Clinical Chemistry 43: 333-337, 1997;
(Clinical Chemistry. 1997;43:333-337.)
© 1997 American Association for Clinical Chemistry, Inc.
Medians for second-trimester maternal serum 
-fetoprotein, human chorionic gonadotropin, and unconjugated estriol; differences between races or ethnic groups
Peter A. Benna,
Jonathan M. Clive and
Roxanne Collins
University of CT Health Center, 1 Division of Human Genetics, Department of Pediatrics, and 2 Lowell P. Weicker Clinical Research Center, 263 Farmington Ave., Farmington, CT 06030-6140.
a Author for correspondence. Fax 203-679-3616.
 |
Abstract
|
|---|
Second-trimester maternal serum 
-fetoprotein (AFP), human chorionic
gonadotropin (hCG), and unconjugated estriol (uE3) are routinely
measured in screening fetuses at high risk for Down syndrome or open
neural tube defects (ONTD). For test interpretation, individual patient
values of these three analytes are related to population-derived median
values. We evaluated data from >21 000 pregnancies to determine the
extent of race-specific differences in median concentrations. For
samples at most gestational ages, median AFP, hCG, and uE3 values for
white, black, Hispanic, and other patients were all significantly
different. Differences remained significant even when data were
corrected for patient weights. For each analyte, the extent of the
variation was not the same at different gestational ages. Differences
in median values across race/ethnicity groups appear to have only a
small impact in Down syndrome screening but it may be appropriate to
use alternative sets of AFP medians or adjustment factors to AFP
medians for some Asian populations receiving ONTD screening.
Key Words: indexing terms: triple marker testing • maternal serum screening • Down syndrome • open neural tube defects
|
Introduction
|
|---|
Increased second-trimester maternal serum -fetoprotein
(AFP)1
is associated with fetal
open neural tube defects (ONTD), other congenital anomalies, fetal and
perinatal loss, low birth weight, and pregnancy complications
(1)(2). Low
concentrations of AFP together with increased human chorionic
gonadotropin (hCG) and low concentrations of unconjugated estriol (uE3)
are frequently found in women carrying Down syndrome fetuses, and
measurement of these three analytes (triple marker testing) is widely
used for second-trimester screening for high risk pregnancies
(3). A separate screening protocol identifies some cases
of trisomy 18, which is associated with low concentrations of AFP, hCG,
and uE3 (4).
Effective and accurate screening relies upon the establishment of the
median concentrations of AFP, hCG, and uE3 in pregnant women. AFP
concentrations are higher in black patients relative to white patients,
and an adjustment of 10–15% is generally made to allow for this
difference (5)(6)(7)(8). A number of studies have noted that hCG
concentrations were higher in black patients, although the magnitude of
the reported difference has varied considerably (9)(10)(11)(12)(13).
uE3 values have been reported to show little difference in black vs
white patients (9)(11)(12).
In this report we reinvestigate the racial or ethnic group differences
in AFP, hCG, and uE3 concentrations and note significant differences in
medians across race for all three analytes.
|
Materials and Methods
|
|---|
Data were collected from all AFP, hCG, and uE3 tests performed on
singleton, nondiabetic pregnancies at 15 to 23 weeks of gestation for
test referrals to our laboratory from April 15, 1993, to August 22,
1995. AFP testing involved 21 480 patients; hCG and uE3 measurement
was carried out for 21 218 women. During this interval all AFP and hCG
concentrations were measured by Hybritech Tandem E immunoassays
(Hybritech, San Diego, Ca), and uE3 was measured by a RIA (DSL,
Webster, TX). A cutoff of 2.0 multiples of the median (MoM) was used to
define the patients at increased risk for a fetus with an ONTD and a
mid-trimester risk of 1:270 was used as a cutoff to define high risk
for Down syndrome. Full details of the screening strategy and policies
have been published elsewhere (14).
All tests were performed in duplicate with variation (CV) not exceeding
10%. Within-assay and between-assay variation was monitored with
criteria based on those suggested by Westgard et al. (15)
with three control preparations (representing high, low, and middle
range values) used in each run. Long-term assay stability was evaluated
by monthly review of mean and median MoM values for each analyte
together with mean and range values of control preparations.
