Prenat Diagn 2005; 25: 47–50.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/pd.1082
Down syndrome screening marker levels in women with a
previous aneuploidy pregnancy
Howard S. Cuckle1*, K. Spencer2,3and K. H. Nicolaides3
1Reproductive Epidemiology, University of Leeds, Leeds, UK
2Clinical Biochemistry Department, Harold Wood Hospital, Romford, Essex, UK
3Harris Birthright Centre for Fetal Medicine, King’s College Hospital, London, UK
assumed to have the same marker distribution as those without a family history. This assumption needs to be
In Down syndrome screening programmes, women with a previous affected pregnancy are
in three centres, Leeds, Romford and the Fetal Medicine Centre, London. For each woman with a previous
aneuploidy (case), five unaffected pregnancies to women without a history were selected as controls. The
markers tested included maternal serum free β-human chorionic gonadotrophin (hCG), pregnancy-associated
plasma protein A (PAPP-A), α-fetoprotein, unconjugated estriol and ultrasound nuchal translucency thickness.
Information on previous aneuploidy pregnancies was routinely sought on the test request forms
Patau’s syndrome. There was a statistically significant difference between cases and controls, in the distribution
of free β-hCG and PAPP-A levels, adjusted for gestation. On average, free β-hCG was increased by 10% in
a subsequent pregnancy after aneuploidy (p < 0.005, Wilcoxon rank sum test) and for PAPP-A the increase
was 15% (p < 0.0001). No other marker was significantly different.
There were 375 cases: 303 with previous Down syndrome, 63 with Edwards syndrome and 9 with
Until data from sufficient affected pregnancies are available for study, it would be prudent to assume that the
same increase as in unaffected pregnancies applies. Copyright 2005 John Wiley & Sons, Ltd.
Risk calculation algorithms need to be modified to take account of the increased marker levels.
KEY WORDS: Down syndrome; screening; markers; previous aneuploidy
The standard method of interpreting a Down syndrome
screening test is to estimate the risk of an affected
pregnancy from the maternal age-specific risk and the
marker profile. If the woman has had a previous preg-
nancy with aneuploidy, the normal age-specific risk will
not apply. In a small proportion of cases, there will
be a parental structural chromosome rearrangement and
the recurrence risk can be quite high, depending on
the specific parental genotype. In the vast majority of
cases, the risk is more modest and can be readily cal-
culated by adding a fixed amount to the age-specific
risk, expressed as a percentage. One estimate for women
with a previous Down syndrome pregnancy is the addi-
tion of 0.77%, 0.54% and 0.42% for risks calculated in
the first trimester, second trimester and at term respec-
tively (Cuckle and Arbuzova, 2004). Similar corrections
apply for Edwards and Patau’s syndrome (Nicolaides
et al., 1999). Once the correct age-specific risk has been
obtained, the screening-related risk is usually calculated
in the same way as any other pregnancy. However, this
assumes that the marker distributions are the same in
*Correspondence to: Prof. Howard S. Cuckle, Reproductive
Epidemiology, Leeds Screening Centre, Gemini Park, Leeds
women with and without a previous aneuploidy preg-
nancy, and we now have evidence that this is not the
MATERIALS AND METHODS
Information on previous aneuploidy pregnancies is rou-
tinely sought on the Down syndrome screening test
request forms used in Leeds, Romford and the Fetal
Medicine Centre (FMC), London. In Leeds, those
requesting either first or second trimester tests are asked
about previous pregnancies with Down syndrome or
Edwards syndrome. In Romford and FMC, the request
is only made in the first trimester, but information about
a previous Patau’s syndrome is also sought.
In Leeds, women screened in both the first and second
trimester are tested for maternal serum free β-human
chorionic gonadotrophin (hCG), α-fetoprotein (AFP)
and unconjugated estriol (uE3) using time-resolved flu-
orescent assay (DELFIA, Perkin-Elmer Life Sciences,
Turku, Finland). In addition, all first-trimester screen-
ing tests included pregnancy-associated plasma pro-
tein A (PAPP-A) measurement (DELFIA) and ultra-
sound nuchal translucency (NT) thickness. Among those
screened in the second trimester, a proportion also had
inhibin measured, but there were too few with a previ-
ous aneuploidy to be included in the study. In Romford
and FMC, first-trimester screening tests are based on
Copyright 2005 John Wiley & Sons, Ltd.
