J. Perinat. Med.
2015; 43(4): 403–408
Gilles Faron*, Ronald Buyl and Walter Foulon
Does recent sexual intercourse during pregnancy
affect the results of the fetal fibronectin rapid
test? A comparative prospective study
Abstract
Objective: We conducted a prospective comparative cohort
study to determinate the influence of coitus on quantita-
tive fetal fibronectin test results under normal pregnancy
conditions. We also compared values obtained in cervical
and vaginal secretions.
Methods: In a population of women with normal sin-
gleton pregnancies between 22 and 28 weeks gestation,
we have performed (cervical and vaginal) quantitative
fetal fibronectin tests in two separate groups classified
according to timing after coitus (one group of women had
intercourse within 24  h before sampling and the control
group had intercourse  > 24 h before sampling). The main
outcome measures were the proportion of positive tests
in both groups and the correlation between cervical and
vaginal values through the Pearson correlation coefficient.
Results: Both groups were similar in terms of general
characteristics and pregnancy outcomes. The propor-
tions of positive results in the vaginal secretions were
7.5% and 25.0% (P = 0.007) in the control and coitus group,
respectively. In the cervical secretions, the proportions of
positive tests were greater, but not statistically different
(39.7% and 40.0%, respectively). The Pearson correlation
coefficients were very low ( < 0.3) indicating poor correla-
tion between both sampling locations. Even if the cervi-
cal values were generally greater than the vaginal values,
they were lower in 26% of the women.
Conclusions: Coitus definitely interferes with vaginal fetal
fibronectin test results. In cervical secretions, the positive
rate was so high that coitus had no influence, but cervical
sampling in this location should be avoided.
*Corresponding author: Gilles Faron, Department of Obstetrics, UZ
Brussel (Vrije Universiteit Brussel), Laarbeeklaan 101, 1090 Jette,
Brussels, Belgium, Tel.: +3224763682, Fax: +3224776790,
E-mail: gilles.faron@uzbrussel.be
Ronald Buyl: Faculty of Medicine and Pharmacy, Department of
Biostatistics and Medical Informatics, Vrije Universiteit Brussel,
Brussels, Belgium
Walter Foulon: Department of Obstetrics, Universitair Ziekenhuis
Brussel, Vrije Universiteit Brussel, Brussels, Belgium
Unauthenticated
Download Date | 6/27/16 6:45 AM
Keywords: Coitus; fetal fibronectin test; pregnancy;
sexual intercourses.
DOI 10.1515/jpm-2014-0127
Received April 15, 2014. Accepted May 19, 2014 . Previously published
online June 14, 2014.
Introduction
Spontaneous preterm delivery (sPTD) is a worldwide
problem with respect to medical and socioeconomic
burdens [16]. Attempting to predict women at risk for sPTD
early in pregnancy with non-invasive and accurate tools
is the first step to beneficial interventions [7]. The fetal
fibronectin (ffn) test is the most common biochemical
marker used in this context, but the low positive predic-
tive value is a concern [14]. Disturbed vaginal flora [13],
cervical bleeding, or the presence of sperm are conditions
suspected to interfere with test results. A previous paper
has indeed suggested that recent coitus ( < 72 h) may yield
false-positive results because of the presence of an isoform
of fibronectin in the seminal plasma; however, the study
was neither conducted with pregnant women nor with the
current quantitative test, and it had a low power for cor-
relating ffn test results and time between intercourse and
sampling. Moreover, some women were tested more than
once, so the data were not independent and that leads to
false proportions [15]. Finally, the best site to sample is still
unclear based on the literature; it has been suggested that
sampling should be from the ectocervix [11], the fornix [1],
or both interchangeably [6, 9]. Consequently, we under-
took a new prospective comparative study in an actual
clinical setting to challenge the efficacy of ffn testing in
cervical and vaginal fluid secretions, and a short or longer
time after coitus.
Methods
This was a prospective, non-randomized study conducted in ante-
natal consultation at the Free University of Brussels (UZ Brussel)
404      Faron et al., Does recent sexual intercourse during pregnancy affect the results of the fetal fibronectin rapid test?

