- 1 -
Publisher’s Version/PDF in open access
on editor web site:
Troude Pénélope, Bailly Estelle, Guibert Juliette, Bouyer Jean, La Rochebrochard Elise (de),
for the DAIFI Group, 2012, “Spontaneous pregnancies among couples previously treated by
in vitro fertilization”, Fertility and Sterility, 98(1), p. 63-68. DOI: 10.1186/1471-2288-12-
Spontaneous pregnancies among couples previously
treated by in vitro fertilization
Pénélope Troude a,b,c,d, MD, Estelle Bailly a, MSc, Juliette Guibert e, MD, Jean Bouyer a,b,f,
PhD, and Elise de La Rochebrochard a,b,f, PhD, for the DAIFI Group
a Institut National d’Etudes Démographiques (INED), F-75020 Paris, France
b Inserm, CESP Center for Research in Epidemiology and Population Health U1018, F-94276
Le Kremlin-Bicêtre, France
c AP-HP, Hôpital Lariboisière, F-75475 Paris, France
- 2 -
d Univ Paris Diderot, Sorbonne Paris Cité, Service de Santé Publique et Economie de la
Santé, F-75475 Paris, France
e Laboratoire de Procréation Médicalement Assistée, Institut Mutualiste de Montsouris, F-
75014 Paris, France
f Univ Paris-Sud, UMRS 1018, F-94276 Le Kremlin-Bicêtre, France
DAIFI Group members include: Institut National d’Etudes Démographiques (INED) - Institut
National de la Santé et de la Recherche Médicale (INSERM) – Université Paris Sud XI: Elise
de La Rochebrochard (national coordinator), Annie Bachelot, Estelle Bailly, Jean Bouyer,
Juliette Guibert, Henri Leridon, Patricia Thauvin, Laurent Toulemon, Pénélope Troude;
Auvergne: Rusudan Peikrishvili, Jean-Luc Pouly (CHU Estaing, Clermont-Ferrand 63);
Basse-Normandie: Isabelle Denis, Michel Herlicoviez (CHU Clémenceau, Caen 14);
Franche-Comté: Christiane Joanne, Christophe Roux (CHR Saint-Jacques, Besançon 25);
Haute-Normandie: Catherine Avril, Julie Roset (Clinique Mathilde, Rouen 76); Ile-de-
France: Joëlle Belaisch-Allart, Olivier Kulski (Center Hospitalier des 4 Villes, Sèvres 92);
Jean-Philippe Wolf, Dominique de Ziegler (Cochin, Paris 75); Philippe Granet, Juliette
Guibert (Institut Mutualiste Montsouris, Paris 75); Provence-Alpes-Côte d’Azur: Claude
Giorgetti, Géraldine Porcu (Institut de Médecine de la Reproduction, Marseille 13).
None of the authors have any conflict of interest.
Dr Penelope Troude
Inserm CESP U1018
94276 Le Kremlin-Bicêtre
- 3 -
Tel. +33 1 45 21 21 50
Fax +33 1 45 21 20 75
Capsule: Among couples previously treated by IVF, one in five has a live birth following a
spontaneous pregnancy. This should give hope to unsuccessfully treated couples, especially
young couples with unexplained infertility.
- 4 -
Objective: To determine the frequency of live births following spontaneous pregnancy (BSP)
and to examine their associated factors among couples who have unsuccessfully or
successfully experienced fertility treatments.
Design: Retrospective cohort.
Setting: Eight French IVF centers.
Patients: 2,134 couples who began IVF treatment in the centers in 2000-2002 and were
followed up by a postal questionnaire sent 7-9 years after they started treatment in the
Main outcome measures: Rates of live birth following spontaneous pregnancy, factors
associated with BSP. Univariate and multivariate analyses were conducted using logistic
Results: The BSP rate was 17% (218/1,320) among couples who had previously had a child
through medical treatment and 24% (193/814) among couples who had remained childless
after treatment. In both groups, the probability of BSP was higher among younger women
and increased with a smaller number of IVF attempts. Probability was also higher when the
cause of infertility was unexplained.
