Obstetrics

Preeclampsia: Should Plasma Albumin

Level Be a Criterion for Severity?
Janie Benoit, MD,1 velyne Rey, MD, MSc, FRCPC2
1

Department of Obstetrics and Gynecology, University of Montreal, Montreal QC

CHU Sainte-Justine, Departments of Obstetrics and Gynecology and Medicine, University of Montreal, Montreal QC

Abstract
Objectives: To determine whether plasma albumin level (ALB) is
associated with preeclampsia (PE) complications and to define
its role in the evaluation of the severity of PE.
Methods: We conducted a retrospective study of women with PE at
CHU Ste-Justine from January 2005 to December 2008 who had
an ALB measurement before delivery. Severe PE was defined
according to the Society of Obstetricans and Gynaecologists of
Canada guidelines, except for the ALB and gestational age at
diagnosis. We compared ALB between women with non-severe
PE and with severe PE. The rates of maternal complications and
adverse perinatal outcomes associated with PE were compared
between women with an ALB < 20 g/L and those with an
ALB 20 g/L.
Results: Among the 459 women included, 62.7% met the criteria for
a diagnosis of severe PE. The mean gestational age was
33.84.1 weeks at diagnosis of PE and 34.63.9 weeks at
delivery. After adjustment for gestational age and number of
fetuses, ALB was found to be lower in women with severe PE
(23.53.7 vs. 25.92.8 g/L; P < 0.05). Thrombocytopenia,
increased serum ALT, severe proteinuria, and a need for platelet
transfusion were more frequent among the 44 women with an
ALB < 20 g/L (P < 0.05). Two women who had an ALB < 20 g/L
and no other criteria for PE developed PE before 34 weeks
gestation.
Conclusion: A plasma albumin level below 20 g/L was not found to
be an independent marker for severe preeclampsia because
all women with a low plasma albumin level had other adverse
conditions.

Rsum
Objectifs: Dterminer si le taux plasmatique dalbumine (ALB) est
associ aux complications de la prclampsie (PE) et dfinir
son rle dans lvaluation de la gravit de la PE.

Mthodes: Nous avons men une tude rtrospective qui portait sur
les femmes prsentant une PE, au CHU Ste-Justine entre janvier
2005 et dcembre 2008, chez lesquelles une mesure du taux ALB
avait t effectue avant laccouchement. Le terme PE grave
a t dfini en fonction des directives cliniques de la Socit des
obsttriciens et gyncologues du Canada, sauf en ce qui a trait
au taux ALB et lge gestationnel au moment du diagnostic.
Nous avons compar les taux ALB constats chez des femmes
prsentant une PE non grave ceux des femmes prsentant
une PE grave. Nous avons galement compar les taux de
complications maternelles et dissues prinatales indsirables
associs la PE des femmes prsentant un taux ALB < 20 g/l
ceux des femmes prsentant un taux ALB 20 g/l.
Rsultats: Parmi les 459 femmes tudies, 62,7% ont satisfait aux
critres dun diagnostic de PE grave. Lge gestationnel moyen
tait de 33,84,1semaines au moment du diagnostic de PE
et de 34,63,9 semaines au moment de laccouchement. la
suite de la neutralisation de leffet de lge gestationnel et du
nombre de ftus, nous avons constat que le taux ALB tait
infrieur chez les femmes prsentant une PE grave (23,53,7,
par comp. avec 25,92,8 g/l; P < 0,05). La thrombocytopnie, la
hausse de lALT srique, la protinurie grave et la ncessit de
procder une transfusion plaquettaire taient plus frquentes
chez les 44 femmes prsentant une taux ALB < 20g/l (P < 0,05).
Deux femmes qui prsentaient un taux ALB < 20 g/l sans aucun
autre critre de PE en sont venues connatre une PE avant
34 semaines de gestation.
Conclusion: Nous ne sommes pas parvenus tablir quun
taux plasmatique dalbumine en de de 20 g/l constituait un
marqueur indpendant de la prclampsie grave, puisque
toutes les femmes chez lesquelles un faible taux plasmatique
dalbumine a t constat prsentaient dautres tats
indsirables.