between January 2011 and November 2013. Healthy women with
uncomplicated and singleton pregnancies were invited to partici-
pate. Gestational age had to be ascertained based on the first day of
the last menstrual cycle and a first trimester ultrasound. We provided
oral information about the study applying to approximately 20 weeks
gestation and a brochure was given to the woman.
Women were asked to have coitus around the next consultation,
and preferably within the 72 h before. If women reported more than
one episode of coitus within this 72-h period, only the hour of the
last episode of coitus was taken into account. When the patient did
not have recent intercourse ( < 24 h), for any reason, but still agreed
to participate in the study, she was placed in the control group. The
study was conducted according to the declaration of Helsinki, writ-
ten informed consent was obtained from those women agreeing to
participate in the study, which was approved by the hospital Ethics
committee. The enrolled patients were tested once between 22+0 and
28+6 weeks gestation. Two groups were created according to the tim-
ing of the last coitus (a control group with intercourse  > 24 h before
sampling, and a coitus group with intercourse within 24  h of sam-
pling). We chose 24 h as cut-off because it is the most generalized cut-
off chosen in the literature concerning ffn test studies, but we have
also looked at ffn values after 36, 48, and 72  h and more to have a
continuum of values according to time after coitus. Exclusion criteria
included vaginal bleeding, placenta previa, cerclage, clinical vaginal
infection, prior history of conisation, and rupture of membranes with
regular contractions.
rupture of membranes, and steroid administration), and the birth
weight were also noted.
The main objective of this study was to compare the propor-
tion of positive vaginal and cervical ffn tests in both groups. On the
basis of an expected proportion of positive vaginal ffn tests of 3%
between 22 and 28 weeks gestation among a normal pregnant popu-
lation [9] and an increase to 22% of positive tests 24 h after coitus as
suggested elsewhere [15], we calculated that 60 women per group
would be required to demonstrate a statistically significant differ-
ence between the two groups (assuming power of 80% and a α error
of 0.01).
As a secondary outcome, we have studied the evolution of con-
centration values of ffn in vaginal and cervical secretions according
to the timing of the last coitus. The different positive and negative
rates between cervical and vaginal tests were compared.
The correlation between the cervical and vaginal concentrations
of ffn in general and in both groups was sought graphically and with
determination of the Pearson correlation coefficient.
All calculations were carried out using IBM (New York, USA) SPSS
software (version 22.0). Baseline data have been summarized for each
group and compared (comparison of two proportions or two means
and 95% confidence interval (CI)). Unpaired t-tests, χ2 for trends, two-
sided Fisher’s exact test, and the 95% CI were used when appropriate.
A P-value  < 0.05 was considered to be statistically significant.

During the speculum examination, two separate Dacron swabs
of both cervical (maximum, 1 cm inside the cervix) and vaginal (for-
nix) secretions were collected for quantitative dosage of ffn. Sam-
pling was obtained gently to avoid (cervical) bleeding. Samples for
ffn were placed in a preservative buffer in a tube and immediately
frozen at –18°C until analysis, which was performed within 1 month.
The concentration of ffn was determined with the use of a quanti-
tative enzyme-linked immunosorbent-assay (Rapid ffn 10Q analyzer
test®; Hologic, Bedford, MA, USA). A concentration  > 50 ng/mL was
considered positive; the maximum quantification using this tech-
nique was limited to 500 ng/mL.
Secondarily, we made a clinical evaluation of the cervix via a
vaginal examination with a gloved index finger.
The clinical data included maternal age, ethnicity, parity, and
history of preterm birth. The term (in weeks) and mode of delivery, all
events related to preterm labor (tocolysis, hospitalization, premature
Results
We included 60 women in the coitus group and 80 women
in the control group. None were lost during follow-up. We
encountered technical problems with the dosage of two
cervical ffn specimens, thus we only had seventy-eight
cervical results in the control group.
The general characteristics and pregnancy outcomes
are shown in Table 1. Both groups were similar; there were
no differences in age, parity, ethnicity, history of sPTD, and
pregnancy outcome, except for the mean time between
last coitus and sampling, which was clearly different with
a ratio of 1:26. The range interval between intercourse and

Table 1 General characteristics and pregnancy outcome.