Conclusions: Our results should give hope to couples who have been unsuccessfully treated
by IVF, especially young couples with unexplained infertility. Nonetheless, it should be
remembered that the BSP rates are cumulative rates observed over a long period of time and
that these couples have a very low monthly probability of conceiving.
Keywords: Fertilization in vitro, spontaneous pregnancy, follow-up studies, France
- 5 -
The issue of live births following spontaneous pregnancy (BSP) among infertile couples who
have had in vitro fertilization (IVF) treatment may seem an unusual one in such a context (1)
and may appear anecdotal to physicians, especially if treatment has failed (2). However, in
view of some recent publications, BSP after IVF may not be such a rare occurrence.
The question of spontaneous pregnancies has been approached in studies on drop-out from
fertility care, since 5% to 7% of couples may have dropped out from the IVF waiting list
because of spontaneous pregnancy (3-5). Because these studies focused only on reasons for
treatment dropout, BSP that occurred at a later date were not taken into account. Some
studies have evaluated occurrence of BSP specifically among couples who previously had a
child through IVF and intracytoplasmic sperm injection (ICSI) (1, 2, 6). Reported BSP rates
ranged from 16% to 17% with a follow-up between two and five years. In another recent
study conducted in Denmark, rate of delivery after spontaneous pregnancy among couples
who had previously conceived through intrauterine insemination or assisted reproduction
techniques was 17% five years after treatment initiation (7). Other studies have focused on
couples who had been unsuccessfully treated by IVF or ICSI and the BSP rate varied
considerably (8-11). With various durations of follow-up (two to ten years), most of these
studies reported spontaneous pregnancy or BSP rates that ranged from 11% to 19%, except
one that observed a BSP rate as high as 32% (9).
Comparison between studies on BSP is difficult since the populations (infertile couples or
treated couples), outcome (pregnancy or live birth) and duration of follow-up differ
considerably. Moreover, most studies of unsuccessfully treated couples included less than
200 couples. Although the results varied widely, all published studies showed that BSP may
- 6 -
occur after IVF in as many as one or two couples out of 10. Our objective was to estimate the
spontaneous live birth rate and to examine its associated factors among couples who had
received IVF treatment for infertility, whether successfully or unsuccessfully.
- 7 -
MATERIALS AND METHODS
The DAIFI cohort
The DAIFI cohort (Outcome after In Vitro Fertilization treatment Initiation) is a retrospective
cohort including couples who had begun IVF treatment between 2000 and 2002 in one of the
eight participating French IVF centers (the inclusion centers). This study received
Institutional Review Board approval from the French Data Protection Authority
(authorization number 05-1334). Information on long-term outcome was collected by postal
questionnaire completed by the couples from 2008 through 2010. The questionnaire included
sociodemographic information, fertility and medical treatment history, occurrence of
spontaneous pregnancies and recourse to adoption. Of the 4,029 couples contacted, 2,321
completed the postal questionnaire. Analysis was restricted to 2,134 couples who were still
together at the time of study (187 separated couples were excluded).
Two study groups: successfully and unsuccessfully treated couples
The 2,134 couples were divided into two groups: the “successfully treated group” of 1,320
couples who had had a child during medical treatment and the “unsuccessfully treated group”
of 814 couples who remained childless after medical treatment. Success during medical
treatments included all births following IVF treatment in the inclusion center as well as all
births following other medical treatments (ovarian stimulation, surgery, artificial
insemination with partner sperm or donor sperm, oocyte donation, IVF, ICSI, FET)
subsequently received in France or abroad.
- 8 -
Births following spontaneous pregnancies
Births following spontaneous pregnancies (BSP) were declared by the couple in the postal
questionnaire. The date of the birth was recorded. Based on the medical records, information
on the dates of first and last IVF attempt in the inclusion center were collected. For
successfully treated couples, the date of birth following medical treatment was collected.