J Obstet Gynaecol Can 2011;33(9):922926

INTRODUCTION
Key Words: Plasma albumin, pregnancy, severe preeclampsia,
complications
Competing Interests: None declared.
Received on January 31, 2011
Accepted on March 15, 2011

922 l SEPTEMBER JOGC SEPTEMBRE 2011

reeclampsia affects from 3% to 14% of pregnancies

worldwide.1 It is a major cause of maternal and
perinatal morbidity and mortality; for example, 40% of
induced preterm births are due to preeclampsia.2 Menzies
et al. have shown that standardized surveillance testing

Preeclampsia: Should Plasma Albumin Level Be a Criterion for Severity?

for end-organ dysfunction in women with preeclampsia

helps to reduce maternal complications.3 The guidelines
on hypertensive disorders of pregnancy published by the
Society of Obstetricians and Gynaecologists of Canada
discuss considering a plasma albumin level of <20g/L
as a criterion of severity, based on studies in non-pregnant
women and on expert opinion.4 However, there are no
data to support this advice.

ALB and gestational age at diagnosis of PE. Severe PE

was defined as preeclampsia with proteinuria 3 g/d, or
with one or more associated conditions, which included
maternal symptoms, eclampsia, severe hypertension,
pulmonary edema, placenta abruption, abnormal maternal
laboratory testing with symptoms, or fetal morbidity.4 The
first ALB measurement at the time of diagnosis of PE was
used in the assessment of severity.

The objective of our study was to determine whether ALB,

especially below the threshold of 20g/L, was associated
with complications of preeclampsia. We considered the
following maternal complications: need for blood or platelet
transfusion, disseminated intravascular coagulopathy,
hepatic hematoma, hepatic rupture, cerebral hemorrhage,
cerebral edema, eclampsia, cerebro-vascular accident, and
death. Perinatal complications considered were placenta
abruption, intrauterine fetal demise, and perinatal death.

The first 50 files were reviewed by both investigators for

quality control of the collected data.

METHODS

We conducted a retrospective study involving women

with a diagnosis of PE managed at the CHU Ste-Justine,
Montreal, from January 2005 to December 2008. We
included women with a confirmed diagnosis of PE
who delivered in our institution and who had an ALB
measurement taken between diagnosis of PE and delivery.
We excluded women who delivered before 20 weeks
gestation or who had chronic conditions that could lower
ALB, such as nephrotic syndrome, hepatic cirrhosis, renal
or cardiac insufficiency, renal transplant, malnutrition with
a BMI <18 kg/m2, or chronic malabsorption. If the same
patient had more than one pregnancy complicated by
PE during the study period, only the first pregnancy was
included.
Women with PE were identified by the Medical Record
Department, using ICD 9 codes (hypertension and
pregnancy), and their eligibility for the study was verified
by one of the investigators. The charts of eligible women
were reviewed, and information on baseline maternal
characteristics, pregnancy details, PE features, and delivery
was collected. Preeclampsia was classified as severe or nonsevere according to the criteria of severity recommended
in the 2008 SOGC guidelines,4 with the exception of

ABBREVIATIONS
ALB

plasma albumin level

PE

preeclampsia

PIERS

Pre-eclampsia Integrated Estimate of RiSk

ROC

receiver operating characteristic

Statistical analyses were performed using SPSS version

16.0 (IBM Corp., Somers NY). Continuous variables were
expressed as mean standard deviation and categorical
variables as percentages. Comparisons between groups
were made using Student t test, Mann-Whitney U test,
or analysis of variance for continuous variables, and
chi-square or Fisher exact test for categorical variables.
Logistic regression analysis was used to adjust the ALB for
the number of fetuses and gestational age at the time of
measurement. An ROC curve was constructed to illustrate
the relationship between ALB and severity of PE, and
P<0.05 (two tailed) was considered statistically significant.
On the basis of our local prevalence of severe preeclampsia,
we estimated that a review of charts covering four years
would allow the detection of a 20% difference between the
number of women with severe and those with non-severe
preeclampsia among those with an albumin level <20g/L.
The study protocol was approved by Ste-Justine hospital
research ethics committee
RESULTS