Received: 15 July 2004
Revised: 4 October 2004
Accepted: 25 October 2004
H. S. CUCKLE ET AL.
maternal serum free β-hCG and PAPP-A, measured by
time-resolved amplified cryptate emission (Brahms AG,
Berlin, Germany), together with NT.
All serum markers were expressed in multiples of
the median for the appropriate gestation (MoMs) and
corrected for maternal weight. The NT values from
Leeds were also expressed in MoMs using either the
general median equation published by Nicolaides et al.
(1998), or, more recently, centre-specific medians (Log-
ghe et al., 2003), whereas those from Romford and
FMC were expressed in deviations from the gestation-
specific median (delta-values), as outlined by Spencer
et al. (2003) using the general equation.
A total of 375 unaffected pregnancies to women with
a previous aneuploidy were available for study. For
each case, five pregnancies to women without previous
aneuploidy were selected as controls. Matching criteria
were same age, within 2 years, and same period of
testing, within 4 weeks.
The non-parametric Wilcoxon rank sum test was used
to make comparisons between the distribution of marker
levels in cases and controls.
Table 1 shows the median marker level in cases and
controls. There was a statistically significant increase in
the level of both free β-hCG and PAPP-A among women
who had a previous affected pregnancy compared to
other women. Overall, the increase was 10% for free
β-hCG and 15% for PAPP-A.
The PAPP-A result was significant in each cen-
tre separately and for each type of previous ane-
uploidy. The median PAPP-A level was higher in
Edwards and Patau’s than in Down syndrome, but there
was no statistically significant heterogeneity (p = 0.19,
The free β-hCG result was significant in Leeds, but
not in Romford and FMC, and for Down and Edwards,
but not for Patau’s syndrome. As for PAPP-A, the
median was higher in Edwards and Patau’s than in
Down syndrome, but there was no statistically significant
heterogeneity (p = 0.54).
In Leeds, 146 of the cases had free β-hCG tests in
the first trimester and 46 in the second trimester. The
medians were 1.17 MoM (p < 0.01) and 1.10 MoM
(p = 0.82), respectively. There was insufficient data to
conclude about any possible interaction between the
effect and trimester.
No material difference between cases and controls was
found for any of the other markers. Moreover, when
the Leeds AFP and uE3results were compared within
trimester, no significant differences were found.
In addition to the median level, the spread of free
β-hCG also differed between cases and controls. The
standard deviation of log10 values, estimated from the
10 to 90th centile range, was 0.281 in cases and 0.254
in controls (p < 0.01, F-test for the variance ratio). In
contrast, the spread of PAPP-A was similar in cases and
controls, with standard deviations of 0.233 and 0.236
We have shown that free β-hCG and PAPP-A levels
are increased on average in unaffected pregnancies to
women with a previous affected pregnancy.
With existing risk calculation algorithms the free β-
hCG increase will lead to over-estimated risks, whilst
the PAPP-A increase will reduce risks. A simple way
of correcting this would be to divide the free β-hCG
level by 1.10 and PAPP-A by 1.15, then calculate
risks in the same way as usual, assuming that the
proportional increase in levels applies equally to affected
and unaffected pregnancies. Whilst there is as yet
insufficient data to test this assumption, experience with
other screening co-variables indicates that it is likely to
When both markers are measured, as in first-trimester
screening, whilst the two effects will tend to cancel
each other out, a corrected risk will be more accurate
than one that is not corrected. For example, a 36-year-
old woman with a previous affected pregnancy, 1.5
MoM free β-hCG and 1.0 MoM PAPP-A at 10 weeks,
and 1.5 MoM NT at 11 weeks’ gestation has a 1 in
260 Down syndrome risk at term, using the parameters
in Cuckle and Arbuzova (2004). Correcting the serum
levels increases the risk to 1 in 240. When only one
marker is measured, as in second-trimester screening, the
correction would be more important, although our study
included too few second-trimester cases to be certain
that the magnitude of the free β-hCG effect is the same
in both trimesters.