The occurrence and distribution of BSP were studied separately in each group. The two
groups were not compared as time of exposure to conception could vary between the two
groups and could not be properly controlled based on the retrospective data collected in this
study. In particular, periods of contraception may have differed between the two groups and
the corresponding dates were not asked because of the retrospective design of the study over
several years. Moreover, the exposure time was defined differently in each group: for
successfully treated couples, it was defined as the time since birth following treatment; for
unsuccessfully treated couples, it was defined as the time since the last attempt in the
inclusion center (i.e. last fresh or frozen embryo transfer).
The couples’ characteristics were analyzed in relation to BSP. Categorical variables were
compared using the chi-2 test, and quantitative variables were described as median, first and
third quartiles (Q1-Q3). Couples' characteristics included the woman’s and the man’s age at
the first IVF attempt (< 30, 30-34, 35-39, ≥ 40 years) and the highest educational level of the
couple (less than, equal to or higher than the French high school diploma). Medical
characteristics were extracted from the inclusion IVF center database and included type of
infertility (primary vs secondary), cause of infertility (female factor, male factor, mixed,
unexplained), its duration (defined as the interval between first attempting pregnancy and
first oocyte retrieval), characteristics at the first attempt in the inclusion center (number of
- 9 -
oocytes retrieved, number of oocytes fertilized, number of embryos transferred, number of
embryos frozen), overall characteristics of treatment in the center (number of attempts,
number of frozen embryo transfers (FET), at least one embryo transferred, at least one
pregnancy obtained). Information on the path followed by the couples included treatments
before and after IVF in the inclusion center and initiation of adoption procedures.
Multivariate analyses were conducted to assess factors associated with BSP using logistic
regression. Multivariate models were adjusted for inclusion center and included the following
variables: the woman’s age, duration of infertility, cause of infertility, having a child before
IVF in the inclusion center and other variables associated with BSP with a P value <0.20.
Some variables were excluded because of colinearity, such as the man’s age. All statistical
analyses were performed using STATA/SE 10.0 (Stata Press, College Station, TX, USA).
- 10 -
In the successfully treated group (n=1,320), the women’s median age at the first IVF attempt
in the inclusion center was 32 years (Q1-Q3[29-35]). Infertility was due to a female factor in
32%, a male factor in 37%, to both in 18% and was unexplained in 13%. Ninety-one percent
of live births were obtained through IVF in the inclusion center, 6.5% through IVF in another
center and 3% through other treatments (mainly artificial insemination). Among these
successfully treated couples followed up for a median of 6 years, 218 later had a spontaneous
live birth, a BSP rate of 17% (95% CI: 15%, 19%). BSP distribution over time is presented in
figure 1. Most BSP occurred during the third or fourth year after the birth following medical
treatment. The median time between the first birth following medical treatment and the birth
following spontaneous conception was 33 months [22-42].
In the unsuccessfully treated group (n=814), the women’s median age at the first attempt in
the inclusion center was 34 years [31-38]. Infertility was due to a female factor in 43%, a
male factor in 28%, to both in 17% and was unexplained in 12%. Eighteen percent of couples
had undergone other unsuccessful treatments after IVF in the inclusion center. Among the
814 unsuccessfully treated couples followed up for a median of 7 years since the last IVF
attempt in the inclusion center, 193 BSP were reported, or 24% (95% CI: 21%, 27%). The
median time elapsed since the initial pregnancy attempt was 6 years [5-7]. BSP distribution
over time since the last attempt in the inclusion center is presented in figure 2. Almost 60%
of BSP occurred during the first two years after the last IVF attempt in the inclusion center.
Characteristics of couples were compared in relation to BSP in the successfully treated group
and in the unsuccessfully treated group in Table 1. Five variables were significantly
- 11 -
associated with BSP in both groups in the univariate analysis: BSP was more frequent when
the woman and the man were younger, when the infertility was unexplained, when the
technique used at the first attempt was IVF without ICSI and when the number of IVF
attempts undergone by the couple in the inclusion center was small. In the successfully
treated group, BSP was also more frequent when the couple had already had treatment before
attending the inclusion center and when the couple did not have FET in the inclusion center.
