
included alternate forms, where possible, as an additional
control for practice effects. The measures included an
assessment of premorbid intellectual status (Wechsler
Test of Adult Reading7), attention/working memory
(Digit Span and Letter-Number Sequencing from the
187
Wechsler Adult Intelligence Scale, third edition8), mem-
ory (Hopkins Verbal Learning Test9), and auditory com-
prehension (Complex Ideation Subtest of the Boston
Diagnostic Aphasia Exam10). Additionally, each subject
was administered self-report measures designed to assess
the presence of psychological distress (Beck Depression
Inventory-211) and fatigue (Brief Fatigue Inventory12),
and the level of subjective pain was quantified via an
analogue pain scale as described below.
Wechsler Test of Adult Reading7: This is a reading test
comprised of 50 words with irregular pronunciations
that measures a patient’s premorbid level of intellectual
functioning. This test is conormed with the Wechsler
Adult Intelligence Scale (WAIS) 3 and was used to cal-
culate estimated IQ scores based on demographic varia-
bles. It was administered during the baseline assessment
only.
Digit Span8: This analysis assesses working memory
and attention capacity in the auditory modality by ask-
ing the subject to repeat increasingly lengthy series of
digits in both forward and reverse order.
Letter-Number Sequencing Test8: This is a measure of
working memory ability and the capacity to actively
hold information in mind and manipulate it mentally.
Subjects were orally presented with a series of letters
and numbers and asked to sequence the numbers first
and then the letters.
Hopkins Verbal Learning Test-Revised9: This test is a
measure of auditory verbal learning and memory. Sub-
jects were presented with 3 learning trials of a 12-word
list (grouped into 3 semantic categories) and asked for
delayed free recall and delayed recognition. Six equiva-
lent versions of the test were available.
Complex Ideation Subset of Boston Diagnostic Apha-
sia Exam10: This task, part of a larger battery designed
to assess various aspects of aphasia, examines auditory
comprehension of detailed sentences by requiring the
subject to respond ‘‘yes’’ or ‘‘no.’’
Brief Fatigue Inventory12: This inventory is a rapid as-
sessment of fatigue specifically for medically ill patients.
It is a short questionnaire involving 4 questions.
Analogue Pain Scale: We asked the subjects to rate
their level of pain on a 10-point scale, with 1 being
‘‘pain free’’ and 10 being ‘‘the worst pain of my life.’’
Statistical analysis
Demographic comparisons with regard to age, education,
and premorbid intelligence were carried out between
groups using multiple t tests. A 4-way, repeated-measures
multivariate analysis of variance was computed to assess
group differences in neuropsychological function. Labor
status was used as the between-subject variable, whereas
neuropsychological function (attention/working mem-
ory, verbal memory, and auditory comprehension) as
well as distress status, pain, and fatigue levels was used
188 Rana et al

Table I Demographic information for age, education, and IQ

Normal labor Preeclamptics with Preterm labor with
n = 15 magnesium n = 15 magnesium n = 15 Significance
Age (y) Minimum 22 Minimum 18 Minimum 16 NS
Maximum 32 Maximum 38 Maximum 38
Mean 27.07 Mean 26.60 Mean 26.79
SD 3.011 SD 7.519 SD 6.216
Education (y) Minimum 19 Minimum 11 Minimum 11 NS
Maximum 18 Maximum 22 Maximum 22
Mean 13.50 Mean 13.71 Mean 14.54
SD 2.210 SD 2.867 SD 3.045
Estimated premorbid IQ 102.8 100.44 102.08 NS

Table II Immediate verbal memory (HVLTest trial 1 raw score)

Normal labor n = 15 Preeclamptics n = 15 Preterm n = 15
Neuropsychological test Mean SD Mean SD Mean SD
HVLT, time A 4.93 1.542 4.64 1.206 4.50 2.510
HVLT, time B 4.21 1.369 4.82 1.471 5.33 1.966
Maximum score = 12. HVLT, Hopkins Verbal Learning Test.