  Group control   Group coitus   P-value

(n = 80) (n = 60)

Maternal age in years, mean ± SD   30.5 ± 5.8   30.7 ± 4.7   0.82a

Nulliparous, n (%)   37 (46.3)   27 (45.0)   1.00b
Ethnicity (Caucasian/others, n)   67/13   53/7   0.48b
History of PTB (n)   1   5   0.08b
Time between coitus and sampling in hours  ± SD   366 ± 928.2   14 ± 4.4  
Gestational age at sampling in weeks, mean ± SD   25.3 ± 1.8   25.0 ± 1.9   0.34a
Gestational age at delivery in weeks, mean ± SD   39.1 ± 1.7   39.0 ± 2.2   0.76a
Birth weight in kg, mean ± SD   3305 ± 544   3397 ± 570   0.34a
sPTB( < 36 w) (n)   1   2   0.58b
Primary cesarean delivery, n (%)   7 (8.8)   5 (8.3)   1.00b
Induction, n (%)   26 (32.5)   26 (43.3)   0.21b

SD = standard deviation, PTB = preterm birth, aunpaired t-test, bFisher’s exact test.

Unauthenticated
Download Date | 6/27/16 6:45 AM
 Faron et al., Does recent sexual intercourse during pregnancy affect the results of the fetal fibronectin rapid test?      405

sampling was 3–24 h in the coitus group and 26–4344 h in Table 3 Qualitative comparative results between cervical and
the control group. vaginal tests (discrepancy).

In this select low-risk population, eight (5.7%) women

delivered before 36 weeks gestation (three and five in the
coitus and control groups, respectively), three after spon-
taneous labor (giving a sPTD rate of 2.1%), and five for
medical indications (fetal or maternal). The cervical and
vaginal values of ffn for the three women who had sPTDs
were 442 and 159 ng/mL, 28 and 44 ng/mL, and 46 and
103 ng/mL, respectively, indicating that one woman was
positive for both tests, one was negative for both tests,
and the third woman was positive for the vaginal test only.
Two women belonged to the coitus group and one woman
belonged to the control group but this was not statistically
significant (P = 0.58).
The main results are summarized in Table 2. No patient
had an abnormal cervical examination and no patients
received intravenous tocolysis; only three patients needed
corticoprophylaxis (one in the coitus group and two in the
control group; P = 0.58).
Approximately 40% of pregnant women between 22
and 28 weeks gestation in this select low-risk population
had  > 50 ng/mL of ffn in their cervical secretions, whether
or not they had recent intercourse. The mean value of ffn
was higher in cervical fluid than vaginal secretions, but
even if the ffn cervical value was generally higher in the
cervix than the fornix, for 36/138 women (26.1%), the
opposite relationship existed (11/60 women in the control
group and 25/78 women in the coitus group (P = 0.08) had
lower cervical ffn values than vaginal ffn values).
Moreover, if we considered the manufacturer’s pro-
posed cut-off of 50 ng/mL for both vaginal and cervical
samples, the number of qualitative discrepant results
between vaginal and cervical tests was 15 in the coitus
group and 28 in the control group (P = 0.27), for a total of
43/138 discordant results (31.2%; Table 3).
The evolution of ffn values according to the dura-
tion after coitus is shown in Figures 1 and 2. The propor-
tion of positive tests within 12, 13–24, 25–36, 37–48, and
49–72 h was 9/23, 6/37, 2/10, 1/10, and 0/13 in the vaginal
secretions (Table 4) and 14/23, 10/37, 3/9, 4/10, and 4/13
  Group control  Group coitus
–cerv ffn/–vag ffn, n(%)   46 (58.9)  33 (55.0)
+cerv ffn/+vag ffn, n (%)   4 (5.1)  12 (20.0)
+cerv ffn/–vag ffn, n (%)   27 (34.6)  12 (20.0)
–cerv ffn/+vag ffn, n (%)   1 (1.3)  3 (5.0)
ffn = fetal fibronectin, cerv = cervical, vag = vaginal.
in the cervical secretions (χ2 for trends, 7.41 (P = 0.007)
and 1.68 (P = 0.20), respectively). Only 4/70 (5.7%) of the
vaginal samplings were positive in women who had inter-
course  > 36 h before sampling (vs. 17/70  < 36 h before sam-
pling, P = 0.004).
The Pearson correlation coefficient between cervi-
cal and vaginal concentrations of ffn was 0.174 for all
women with vaginal values  < 200 ng/mL and ffn cervi-
cal values  < 300 ng/mL, 0.156 for women belonging to
the control group, and 0.267 for women belonging to the
coitus group.
Discussion
Fetal fibronectin plays an important role in mediating
placental-uterine attachment at the chorio-decidual
interface [3]. Fetal fibronectin is normally present in the
vaginal secretions during the first trimester and after
34 weeks’ gestation, but seldom during the interval [4].
As the sub-clinical first steps of preterm labor involve a
disruption of this interface secondary to contractions or
aggression by microorganisms, the leakage of ffn in the
cervicovaginal fluid allows a test to be used based on its
dosage to screen for sPTD. Yet, the use of the ffn test is not
generalized because of its low predictive value in low-risk
populations. The high negative predictive value ( > 99%)
in women at risk for sPTD presenting with contractions is
particularly relevant; such women can avoid futile trans-
fer to tertiary care center and/or treatment [12].