In the unsuccessfully treated group, BSP was also more frequent when the duration of
infertility was shorter and when two embryos were transferred at the first IVF attempt in the
inclusion center. Multivariate analysis is presented in Table 2. In both groups, the probability
of BSP decreased as the woman’s age increased and the number of IVF attempts in the
inclusion center was greater (P<0.001 for trend). Probability increased when infertility was
due to a female factor and above all when it was unexplained. Two variables were significant
in the multivariate analysis only in the unsuccessfully treated group: for these couples, the
probability of BSP also increased when the duration of infertility decreased and when at least
one embryo was transferred during treatment in the inclusion center.
- 12 -
In a large sample of French couples undergoing IVF treatment for infertility, the rate of live
birth following spontaneous pregnancy was estimated to be 17% (15 to 19%) among
successfully treated couples and 24% (21 to 27%) among unsuccessfully treated couples.
Among unsuccessfully treated couples, 60% of BSP occurred in the two years following the
last IVF transfer, suggesting that BSP was the reason for IVF treatment discontinuation
(usually the duration between two attempts was no more than 18 months). Among
successfully treated couples, 58% of BSP occurred during the third and fourth years after the
first birth following medical treatment. Thus, the context of BSP is clearly quite different in
these two groups. Moreover, the BSP rate observed among successfully and unsuccessfully
treated couples cannot be directly compared because couples who had had a child through
fertility treatment had been exposed to the possibility of BSP during a shorter period of time
(because of the pregnancy period). Furthermore, couples who obtained the desired child
through fertility treatment may be less likely to try to conceive again and the study did not
include enough information to properly control for such differences in time of exposure to
conception between the two groups (for example contraceptive periods were unknown).
The BSP rate observed among successfully treated couples, 17%, was very close to the 16-
17% observed in other studies in Germany, Denmark and Japan (2, 6, 7). In our study, the
median time to BSP after the birth of a child following medical treatment was three years,
which is in accordance with previous reports (1, 2). The BSP rate observed among
unsuccessfully treated couples, 24%, appears to be slightly higher than most of those
previously reported in the literature, usually lower than 17% (8, 12, 13). However, one
English study has already reported a much higher BSP rate of 32% four to ten years after
- 13 -
referral to a center, in couples who had ceased to try to conceive by assisted reproduction
techniques (9). Previous studies concerned samples of less than 250 couples, whereas our
study has been conducted on more than 800 unsuccessfully treated couples.
In prospective and retrospective cohort studies, information cannot be gathered for all
couples: in our cohort, 58% of contacted couples participated in the postal survey seven to
nine years after IVF initiation in the inclusion center. This participation rate is higher than
that (44%) observed in an English cohort of couples who had undergone IVF four to ten
years earlier (9). Most studies did not report their participation rates, which may have been
even lower. We cannot rule out the possibility that couples who had a BSP are more likely to
respond than couples who did not, and so published BSP rates may be slightly overestimated.
Regarding factors associated with BSP, for both successfully and unsuccessfully treated
couples the probability of BSP decreased as the woman’s age increased, as previously
reported in the literature (1, 6, 9, 14, 15). The cause of infertility was globally associated with
BSP among successfully treated couples and among unsuccessfully treated couples. The
proportion of BSP was higher for couples with unexplained infertility and lower for male-
factor infertility. Previously reported results on cause or type of infertility are mixed: cause or
diagnosis of infertility have been reported as associated factors in some studies (1, 9, 14) and
not in others (6, 15, 16). For unsuccessfully treated couples, we observed 32% of BSP among
couples with unexplained infertility while Cahill et al. reported 35% of spontaneous
pregnancy in this subpopulation within three years after IVF treatment (9). Other medical
characteristics also seemed linked to BSP, especially among unsuccessfully treated couples.