as the within-subject factors. Significant findings were
clarified using a series of univariant analyses of variance
for each neuropsychological function and, when indi-
cated, each component variable. Because there were no
differences in age and education between groups, raw
scores on neuropsychological tests were used in these
analyses. To maintain acceptably low type I error rates,
a conservative approach to a priori planned analyses
was used. A P value of less than .05 was considered signif-
icant. To provide a standard metric for comparison
across neuropsychological tests and domains with pub-
lished norms of healthy adults, raw scores were subse-
quently standardized (z-scores). Summary measures
were then calculated for attention, working memory, ver-
bal memory, visual memory, and motor skills for com-
parison purposes.
Results
Fifteen women were recruited into each group. Three of
the 15 preeclamptics were considered severe (by pressure
criteria in 1 and the presence of signs indicating hemol-
ysis, elevated liver enzymes, and low platelet count in 2).
Patients in preterm labor were between gestational
weeks 26 and 34. All participants completed both arms
of the test. Of all gravidas invited to participate in this
investigation, only 2 patients declined, both in the
preterm labor group.
Table I depicts age, level of education, and IQ. There
were no differences between the groups in any of these
variables, nor were differences in auditory comprehen-
sion and fatigue level apparent among groups. However,
the distress level was slightly higher among preeclamp-
tics and women laboring preterm, compared with those
delivering at term. Pain was nonsignificantly higher in
normal laboring women. Immediate Verbal Memory
(Hopkins Verbal Learning Test) was similar among
groups and within groups (Table II). Delayed verbal
memory, computed by the delayed component of the
Hopkins Verbal Learning Test, was nonsignificantly
lower in the preeclamptics antepartum. Scores improved
postpartum in all groups, but the increase was greater in
the preeclamptic patients, erasing the small difference
noted during time period A (Table III).
Regarding attention, derived from the Digit Span,
there was no difference in individual groups before and
after delivery. The preeclamptics receiving magnesium
had better scores at both end points (less than 0.05),
whereas preterm women receiving magnesium had
poorer scores, compared with normal laboring women,
although not statistically significant (Table IV). Work-
ing memory, computed by Letter Number Sequencing,
improved modestly after delivery in each group,
preeclamptics receiving MgSO4, scoring better than
the other groups, both before and after delivery,
(P ! .05). Preterm patients receiving magnesium scored
the lowest in working memory (Table V).
We also performed age-corrected comparisons of
the attention and working memory, using z-scores to
compare them with normative nonpregnant controls.
The normal labor group (group A) had mild to moder-
ately impaired memory, and working memory was low
average. Similarly, in the preeclamptic group, memory
was moderately to severely impaired, compared with
normative standards, and working memory was low
Rana et al 189

Table III Delayed verbal memory (HVLT delayed recall raw score)
Normal labor n = 15 Preeclamptics n = 15 Preterm n = 15
Neuropsychological test Mean SD Mean SD Mean SD
HVLT, time A 7.64 1.865 5.50 2.635 6.17 2.041
HVLT, time B 7.93 1.685 7.90 1.524 6.83 3.189
HVLT, Hopkins Verbal Learning Test. Maximum score = 12.

Table IV Attention (digit span total raw score)

Normal labor n = 15 Preeclamptics n = 15 Preterm n = 15
Neuropsychological test Mean SD Mean S/D Mean S/D
Digit span, time A 15.86 2.983 18.80 3.795 14.60 2.30
Digit span, time B 16.86 2.179 18.80 4.367 14.80 1.30
Maximum score = 30.