Four categories of maternal race or ethnicity were used: white, black,
Hispanic, and other (primarily Oriental and Indian). Grouping was based
on patient self-identification and the information provided to the
laboratory on the laboratory test request form. Median values for each
group were determined for each completed week of gestation. Weight
correction was applied by using the approach proposed by Knight et al.
(16) and Palomaki et al. (17). For each
race/ethnicity group, patients were sorted into 15-lb. weight
intervals, and median MoM values calculated for each interval.
Regressed values were computed by using log linear curves taking into
account the number of observations within each category. The
appropriateness of the weight correction equations (Table 1
) was verified by determining median corrected MoM values that
were close to 1.00 for each gestational week. The weight-corrected
concentration of each analyte for each patient was derived by applying
the formulas to the observed concentrations. Data for women
subsequently determined to carry fetuses with anomalies, patients with
poor pregnancy outcome or pregnancy complications, were not excluded
from the databases.
Descriptive statistics, median tests, and McNemar 
2
tests were carried out with the package from SPSS (Chicago, IL), with
P <0.05 considered statistically significant when testing
hypotheses.
|
Results
|
|---|
Table 2
summarizes raw median values for AFP, hCG, and uE3 for each of
the racial groups. By using the statistical test for medians,
significant differences across races could be demonstrated for AFP at
15 (2 = 27.63), 16 (2 = 49.65), 17
(2 = 30.38), and 20 (2 = 8.42) completed
weeks of gestation (3 df, P <0.05). For hCG,
significant differences in medians across races were noted at 15
(2 = 17.31), 16 (2 = 32.95), 17
(2 = 19.85), and 18 (2 = 10.19) weeks of
gestation. Similarly, for uE3, medians across races were significantly
different at 15 (2 = 24.59), 16 (2 =
107.87), 17 (2 = 71.60), and 18 (2 =
22.18) weeks of gestation. At higher gestational ages, differences were
not statistically significant, possibly because of the limited data
available.
View this table:
[in this window]
[in a new window]
|
Table 2. Median AFP, hCG, and uE3 concentrations (raw and
weight-corrected data) as a function of gestational
age.
|
|
There were significant differences in median patient weights for each
of four racial groups (median weight for whites 68.4 kg (150.7 lbs.),
blacks 73.7 kg (162.5 lbs.), Hispanics 65.5 kg (144.4 lbs.), and other
[57.9 kg (127.7 lbs.)] (2 = 364.00, 3
df, P <0.0001). Separate weight correction
factors were developed for each analyte and for each population (Table 1
) and a second set of data was established with patient weight
correction factors applied to each patient's analyte value.
Significant differences for medians for each of the analytes across
races could still be demonstrated after weight correction. For AFP,
significant differences were noted at 15 (2 = 29.59), 16
(2 = 43.81), 17 (2 = 44.72), and
18(2 = 33.62) weeks of gestation. For hCG, differences
were significant at 15 (2 = 16.71), 16
(2 = 37.79), 17 (2 = 18.67), and 18
(2 = 10.16) weeks. uE3 values were significantly
different at 15 (2 = 33.04), 16 (2 =
123.38), 17 (2 = 71.85), 18 (2 = 29.84),
19 (2 = 8.04), and 20 (2 = 9.71)
weeks.
Fig. 1
shows the regressed data for medians of the weight-corrected
concentrations for each racial group, relative to the values obtained
from white patients. AFP concentrations were higher in black patients
and that difference was approximately constant over the weeks that most
screening was performed (15–18 weeks). AFP in Hispanics also appeared
to be higher, but only at advanced gestational ages. For other patients
AFP appeared to be higher, particularly at later gestational ages. hCG
values in black, Hispanic, and other patients appeared to be higher
than those in white patients except at advanced gestational ages. uE3
values appeared to be higher in black, Hispanic, and other patients
during the time when most screening was performed but were lower in
black patients at advanced gestational age.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 1. AFP median values at 15–23 weeks of gestation
(A), hCG median values at 15–21 weeks of gestation
(B), and uE3 median values at 15–21 weeks of gestation
(C) for white, black, Hispanic, and other patients
(expressed relative to the value for white patients).