The proportion of BSP decreased when the number of IVF attempts increased in the
successfully and in the unsuccessfully treated groups, but the impact of this variable seemed
- 14 -
much greater in the unsuccessfully treated group. Other variables were significant in the
multivariate model only among the unsuccessfully treated couples: duration of infertility and
at least one embryo transfer during the treatment in the inclusion center. Globally, factors
associated with BSP can be viewed as indicators of a better fertility prognosis, especially
among unsuccessfully treated couples: couples with BSP are younger and they have better
medical characteristics (shorter duration of infertility and at least one embryo transfer during
IVF treatment). Unexplained infertility appeared strongly associated with BSP (17), and
could probably also be interpreted as an indicator of better fertility prognosis compared with
couples whose infertility is of known origin, and particularly of male-factor origin. In the
unsuccessfully treated group, the BSP rate in couples with unexplained infertility and a
woman's age under 35 years was as high as 45% (n = 21/47; 95% CI: 30%, 60%).
Considering the 37 couples with the best prognostic factors (unexplained infertility, woman’s
age under 35 years and infertility duration ≤ 5 years), this rate even reached 57% (95% CI:
39%, 73%). This result is in agreement with a study conducted in the Netherlands on the
reasons for dropping out from a waiting list for IVF: spontaneous pregnancies occurred
mostly in couples with unexplained infertility and younger than 35 years (5). Based on these
results, it would be very interesting to develop further research with the specific aim of better
characterizing young couples with unexplained infertility. However, these couples only
represent a small proportion of all unsuccessfully treated couples, at least in our French
To conclude, our results should give hope to couples who have previously been treated by
IVF, especially for unsuccessful young couples with unexplained infertility. Nonetheless, it
should be borne in mind that the published rates are cumulative BSP observed over a long
period of time. For example, if unsuccessfully treated couples had a cumulative BSP rate of
- 15 -
24%, they had to wait a median of six years to have their child. This corresponds to a very
low monthly conception probability (fecundability) of around 0.44%, whereas fecundability
in the general population is estimated at 25% (18).
- 16 -
1. Hennelly B, Harrison RF, Kelly J, Jacob S, Barrett T. Spontaneous conception after a
successful attempt at in vitro fertilization/intracytoplasmic sperm injection. Fertil Steril
2. Ludwig AK, Katalinic A, Jendrysik J, Thyen U, Sutcliffe AG, Diedrich K et al.
Spontaneous pregnancy after successful ICSI treatment: evaluation of risk factors in
899 families in Germany. Reprod Biomed Online 2008;17:403-9.
3. Brandes M, van der Steen JO, Bokdam SB, Hamilton CJ, de Bruin JP, Nelen WL et al.
When and why do subfertile couples discontinue their fertility care? A longitudinal
cohort study in a secondary care subfertility population. Hum Reprod 2009;24:3127-
4. Horne G, Farrell C, Pease EH, Brison DR, Falconer DA, Lieberman BA. Waiting for in
vitro fertilization treatment: spontaneous and ART live births. Hum Fertil (Camb)
5. van Dongen AJ, Verhagen TE, Dumoulin JC, Land JA, Evers JL. Reasons for dropping
out from a waiting list for in vitro fertilization. Fertil Steril 2009;94:1713-6.
6. Shimizu Y, Kodama H, Fukuda J, Murata M, Kumagai J, Tanaka T. Spontaneous
conception after the birth of infants conceived through in vitro fertilization treatment.
Fertil Steril 1999;71:35-9.
7. Pinborg A, Hougaard CO, Nyboe Andersen A, Molbo D, Schmidt L. Prospective
longitudinal cohort study on cumulative 5-year delivery and adoption rates among
1338 couples initiating infertility treatment. Hum Reprod 2009;24:991-9.
8. Osmanagaoglu K, Collins J, Kolibianakis E, Tournaye H, Camus M, Van Steirteghem
A et al. Spontaneous pregnancies in couples who discontinued intracytoplasmic sperm
injection treatment: a 5-year follow-up study. Fertil Steril 2002;78:550-6.