Table V Total raw scores for letter number sequencing

Normal labor n = 15 Preeclamptics n = 15 Preterm n = 15
Neuropsychological test Mean SD Mean SD Mean SD
Letter number sequence, time A 9.62 2.063 10.80 3.048 8.40 3.647
Letter number sequence, time B 10.23 1.423 11.70 2.263 10.00 1.00
Maximum score = 30.

average to average. This was also true for patients in the
preterm labor group.
Comment
The administration of a battery of neurocognitive tests to
normotensive and preeclamptic women was undertaken
because of anecdotal claims that mental status changes,
sometimes serious, were associated with preeclampisa,
plus our inability to find proof of these assertions. Our
data, noting that normal pregnant women manifest mild
cognitive defects during labor and after delivery, com-
pared with age-matched normative nonpregnant controls,
confirms a limited literature.1-5 On the other hand,
preeclamptics receiving MgSO4 had better attention and
working memory performance both before and after
delivery, compared with laboring controls and patients
receiving magnesium to halt premature labor, the latter
group surprisingly with the lowest scores.
Women frequently report problems with attention,
concentration, and memory throughout pregnancy and in
the early postpartum period.1-5,13 Keenan et al2 for exam-
ple, observed a pregnancy-related memory loss, which
was limited to the third trimester. The decline was not
attributed to depression, anxiety, sleep deprivation, or
other physical changes associated with pregnancy.
Several explanations for the pregnancy-related
memory loss have been postulated. Some have suggested
hypoestrogenemia-induced memory impairment14,15;
however, pregnancy is a hyperestrogen state, lending little
credence to this explanation. High levels of pregnenolone
and allopregnanolone have been correlated with negative
effects on memory,16 a more plausible explanation in nor-
mal gestation. There is also evidence that high levels of
both endogenous and exogenous cortisol affect hippocam-
pal integrity and therefore explicit memory; free cortisol
levels may rise in pregnancy, lending plausibility to this ex-
planation.16 Oxytocin is also known for its amnestic
effects.2
We could locate only one report of the effect of
magnesium in pregnancy on neurocognitive abilities.
Ghia et al17 evaluated the effects of MgSO4 on attention,
comprehension, and memory in patients experiencing
preterm labor. Using the Paced Auditory Serial Addition
Test, they observed a negative effect of magnesium on at-
tention and rapid information-processing ability but not
on short-term memory or comprehension. Our findings
were similar, in that patients in preterm labor receiving
magnesium therapy had the poorest attention and work-
ing memory (tested by Digit Span and Letter-Number
Sequencing). Additionally, we eliminated pain and emo-
tional stress as causal in these observations.
With the previously cited information in mind, it is
remarkable that our preeclamptics displayed the least
amount of cognitive difficulties, despite the fact that they
were receiving magnesium. This may relate to a literature
suggesting the presence of cerebral ischemia, edema, and
190
other brain lesions in patients with preeclampsia,18-20 but
most are considered transient. Recently residual changes
have also been noted.21 In addition, there are reports us-
ing ultrasound Doppler technology describing cerebral
blood flow patterns in preeclamptics and eclamptics sim-
ilar to those associated with hypertensive encephalopa-
thy.20 In this respect clinical and experimental evidence
that magnesium may be both neuroprotective and a cere-
bral vascular dilator when used in preeclampsia may be
relevant to our neurocognitive observations.18,19,22-29