Table 2 Fetal fibronectin results.

  Control group  Coitus group  P-value

Positive vaginal ffn, n (%)   6 (7.5)  15 (25.0)  0.007a

Vaginal ffn value (ng/mL), mean (95% CI)   20.2 (5.7–34.8)  68.3 (37.0–99.6)  0.003b
Positive cervical ffn, n (%)   31 (39.7)  24 (40.0)  1.00a
Cervical ffn value (ng/mL), mean (95% CI)   71.7 (46.8–96.5)  91.9 (59.1–124.7)  0.31b

ffn = fetal fibronectin, aFisher’s exact test, bunpaired t-test.

Unauthenticated
Download Date | 6/27/16 6:45 AM
406      Faron et al., Does recent sexual intercourse during pregnancy affect the results of the fetal fibronectin rapid test?

Vaginal ffn
>250

200

150
Vaginal ffn

>24 h
100 ≤ 24 h

50

0
0 100 200 300 400 >500
Time since coitus

Figure 1 Vaginal ffn values according to time after coitus.

Cervical ffn
>250

200

150
Cervical ffn

>24 h
100 ≤24 h

50

0
0 100 200 300 400 >500
Time since coitus

Figure 2 Cervical ffn values according to time after coitus.

Sexual intercourse is not a risk factor for sPTD and
coitus is not discouraged, even among women at risk [17].
Based on the only study involving interference due the
Table 4 Evolution of the proportion of vaginal ffn positive tests
according to time after coitus.
  Proportion of positive  
vaginal ffn
  ≤  12 h   9/23 (39.1%)  
12–  ≤  24 h  6/37 (16.2%)  
24–  ≤  36 h  2/10 (20.0%)  
36–  ≤  48 h  1/10 (10.0%)  
48–72 h   0/13 (0.0%)  
 > 72 h   3/47 (6.4%)  
ffn = fetal fibronectin.
presence of semen on ffn test results, the manufacturer
suggests caution in interpreting positive results if inter-
course has occurred in the previous 24 h. This study was
published in 1998 and the authors had used a different
assay than is currently used. Samples were successively
obtained from a small population of fourteen non-preg-
nant women, so interpretation of results must be put into
Cumulative (%) question because the concentrations of ffn were interde-
pendent [15]. We also have no data regarding the inter-
action or adherence of ffn with the vaginal epithelium
9/23 (39.1%)
15/60 (25.0%) during or outside the pregnancy, whereas the vaginal
17/70 (24.3%) ecosystem is modified during pregnancy [10]. The authors
18/80 (22.5%) stated that interference of coitus persisted 72 h, but the
18/93 (19.4%) manufacturer continues to propose to wait only 24 h after
21/140 (15.0%)
coitus to perform the test. All such questionable prelimi-
nary data need to be confirmed in a new prospective and