- 17 -
9. Cahill DJ, Meadowcroft J, Akande VA, Corrigan E. Likelihood of natural conception
following treatment by IVF. J Assist Reprod Genet 2005;22:401-5.
10. Filetto JN, Makuch MY. Long-term follow-up of women and men after unsuccessful
IVF. Reprod Biomed Online 2005;11:458-63.
11. de La Rochebrochard E, Quelen C, Peikrishvili R, Guibert J, Bouyer J. Long-term
outcome of parenthood project during in vitro fertilization and after discontinuation of
unsuccessful in vitro fertilization. Fertil Steril 2009;92:149-56.
12. Osmanagaoglu K, Tournaye H, Camus M, Vandervorst M, Van Steirteghem A,
Devroey P. Cumulative delivery rates after intracytoplasmic sperm injection: 5 year
follow-up of 498 patients. Hum Reprod 1999;14:2651-5.
13. Brandes M, Hamilton CJ, de Bruin JP, Nelen WL, Kremer JA. The relative
contribution of IVF to the total ongoing pregnancy rate in a subfertile cohort. Hum
14. Eijkemans MJ, Lintsen AM, Hunault CC, Bouwmans CA, Hakkaart L, Braat DD et al.
Pregnancy chances on an IVF/ICSI waiting list: a national prospective cohort study.
Hum Reprod 2008;23:1627-32.
15. Kupka MS, Dorn C, Richter O, Schmutzler A, van der Ven H, Kulczycki A. Stress
relief after infertility treatment--spontaneous conception, adoption and psychological
counselling. Eur J Obstet Gynecol Reprod Biol 2003;110:190-5.
16. Vardon D, Burban C, Collomb J, Stolla V, Erny R. [Spontaneous pregnancies in
couples after failed or successful in vitro fertilization]. J Gynecol Obstet Biol Reprod
17. Brandes M, Hamilton CJ, van der Steen JO, de Bruin JP, Bots RS, Nelen WL et al.
Unexplained infertility: overall ongoing pregnancy rate and mode of conception. Hum
- 18 -
18. Spira A. Epidemiology of human reproduction. Hum Reprod 1986;1:111-5.
- 19 -
Figure 1. Distribution of BSP over time since birth following treatments among successfully
treated couples (n = 218)
- 20 -
Figure 2. Distribution of BSP over time since last IVF attempt in the inclusion center among
unsuccessfully treated couples (n = 193)
- 21 -
Table 1. Live births following spontaneous conception (BSP) and characteristics of
successfully treated couples (n = 1,320) and unsuccessfully treated couples (n = 814).
couples (n = 1,320)
couples (n = 814)
Woman’s age at 1st attempt (years)
Man’s age at 1st attempt (years)
Highest educational level of the
< French high school diploma
French high school diploma
> French high school diploma
Duration of infertility (years)
Type of infertility
Origin of infertility
Previous child before IVF in
IVF technique at 1st IVF attempt
Oocytes retrieved at 1st IVF attempt
- 22 -
Embryos transferred at 1st IVF
Number of IVF attempts in
Number of frozen embryo transfers
in inclusion center
At least one embryo transferred in
* Couples may have been successfully treated in the center or subsequently treated in another
center: 15 couples who had no embryo transferred in the inclusion center were later
successfully treated elsewhere.
- 23 -
Table 2. Factors associated with live births following spontaneous pregnancy in successfully and unsuccessfully treated infertile couples.
Successfully treated couples
(n = 1,320)
Unsuccessfully treated couples
(n = 814)
Multivariate analysis a
(n = 1,276)
Multivariate analysis a
(n = 787)
Woman’s age at 1st attempt (years)
Duration of infertility (years)
Origin of infertility
Number of attempts in inclusion center
At least one embryo transferred in inclusion center
Previous child before IVF
a Adjusted for inclusion center; b p for trend