One potential bias in our results is that cases are by
definition parous, whereas controls could be nulliparous,
and both hCG and PAPP-A levels vary with parity or
gravidity (for a review, see Wald et al., 1998; Spencer
et al., 2000). However, this is unlikely to be a large
effect since, by matching on maternal age, we will
have minimised the parity difference, and the association
between the markers and parity is weak. Information on
parity has been routinely sought on the test request form
in Leeds since early 1996, so the increase in levels could
be examined for cases tested in this period compared to
their parous controls. The free β-hCG increase for 157
cases was 9% and the PAPP-A increase for 151 cases
Maternal smoking also affects free β-hCG and PAPP-
A levels (Wald et al., 1998; Spencer, 1998, 1999; De
Graaf et al., 2000; Spencer et al., 2004), and several
studies have reported that smoking is less common in
the mothers of infants with Down syndrome. However,
a meta-analysis of all published studies to date has
concluded that smoking does not confer a reduced risk
(Rudnicka et al., 2002), so this will not have biased our
Whilst maternal serum free β-hCG and PAPP-A are
placental products, there is a maternal contribution to
the level in an individual pregnancy. This is clear from
the between-pregnancy correlations reported for total
hCG (Holding and Cuckle, 1994; Dar et al., 1996; Wax
et al., 2000), free β-hCG (Spencer, 1997, 2001; Spencer,
2002) and PAPP-A (Spencer, 2001; 2002). Maternal
factors controlling the production or metabolism of these
Copyright 2005 John Wiley & Sons, Ltd.Prenat Diagn 2005; 25: 47–50.
SCREENING MARKER LEVELS IN PREVIOUS ANEUPLOIDY
Table 1—Median marker level in cases with previous aneuploidy and controls (MoMs except for NT in Romford/FMC, which
is expressed in delta units)
Previous pregnancy with syndrome
— 1.15 (192)a
— 1.01 (91)
0.98 (144)1.04 (48)— 0.99 (192)1.00 (960)
Statistical significance compared with controls (p):a< 0.05;b< 0.01;c< 0.005;d< 0.001;e< 0.0001.
analytes probably also impact on the increased levels in
women with a previous aneuploidy.
Altered levels of several other analytes have been
found in blood samples from women with previous
aneuploidy pregnancies: neutrophil alkaline phosphatase
(Vergnes et al., 1988), thyroid antibodies (Cuckle et al.,
1998), Cu, Zn-superoxide dismutase (Arbuzova, 1998),
follicle stimulating hormone (van Montfrans et al., 1999;
Nasseri et al., 1999) and homocystine (James et al.,
1999). It is possible that the same biological mechanism
underlies both these and the changes we have observed.
However, not all of these studies allowed fully for
maternal age, as we have done, which may have
confounded their results.
Arbuzova S. 1998. Why it is necessary to study the role of the
mitochondrial genome in trisomy 21 pathogenesis? Downs Syndr
Res Pract 5(3): 26–29.
Cuckle H, Arbuzova S 2004. Multianalyte maternal serum screening
for chromosomal defects. In Genetic Disorders and the Fetus:
Diagnosis, Prevention and Treatment, (5th edn), Milunsky Aubrey.
(ed). Johns Hopkins University Press: Baltimore; 795–835.
Cuckle HS, Wald N, Stone R, Densem J, Haddow J, Knight G. 1998.
Maternal serum thyroid antibodies in early pregnancy and fetal
Down’s syndrome. Prenat Diagn 8: 439–445.
Dar H, Merksamer R, Berdichevsky D, David M. 1996. Maternal
serum markers levels in consecutive pregnancies: a possible genetic
predisposition to abnormal levels. Am J Med Genet 61(2): 154–157.
De Graaf IM, Cuckle HS, Pajkrt E, Leschot NJ, Bleker OP, Van
Lith JMM. 2000. Co-variables in first trimester maternal serum
screening. Prenat Diagn 20(3): 186–189.
Holding S, Cuckle HS. 1994. Maternal serum screening for Downs
syndrome taking account of the result in a previous pregnancy.
Prenat Diagn 14: 321–322.