Unauthenticated
Download Date | 6/27/16 6:45 AM
 Faron et al., Does recent sexual intercourse during pregnancy affect the results of the fetal fibronectin rapid test?      407
comparatively larger study conducted in pregnant women
with independent samples and with the last ffn test kit
available on the market. We have chosen 24  h as cut-off
for two reasons (according to the manufacturer’s recom-
mendation brochure and in accordance with the previous
data and expected proportion of approximately 20% posi-
tive tests within the 24 h after coitus, which makes us able
to calculate the sample size needed).
Although not randomized, this prospective study gives
a strong recommendation with high-quality evidence not
to use the ffn test, not only after 24 h as generally recom-
mended, but within 36 h after sexual intercourse (GRADE
for clinical guidelines: 1A) [5]. In the clinical setting at the
time of admission, this information must be taken into
account before sampling. Recent coitus affects the results of
the test when sampling is obtained from the fornix, increas-
ing the risk of false-positives. Clinical decisions, such as
admission, transfer to a maternal intensive care (MIC) unit,
or administration of corticosteroids, should not be based
on this result in a post-coital context. Another biochemi-
cal test based on the qualitative dosage of phosphorylated
insulin-like growth factor binding protein-1 (phIGFBP-1) on
cervical secretions, which is not influenced by the presence
of semen [8] and/or measurement of cervical length with
vaginal ultrasound, are preferable. To wait 24 or 48 h and
repeat the test again makes no sense because clinical deci-
sions should have been made before then. In contrast, if the
ffn test has been performed but remains negative, the inter-
pretation retains all its value, even a short time after coitus
and the probability to deliver within a week is  < 1% [6].
We have not been able to demonstrate a linear or
asymptotic relationship between the time after coitus and
ffn values (vaginal or cervical; see Figures 1 and 2). More­
over, there is a great discrepancy between vaginal and
cervical results. The presence of concomitant macroscopic
blood in the sampling has been discussed elsewhere [2]
and could explain some outliers, but our rigorous tech-
nique of sampling allowed us to discard such confounding
factors. We had also specifically asked women to avoid the
use of lubricants during coitus and vaginal douching in
the interval time between last coitus and the sampling, but
showering, toilet use, or bathing are probably sufficient
methods to dilute vaginal semen. It would be interesting
to perform the test after coitus and vaginal douching to
check if the test is fully negative. The best explanation for
this unexpected result is that the median concentration
of ffn in seminal plasma is 21 μg/mL, but the concentra-
tion can also be as low as 2 μg/mL [15]. In this case, after
dilution in the vaginal secretions the ffn concentration is
probably too low to test positive. It also explains why the
proportion of positive tests after 24 h is not  > 25%. Usually,
Unauthenticated
Download Date | 6/27/16 6:45 AM
the amount of ffn is higher in cervical secretions than in
the vagina, but not in 32% of women when the ffn test is
done within 24  h after coitus. One explanation could be
that it occurs when the seminal ffn concentration reaches
a higher range (i.e., 100 μg/mL). In this case, even after
dilution by vaginal secretions the fornical ffn concentra-
tion remains higher than in the ectocervix.
Although the manufacturer of the ffn assay recom-
mends that sampling should be obtained from the fornix,
there are many published studied on the ffn test which
have been conducted with cervical sampling. This study
had as a secondary outcome the opportunity to assess the
difference between intracervical and fornical ffn quantita-
tive values. The observed difference in the ratio of vaginal
positive tests in the two groups was not observed with cer-
vical sampling; approximately 40% of cervical samples
are positive either a short or long time after intercourse.
The presence of ffn inside the cervix is so frequent and so
high that the impact of the semen appears to be minimal;
the mean value is not significantly enhanced. For clinical
purposes, these high values in low-risk patients clearly
indicate that intracervical sampling should be avoided or
the level of positivity should be enhanced.
Conclusions
Between 22 and 28 weeks’ gestation, coitus interferes with
the quantitative ffn test within 36 h, but the evolution
of ffn concentrations according to the time after coitus
cannot be mathematically predicted. Moreover, this test
should definitively be performed solely in the posterior
fornix and not in the (ecto) cervix due to the unacceptable
high false-positive rate of the test in cervical secretions.
References
[1] Bartnicki J, Casal D, Kreaden US, Saling E, Vetter K. Fetal fibronec-
tin in vaginal specimens predicts preterm delivery and very-low-
birth-weight infants. Am J Obstet Gynecol. 1996;174:971–4.
[2] Feinberg RF, Wang CL. Monoclonal antibody FDC-6 exhibits
binding to human plasma fibronectin: a caveat for cervico-
vaginal oncofetal fibronectin testing? Am J Obstet Gynecol.
1994;171:1302–8.
[3] Goffinet F. Primary predictors of preterm labour. Br J Obstet
Gynecol. 2005;112(Suppl 1):38–47.
[4] Goldenberg RL, Klebanoff M, Carey JC, Macpherson C, Leveno KJ,
Moawad AH, et al. Vaginal fetal fibronectin measurements from
8 to 22 weeks gestation and subsequent spontaneous preterm
birth. Am J Obstet Gynecol. 2000;183:469–75.
[5] Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-
Coello P, et al. (for the GRADE working group). Rating quality
408      Faron et al., Does recent sexual intercourse during pregnancy affect the results of the fetal fibronectin rapid test?