Among the 797 patients identified by the medical record

department as having hypertension in pregnancy, 459
patients were eligible and included in the study. Most of
those included were Caucasian (74.9%) and nulliparous
(68.8%). Mean maternal age was 30.05.6 years, and mean
BMI (based on first trimester weight) was 26.46.3kg/m2.
Mean gestational age at diagnosis of preeclampsia was
33.74.1 weeks and at delivery was 34.63.9 weeks.
Delivery before 34 weeks of gestation occurred in 49% of
women. The rate of intrauterine growth restriction (below
the 5th percentile using Canadian standards5) was 22%.
Severe PE was identified in 288 women (62.7%) on the
basis of one or more of the following parameters: severe
proteinuria (31.6%), severe hypertension (29.8%), hepatic
involvement (27.3%), thrombocytopenia (9.2%), elevated
serum creatinine (5.0%), or eclampsia (0.2%).
SEPTEMBER JOGC SEPTEMBRE 2011 l 923

Obstetrics

Table 1. Maternal characteristics according to plasma albumin level

Women with albumin
level 20 g/L
n = 415

Women with albumin

level < 20 g/L
n = 44

29.9 5.6

30.75.7

Age, years
Ethnicity
Caucasian

304 (73.3)

39 (88.6)

Black

094 (22.7)

05 (11.4)

Other

17 (4.1)

Weight in first trimester, kg

71.317.6

66.615.6

BMI, kg/m2

26.56.3

25.15.7

BMI > 30 kg/m

099 (23.9)

08 (18.2)

Primigravid

194 (46.7)

22 (50.0)

Nulliparous

283 (68.2)

33 (75.0)

Singleton

383 (92.3)

39 (88.6)

02 (4.5)

48 (11.6)

Other pathology
Pre-existing hypertension
Pre-gestational diabetes

17 (4.1)

3 (6.8)

052 (13.1)

4 (9.8)

Asthma

35 (8.4)

05 (11.4)

Thyroid disease

32 (7.7)

4 (9.1)

Previous PE

31 (7.5)

3 (6.8)

Gestational diabetes

Data are expressed as mean standard deviation or numbers (percentages)

No P values were significant.

Women with severe PE had a significantly lower ALB

(23.52.8g/L) than women with non-severe PE
(25.52.8g/L; P<0.05) after adjustment for gestational
age at assessment and the number of fetuses (logistic
regression analysis).
An ALB <20g/L was observed in 44 women (9.6%).
No statistical difference in maternal characteristics
was observed between women with an ALB <20g/L
and those with an ALB 20 g/L (Table 1). However,
pregnancy outcomes such as premature delivery, severity
of preeclampsia, length of hospitalization, and newborn
weight were worse in women with an ALB <20g/L
than in those with ALB 20 g/L (Table 2). The rates of
pulmonary edema, elevated serum creatinine, and newborn
weight <5th percentile were similar in the two groups.
Among the 44 women with an ALB <20g/L, 42 had at
least one other criterion of severity, and the remaining two
women had a diagnosis of PE before 34 weeks gestation.
The area under the ROC curve for the performance of
ALB in detecting severe PE was 0.67 (95% CI 0.62 to 0.72)
(Figure). No plasma albumin level threshold was clearly
associated with severe PE.
924 l SEPTEMBER JOGC SEPTEMBRE 2011

DISCUSSION

In this observational retrospective study, we observed that

women with a plasma albumin level <20g/L presented
with many complications of preeclampsia. However, such
women had at least one other maternal criterion to qualify
for severe PE as defined by the Society of Obstetricians
and Gynaecologists of Canada guidelines.4 Thus, we were
not able to study pregnancy complications in women with
an ALB <20g/L who had no other criteria for severe PE.
Therefore, the ALB threshold of 20g/L does not identify
women who would not otherwise have been identified by
other well-studied criteria for severe PE.
In a non-pregnant population, a low plasma albumin
level is associated with severity in various inflammatory
diseases6,7 and with post-surgical mortality and morbidity.8
During pregnancy, ALB decreases and changes over
time.9,10 This is explained by an increase in plasma and
interstitial volume, and possibly by an increase in albumin
metabolism.8 Moreover, ALB is lower in women with PE
than in healthy pregnant women.8,11 Increased capillary
permeability secondary to endothelial damage seems
to be partly responsible for this finding.11 Data on ALB

Preeclampsia: Should Plasma Albumin Level Be a Criterion for Severity?