James SJ, Pogribna M, Pogribny IP, et al. 1999. Abnormal folate
reductase gene may be maternal risk factors for Down syndrome.
Am J Clin Nutr 70: 495–501.
Logghe H, Cuckle H, Sehmi I. 2003. Centre-specific ultrasound
nuchal translucency medians needed for Down’s syndrome
screening. Prenat Diagn 23: 389–392.
Nasseri A, Mukherjee T, Grifo JA, Noyes N, Krey L, Copper-
man AB. 1999. Elevated day 3 serum follicle stimulating hormone
and/or estradiol may predict fetal aneuploidy. Fertil Steril 71(4):
Nicolaides KH, Sebire NJ, Snijders RJM. 1999. The 11–14 week
scan. The diagnosis of fetal abnormalities, Parthenon: New York.
Nicolaides KH, Snijders RJ, Cuckle HS. 1998. Correct estimation of
parameters for ultrasound nuchal translucency screening. Prenat
Diagn 18(5): 519–523.
Rudnicka AR, Wald NJ, Huttly W, Hackshaw AK. 2002. Influence of
maternal smoking on the birth prevalence of Down syndrome and
on second trimester screening performance. Prenat Diagn 22(10):
maternal serum alpha-fetoprotein and free beta hCG: implications
for Down syndrome screening in subsequent pregnancies. Prenat
Diagn 17(1): 39–45.
Spencer K. 1998. The influence of smoking on maternal serum AFP
and free beta hCG levels and the impact on screening for Down’s
syndrome. Prenat Diagn 18: 225–234.
Spencer K. 1999. The influence of smoking on maternal serum PAPP-
A and free beta hCG levels in the first trimester of pregnancy.
Prenat Diagn 19(11): 1065–1066.
Spencer K. 2001. Between pregnancy biological variability of first
trimester markers of Down syndrome: implications for screening in
subsequent pregnancies. Prenat Diagn 21: 445–447.
Spencer K. 2002. Between pregnancy biological variability of first
trimester markers of Down syndrome and the implications for
screening in subsequent pregnancies: an issue revisited. Prenat
Diagn 22: 874–876.
Spencer K, Bindra R, Cacho AM, Nicolaides KH. 2004. The impact
of correcting for smoking status when screening for chromosomal
anomalies using maternal serum biochemistry and fetal nuchal
translucency thickness in the first trimester of pregnancy. Prenat
Diagn 24: 169–173.
Spencer K,Ong CYT,Liao AWJ,
influence of parity and gravidity on first trimester markers of
chromosomal abnormality. Prenat Diagn 20: 792–794.
Spencer K, Bindra R, Nix ABJ, Heath V, Nicolaides KH. 2003. Delta
NT or NT MoM: Which is the most appropriate method for
calculating patient specific risks for trisomy 21 in the first trimester?
Ultrasound Obstet Gynecol 22: 142–148.
van Montfrans JM, Dorland M, Oosterhuis GJ, van Vugt JM, Rekers-
Mombarg LT, Lambalk CB. 1999. Increased concentrations of
Copyright 2005 John Wiley & Sons, Ltd. Prenat Diagn 2005; 25: 47–50.
H. S. CUCKLE ET AL.
follicle-stimulating hormone in mothers of children with Down’s
syndrome. Lancet 353: 1853–1854.
Vergnes H, Grozdea J, Brisson-Lougarre A, et al. 1988. An enzymatic
marker in mothers of trisomy 21 children: neutrophil alkaline
phosphatase. Enzyme 39(3): 174–180.
Wald NJ, Kennard A, Hackshaw A, McGuire A. 1998. Antenatal
screening for Down’s syndrome. Health Technol Assess 2(1): i–iv,
Wax JR, Lopes AM, Benn PA, Lerer T, Steinfeld JD, Ingardia CJ.
2000. Unexplained elevated midtrimester maternal serum levels
of alpha fetoprotein, human chorionic gonadotropin, or low
unconjugated estriol: recurrence risk and association with adverse
perinatal outcome. J Matern Fetal Med 9(3): 161–164.
Copyright 2005 John Wiley & Sons, Ltd.Prenat Diagn 2005; 25: 47–50.