of evidence and strength of recommendations GRADE: an
emerging consensus on rating quality of evidence and
strength of recommendations. Br Med J. 2008;336:924–6.
[6] Honest H, Bachmann LM, Gupta JK, Kleijnen J, Khan KS.
­Accuracy of cervicovaginal fetal fibronectin test in predicting
risk of spontaneous preterm birth: systematic review. Br Med J.
2002;325:301–10.
[7] Iams JD, Romero R, Culhane JF, Goldenberg RL. Primary, sec-
ondary, and tertiary interventions to reduce the morbidity and
mortality of preterm birth. Lancet. 2008;371:164–75.
[8] Kallioniemi H, Rahkonen L, Heikinheimo O, Paavonen J. Early
pregnancy vaginal fluid phosphorylated insulin-like growth
factor binding protein-1 predicts preterm delivery. Prenat
Diagn. 2013;33:378–83.
[9] Lockwood CJ, Senyei AE, Dische MR, Casal D, Shah KD, Thung
SN, et al. Fetal fibronectin in cervical and vaginal secretions as a
predictor of preterm delivery. N Engl J Med. 1991;325:669–74.
[10] Mardh PA. The vaginal ecosystem. Am J Obstet Gynecol.
1991;165:1163–8.
[11] Morrison JC, Allbert JR, McLaughlin BN, Whitworth NS, Roberts
WE, Martin RW. Oncofetal fibronectin in patients with false
labor as a predictor of preterm delivery. Am J Obstet Gynecol.
1993;168:538–42.
[12] Pelaez LM, Fox NS, Chasen ST. Negative fetal fibronectin: who
is still treating for threatened preterm labor and does it help?
J Perinat Med. 2008;36:202–5.
[13] Rossel Goffeng A, Holst E, Milsom I, Lindstedt G, Lundberg
PA, Andersch B. Factors influencing vaginal fetal fibronectin
concentrations during pregnancy. Gynecol Obstet Invest.
1999;47:83–8.
[14] Sanchez-Ramos L, Delke I, Zamora J, Kaunitz AM. Fetal
fibronectin as a short-term predictor of preterm birth in
symptomatic patients: a meta-analysis. Obstet Gynecol.
2009;114:631–40.
[15] Shimoya K, Hashimoto K, Shimizu T, Saji F, Murata Y. Effect
of sexual intercourse on fetal fibronectin concentration in
­cervicovaginal secretions. Am J Obstet Gynecol. 1998;179:255–6.
[16] Slattery MM, Morrison JJ. Preterm delivery. Lancet.
2002;360:1489–97.
[17] Yost NP, Owen J, Berghella V, Thom E, Swain M, Dildy GA, 3rd
et al. Effect of coitus on recurrent preterm birth: a prospective
study. Obstet Gynecol. 2006;107:793–7.
The authors stated that there are no conflicts of interest regarding
the publication of this article.

Unauthenticated
Download Date | 6/27/16 6:45 AM