Table 2. Pregnancy outcomes according to plasma albumin level

Women with albumin
level 20 g/L
n = 415

Women with albumin

level < 20 g/L
n = 44

Gestational age at diagnosis, wk

34.04.1

31.83.7

< 0.001

Diagnosis < 34 weeks

185 (44.6)

31 (70.5)

0.001

118 (28.4)

19 (43.2)

0.03

15 (3.6)

1 (2.3)

NS

31 (7.5)

11 (25.0)

0.001

106 (25.6)

19 (43.2)

0.01

Women with PE

Criteria of severity
Severe hypertension
Pulmonary edema
Platelet concentration < 100 10 /L
9

Hepatic involvement
Increase in serum creatinine
Severe proteinuria
Placental abruption

19 (4.6)

4 (9.1)

NS

108 (26.3)

36 (81.8)

< 0.001

11 (4.5)

1 (2.4)

NS

1 criterion of severity

246 (59.4)

42 (95.5)

< 0.001

Hospitalization stay, days

6.017.8

11.79.0

< 0.001

04 (1.0)

4 (9.1)

0.04

Platelet transfusion
Postpartum deterioration

088 (35.8)

17 (40.5)

NS

Gestational age at delivery, wk

34.83.9

32.83.5

< 0.001

Delivery 34 weeks gestation

149 (35.9)

25 (56.8)

0.09

2.21.0

1.80.9

0.003

082 (21.8)

09 (24.3)

NS

16 (3.9)

1 (2.3)

NS

Newborn weight, kg
Newborn weight < 5th centile
Perinatal mortality

Data are expressed as mean standard deviation or numbers (percentages)

and severe preeclampsia are so limited that the decision

to add the plasma albumin level threshold of 20g/L to
the criteria for severe PE in the SOGC guidelines was
based not only on expert opinion, but also on studies
performed in non-pregnant populations, which showed
that edema develops at this threshold.6,7,12 The PIERS
study evaluated the performance of the ALB threshold of
18g/L in predicting severe adverse maternal or perinatal
outcomes, and the PIERS investigators did not observe
an association (P=0.328 and P=0.438 respectively).13
These investigators did not retain ALB as a significant
predictor of adverse maternal outcome in their prospective
assessment model.14

Receiver operating characteristic curve of plasma

albumin level in relation to preeclampsia severity
1.0

Sensitivity

0.8

0.6

0.4

0.2

0.0

0.2

0.4

0.6

1 - Specificity

0.8

1.0

We are aware that we studied a specific group of women:

most of them presented with severe preeclampsia and
delivered prematurely. This is explained by the fact that
our institution is a tertiary care centre with a high rate of
inward transfer for severe or early onset preeclampsia.
However, the large number of women with severe
preeclampsia in our study allows us to extract much useful
SEPTEMBER JOGC SEPTEMBRE 2011 l 925

Obstetrics

and reliable information. A limitation of the study is that it

was performed retrospectively. A large number of patients,
especially women with non-severe preeclampsia, were
excluded from the study because no plasma albumin level
had been measured before delivery.
We did not find that pulmonary edema was more frequent
in women with ALB <20g/L. This may be explained by
the fact that the incidence of pulmonary edema was low
in the study group; pulmonary endothelial damage may be
more important than ALB in causing this complication. It
is possible that ALB is associated with peripheral edema,
but this was not investigated in this study. However, this is
a subjective variable and is not a cause of severe maternal
morbidity. We observed a strong association between
ALB <20g/L and severe proteinuria. This is also seen in
patients with nephrotic syndrome, in whom protein is lost
in the urine over many weeks.15 This condition, however, is
different from PE, which develops more acutely.
CONCLUSION