Our data have relevance in relation to an issue noted
in the introduction that obstetricians performing inter-
ventional procedures have an ethical and legal respon-
sibility to provide patients with relevant information
when obtaining informed consent, and an important
consideration is the patient’s status at the time of the
consent process. These data may aid physicians by
reinforcing their conclusions of patients’ comprehen-
sion, heretofore made intuitively alone. Here attention
and comprehension are key elements that, of course can
be affected by the medical condition including the
medication administered. In this respect, Ghia et al
(12) have noted that magnesium sulphate has an affect
on attention and working memory. In this study, paced
auditory serial addition Test scores revealed deficits in
information processing ability in preterm pregnant
patients receiving magnesium sulphate.
The adequacy of informed consent has been the subject
of few studies that uniformly note flaws in the design and
patient recall of informed consent.30-34 Our data may be
helpful in interpreting and furthering such studies.
The shortcomings of our study, some already alluded
to, include absence of studying preeclamptics prior to
the initiation of magnesium and the limited number of
patients studied, raising the possibility of type I or II
errors. However, and as noted, magnesium should have
aggravated cognitive problems, and obviously future
studies can involve only the mildest preeclamptics
because withholding or delaying such therapy in more
severe cases would not be ethical. Finally, because this
was a pilot investigation, more studies will be needed to
confirm and elaborate on our findings.
In conclusions, despite anecdotal reports to the con-
trary, preeclampsia does not decrease cognitive function,
at least after magnesium therapy has commenced. Pre-
eclamptic gravidas have an increased probability that their
delivery will be expeditious, often by an operative proce-
dure. Thus, it is reassuring that the disease does not appear
to compromise their ability to give informed consent.
References
1. Brindle PM, Brown MW, Brown J, Griffith HB, Turner GM. Ob-
jective and subjective memory impairment in pregnancy. Psychol
Med 1991;21:647-53.
Rana et al
2. Keenan PA, Yaldoo DT, Stress ME, Fuerst DR, Ginsburg KA. Explicit
memory in pregnant women. Am J Obstet Gynecol 1998;179:731-7.
3. De Groot RH, Hornstra G, Roozendaal N, Jolles J. Memory per-
formance, but not information processing speed, may be reduced
during early pregnancy. J Clin Exp Neuropsychol 2003;25:482-8.
4. Parsons C, Redman S. Self-reported cognitive change during preg-
nancy. Aust J Adv Nurs 1991;9:20-9.
5. Sharp K, Brindle PM, Brown MW, Turner GM. Memory loss dur-
ing pregnancy. Br J Obstet Gynaecol 1993;100:209-15.
6. Shetty DN, Pathak SS. Correlation between plasma neurotransmit-
ters and memory loss in pregnancy. J Reprod Med 2002;47:494-6.
7. Wechsler Test of Adult Reading Manual. San Antonio (TX): The
Psychological Corporation; 2001.
8. Wechsler D. Wechsler Adult Intelligence Scale. 3rd edition. San
Antonio (TX): The Psychological Corporation; 1997.
9. Brandt J, Benedict RH. Hopkins Verbal Learning Test-Revised
Professional Manual. Odessa (FL): Psychological Assessment
Resources, Inc; 1998.
10. Goodglass H, Kaplan E. The Boston Diagnostic Aphasia Exami-
nation. Philadelphia (PA): Lea & Febiger; 1983.
11. Beck AT, Steer RA, Brown GK. BDI-II Manual. San Antonio
(TX): The Psychological Corporation; 1996.
12. Mendoza TR, Wang XS, Cleeland CS, Morrissey M, Johnson BA,
Wendt JK, et al. The rapid assessment of fatigue severity in cancer
patients: use of the brief fatigue inventory. Cancer 1999;85:1186-96.
13. Kane FJ, Harman WJ, Keller MH, Erving JA. Emotional and cog-