Values are based on the weight-corrected data and regressed with a log
linear (A, C) or an exponential (B)
model taking into account the number of observations at each
gestational week.
|
|
|
Discussion
|
|---|
On the basis of analysis of >21 000 samples, we observed that
median values for second-trimester maternal serum AFP, hCG, and uE3 all
differ between racial/ethnic groups and that the extent of the
variation is gestational age dependent. Significant differences in
analyte medians existed even after adjustments were made for patient
weight.
AFP values are higher in American black pregnant women, although the
magnitude of the difference appears to vary in different studies
(5)(6)(7)(8). Crandall et al. (5) noted slightly
lower AFP concentrations in Hispanics and somewhat higher AFP
concentrations in Orientals. Similarly, hCG values have been noted to
be higher in black patients (9)(10)(11)(12)(13) and possibly Orientals
(13). uE3 concentrations have been reported to be similar
in black and white women
(9)(11)(12). The differences
observed in the various studies may be attributable to the inconsistent
use of weight correction, heterogeneity in the gestational ages of the
screened populations, maternal age differences, and regional
differences in the subpopulations of women included in each group. The
reliability of gestational age determination may also be a source of
variability in establishing medians for different groups. Gestational
age determination on the basis of time from the last menstrual period
can show a bias towards overestimation (18), and the
preferential use of ultrasound biometry to determine gestational age in
some socioeconomic groups could lead to minor differences in median
values.
For Down syndrome screening there appears to be little overall clinical
significance to the differences in medians for the various
race/ethnicity groups. Changes in the Down syndrome risk with
race-specific medians should be relatively minor because of the
counterbalancing effect of adjustments to the different analytes in the
three-test combination. For example, consider a Hispanic patient
screened at 16 weeks with median values (1.00 MoM) of AFP, hCG, and uE3
when interpreted against a Hispanic patient-derived regressed database.
If a primarily white population-regressed database had been used, the
AFP value would have been 1.00 MoM, hCG value 1.06 MoM, and the uE3
value 1.12 MoM. The higher hCG MoM value would increase the Down
syndrome risk but the higher uE3 MoM value reduces risk. Similar
counterbalancing combinations occur for other race/ethnicity groups. A
precise estimation of the clinical significance associated with the
observed differences in medians would require a recalculation of all
patients' risks by using group-specific medians and an evaluation of
the revised detection rates and false-positive rates.
Of somewhat more concern is the impact for neural tube defect screening
and the possible need to use group-specific medians for AFP in
different populations. We observed higher AFP medians for patients
grouped as other (primarily Oriental and Indian), particularly at later
gestational ages. The observation is based on relatively few patients
but does appear to be consistent with the data of Crandall et al.
(5). The observed AFP medians for this group exceeded
those for blacks and it is usual to adjust AFP MoMs for the black
population. If confirmed, a correction for AFP concentration or the use
of group-specific medians for at least some Asian populations would be
appropriate.
|
Acknowledgments
|
|---|
We thank the Lowell P. Weicker Clinical Research Center,
University of Connecticut Health Center, for support (J.M.C.).
|
Footnotes
|
|---|
1 Nonstandard abbreviations: AFP, -fetoprotein; ONTD,
open neural tube defects; hCG, human chorionic gonadotropin; uE3,
unconjugated estriol; and MoM, multiple of the median. 
|
References
|
|---|
-
Knight GJ, Palomaki GE. Maternal alpha-fetoprotein and the detection of open neural tube defects. Elias S Simpson JL eds. Maternal serum screening for fetal genetic disorders 1992:41-58 Churchill Livingstone New York. .
-
Burton BK. Unexplained elevated maternal serum alpha-fetoprotein and adverse perinatal outcome. Elias S Simpson JL eds. Maternal serum screening for fetal genetic disorders 1992:109-119 Churchill Livingstone New York. .