Our data suggest that even if the plasma albumin level is

lower in women with severe PE, the threshold ALB of
20 g/L is not an independent criterion for severe PE. We
therefore question the value of including a plasma albumin
level <20g/L in the list of criteria for severe preeclampsia.
Further studies will be required to confirm these results.
ACKNOWLEDGEMENTS

We thank Mrs Cline Durand from the Medical Record

Department at the CHU Ste-Justine for her help.
REFERENCES
1. Cunningham FG, Lindheimer MD. Hypertension in pregnancy.
N Engl J Med 1992;326:92732.

926 l SEPTEMBER JOGC SEPTEMBRE 2011

2. Meis PJ, Goldenberg RL, Mercer BM, Iams JD, Moawad AH, Miodovnik
M, et al. The preterm prediction study: risk factors for indicated preterm
births. Maternal-Fetal Medicine Units Network of the National Institute
of Child Health and Human Development. Am J Obstet Gynecol
1998;178:5627.
3. Menzies J, Magee LA, Li J, MacNab YC, Yin R, Stuart H, et al. Instituting
surveillance guidelines and adverse outcomes in preeclampsia. Obstet
Gynecol 2007;110:1217.
4. Magee LA, Helewa M, Moutquin J-M, von Dadelszen P, Cardew S, Ct
A-M, et al. Diagnosis, evaluation and management of the hypertensive
disorders of pregnancy. SOGC Clinical Practice Guidelines, No. 206,
March 2008. J Obstet Gynaecol Can 2008;30(3 Suppl 1):S1S48.
5. Kramer MS, Platt RW, Wen SW, Joseph KS, Allen A, Abrahamowicz M,
et al. A new and improved population-based Canadian reference for birth
weight for gestational age. Pediatrics 2001;108:E35.
6. Manning RD Jr, Guyton AC. Effects of hypoproteinemia on fluid
volumes and arterial pressure. Am J Physiol 1983;245:H28493.
7. Manning RD Jr. Effects of hypoproteinemia on renal hemodynamics,
arterial pressure, and fluid volume. Am J Physiol 1987;252:F918.
8. Horne CH, Howie PW, Goudie RB. Serum-alpha2-macroglobulin,
transferrin, albumin, and IgG levels in preeclampsia. J Clin Pathol
1970;23:5146.
9. Larsson A, Palm M, Hansson LO, Axelsson O. Reference values for
clinical chemistry tests during normal pregnancy. BJOG 2008;115:87481.
10. Larsson A, Palm M, Hansson LO, Basu S, Axelsson O. Reference values
for alpha1-acid glycoprotein, alpha1-antitrypsin, albumin, haptoglobin,
C-reactive protein, IgA, IgG and IgM during pregnancy. Acta Obstet
Gynecol Scand 2008;87:10848.
11. Bhatia RK, Bottoms SF, Saleh AA, Norman GS, Mammen EF, Sokol RJ.
Mechanisms for reduced colloid osmotic pressure in preeclampsia.
Am J Obstet Gynecol 1987;157:1068.
12. Fadnes HO, Pape JF, Sundsfjord JA. A study on oedema mechanism in
nephrotic syndrome. Scand J Clin Lab Invest 1986;46:5338.
13. Menzies J, Magee LA, MacNab YC, Ansermino JM, Li J, Douglas MJ, et
al. Current CHS and NHBPEP criteria for severe preeclampsia do not
uniformly predict adverse maternal or perinatal outcomes. Hypertens
Pregnancy 2007;26:44762.
14. von Dadelszen P, Payne B, Ansermino JM, Broughton Pipkin F,
Ct A-M, Douglas MJ. Prediction of adverse maternal outcomes in
pre-eclampsia: development and validation of the fullPIERS model.
Lancet 2011;377:21927.
15. Margarson MP, Soni N. Serum albumin: touchstone or totem?
Anaesthesia 1998;53:789803.