nitive disturbance in the early perperium. Br Soc Clin Psychol
1968;114:99-102.
14. Kampen DL, Sherwin BB. Estrogen use and verbal memory in
healthy menopausal women. Obstet Gynecol 1994;83:979-83.
15. Phillips SM, Sherwin BB. Effects of estrogen on memory function in
surgically menopausal women. Psychoneuroendocrinology 1992;
17:485-95.
16. Nolten WE, Lindheimer MD, Rueckert PA, Oparil S, Ehrlich EN.
Diurnal patterns and regulation of cortisol secretion in pregnancy.
J Clin Endocrinol Metab 1980;51:466-72.
17. Ghia N, Spong CY, Starbuck VN, Scialli AR, Ghidini A. Magnesium
sulfate therapy affects attention and working memory in patients
undergoing preterm labor. Am J Obstet Gynecol 2000;183:940-4.
18. Rutherford JM, Moody A, Crawshaw S, Rubin PC. Magnetic res-
onance spectroscopy in preeclampsia: evidence of cerebral ische-
mia. BJOG 2003;110:416-23.
19. Schwartz RB, Feske SK, Polak JF, DeGirolami U, Iaia A, Beckner
KM, et al. Preeclampsia-eclampsia: clinical and neuroradiographic
correlates and insights into the pathogenesis of hypertensive en-
cephalopathy. Radiology 2000;217:371-6.
20. Williams K, Galerneau F. Maternal transcranial Doppler in
pre-eclampsia and eclampsia. Ultrasound Obstet Gynecol 2003;
21:507-13.
21. Zeeman GG, Fleckenstein JL, Twickler DM, Cunningham FG. Cere-
bral infarction in eclampsia. Am J Obstet Gynecol 2004;190:714-20.
22. Van der Grond J, Balm R, Kappelle LJ, Eikelboom BC, Mali WP.
Cerebral metabolism of patients with stenosis or occlusion of the
internal carotid artery. A 1H-MR spectroscopic imaging study.
Stroke 1995;26:822-8.
23. Feng DF, Zhu ZA, Lu YC. Effects of magnesium sulfate on trau-
matic brain edema in rats. Chin J Traumatol 2004;7:148-52.
24. Peker S, Abacioglu U, Sun I, Konya D, Yuksel M, Pamir NM.
Prophylactic effects of magnesium and vitamin E in rat spinal
cord radiation damage: evaluation based on lipid peroxidation
levels. Life Sci 2004;75:1523-30.
25. Kaplan S, Ulus AT, Tutun U, Aksoyek A, Ozgencil E, Saritas Z,
et al. Effect of MagSO4 usage on spinal cord ischemia-reperfusion
injury: electron microscopic and functional evaluation. Eur Surg
Res 2004;36:20-5.
26. Westermaier T, Zausinger S, Baethmann A, Schmid-Elsaesser R.
Dose finding study of intravenous magnesium sulphate in transient
Rana et al
focal cerebral ischemia in rats. Acta Neurochir (Wien) 2005;147:
525-32.
27. Sang N, Meng ZQ. Blockade of magnesium sulfate on transient out-
ward KC current and delayed rectifier KC current in acutely iso-
lated rat hippocampal neurons. Yao Xue Xue Bao 2002;37:510-5.
28. Lin JY, Chung SY, Lin MC, Cheng FC. Effects of magnesium
sulfate on energy metabolites and glutamate in the cortex during
focal cerebral ischemia and reperfusion in the gerbil monitored by
a dual-probe microdialysis technique. Life Sci 2002;71:803-11.
29. Belfort MA. Is high cerebral perfusion pressure and cerebral flow
predictive of impending seizures in preeclampsia? A case report.
Hypertens Pregnancy 2005;24:59-63.
30. Robinson G, Merav A. Informed consent: recall by patients tested
postoperatively. Ann Thorac Surg 1976;22:209-12.
191
31. Hutson MM, Biaha JD. Patients’ recall of preoperative instruction
for informed consent for an operation. J Bone Joint Surg Am 1991;
73:160-2.
32. White CS, Mason AC, Feechan M, Templeton PA. Informed con-
sent for percutaneous lung biopsy: comparison of two consent pro-
tocols based on patient recall after the procedure. Am J Roentgenol
1995;165:1139-42.
33. White LJ, Jones JS, Felton CW, Pool LC. Informed consent for
medical research: common discrepancies and readability. Acad
Emerg Med 1996;8:745-50.
34. Miller CK, O’Donnell DC, Searight HR, Barbarash RA. The
Deaconess Informed Consent Comprehension Test: an assess-
ment tool for clinical research subjects Pharmacotherapy 1996;
16:872-8.