-
Wald NJ, Cuckle HS, Densem JW, et al. Maternal serum screening for Down's syndrome in early pregnancy. BMJ 1988;297:883-887.
-
Canick JA, Palomaki GE, Osathanondh R. Prenatal screening for trisomy 18 in the second trimester. Prenat Diagn 1990;10:546-548.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Crandall BF, Lebherz TB, Schroth PC, Matsumoto M. Alpha-fetoprotein concentrations in maternal serum: relation to race and body weight. Clin Chem 1983;29:531-533.
[Abstract/Free Full Text]
-
Johnson AM. Racial differences in maternal serum alpha-fetoprotein screening. Mizejewski GJ Porter IH eds. Alpha-fetoprotein and congenital disorders 1985:183-196 Academic Press New York. .
-
Baumgarten A. Racial difference and biological significance of maternal serum alpha-fetoprotein. Lancet 1986;ii:573.
-
Macri JN, Kasturi RV, Hu MG, Krantz DA, Douros TJ, Sajda P, Cook EJ. Maternal serum -fetoprotein screening. III. Pitfalls in evaluating black gravid women. Am J Obstet Gynecol 1987;157:820-822.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Simpson JL, Elias S, Morgan CD, Shulman L, Umstot E, Anderson RN. Second trimester maternal serum human chorionic gonadotropin and unconjugated oestriol levels in blacks and whites. Lancet 1990;335:459-460.
-
Muller F, Boue A. A single chorionic gonadotropin assay for maternal serum screening for Down's syndrome. Prenat Diagn 1990;10:389-398.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Canick JA, Panizza DS, Palomaki GE. Prenatal screening for Down syndrome using AFP, uE3 and hCG: effect of maternal race, insulin-dependent diabetes and twin pregnancy. Am J Hum Genet 1990;47:A270.
-
Burton BK, Nieb B. Effect of maternal race and weight (wt) on hCG and uE3 levels in the mid-trimester. Am J Hum Genet 1991;49:A212.
-
Bogart MH, Jones OW, Felder RA, et al. Prospective evaluation of maternal serum human chorionic gonadotropin levels in 3428 pregnancies. Am J Obstet Gynecol 1991;165:663-667.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Benn PA, Horne D, Briganti S, Greenstein RM. Prenatal diagnosis of diverse chromosome abnormalities in a population of patients identified by triple-marker testing as screen positive for Down syndrome. Am J Obstet Gynecol 1995;173:496-501.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Westgard JO, Barry PL, Hunt MR, Groth T. A multi-rule Shewhart chart for quality control in clinical chemistry. Clin Chem 1981;27:493-501.
[Free Full Text]
-
Knight GJ, Palomaki GE, Haddow JE. Use of maternal serum alpha-fetoprotein measurements to screen for Down's syndrome. Clin Obstet Gynecol 1988;31:306-327.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Palomaki GE, Panizza DS, Canick JA. Screening for Down syndrome using AFP, uE3, and hCG: effect of maternal weight. Am J Hum Genet 1990;47:A282.
-
Wald N, Cuckle H, Borcham J. Effect of estimating gestational age by ultrasound cephalometry on the specificity of -fetoprotein screening for open neural-tube defects. Br J Obstet Gynaecol 1982;89:1050-1053.
[Web of Science][Medline]
[Order article via Infotrieve]
The following articles in journals at HighWire Press have cited this article:
|
|
|
|
 
A. A. Arslan, A. Zeleniuch-Jacquotte, A. Lukanova, Y. Afanasyeva, J. Katz, M. Levitz, G. Del Priore, and P. Toniolo
Effects of Parity on Pregnancy Hormonal Profiles Across Ethnic Groups with a Diverse Incidence of Breast Cancer.
Cancer Epidemiol. Biomarkers Prev.,
November 1, 2006;
15(11):
2123 - 2130.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G Vranken, T Reynolds, and J Van Nueten
Medians for second-trimester maternal serum markers: geographical differences and variation caused by median multiples-of-median equations.
J. Clin. Pathol.,
June 1, 2006;
59(6):
639 - 644.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
