Human Reproduction, Vol.24, No.9 pp. 2173– 2182, 2009
Advanced Access publication on May 22, 2009 doi:10.1093/humrep/dep185
ORIGINAL ARTICLE Infertility
Stressful life events are associated with
a poor in-vitro fertilization (IVF)
outcome: a prospective study
S.M.S. Ebbesen 1,3, R. Zachariae 1, M.Y. Mehlsen 1, D. Thomsen 1,
A. Højgaard 2, L. Ottosen 2, T. Petersen 2, and H.J. Ingerslev 2
1
Department of Psychology, University of Aarhus, Jens Chr. Skous vej 4, 8000 Aarhus, Denmark 2The Fertility Clinic, Aarhus University
Hospital, 8200 Aarhus, Denmark
3
Correspondence address. E-mail: [email protected]
background: There is preliminary evidence to suggest an impact of stress on chances of achieving a pregnancy with in-vitro fertilization
(IVF). The majority of the available research has focused on stress related to infertility and going through IVF-treatment, and it is still unclear
whether non-fertility-related, naturally occurring stressors may influence IVF pregnancy chances. Our aim was to explore the association
between IVF-outcome and negative, i.e. stressful, life-events during the previous 12 months.
methods: Prior to IVF, 809 women (mean age: 31.2 years) completed the List of Recent Events (LRE) and questionnaires measuring
perceived stress and depressive symptoms.
results: Women who became pregnant reported fewer non-fertility-related negative life-events prior to IVF (Mean: 2.5; SD: 2.5) than
women who did not obtain a pregnancy (Mean: 3.0; SD: 3.0) (t(465.28) ¼ 2.390, P ¼ 0.017). Logistic regression analyses revealed that the
number of negative life-events remained a significant predictor of pregnancy (OR: 0.889; P ¼ 0.02), when controlling for age, total number of
life-events, perceived stress within the previous month, depressive symptoms, and relevant medical factors related to the patient or treatment procedure, including duration of infertility, number of oocytes retrieved and infertility etiology. Mediation analyses indicated that the
association between negative life events and IVF pregnancy was partly mediated by the number of oocytes harvested during oocyte retrieval.
conclusion: A large number of life-events perceived as having a negative impact on quality of life may indicate chronic stress, and the
results of our study indicate that stress may reduce the chances of a successful outcome following IVF, possibly through psychobiological
mechanisms affecting medical end-points such as oocyte retrieval outcome.
Key words: infertility / stress / negative life events / IVF / treatment outcome
Introduction
Between 10 and 15% of all couples experience fertility problems due
to a variety of causes (Schmidt et al., 1995; Eugster and Vingerhoets,
1999; Juul et al., 1999), and infertility is increasing in the industrialized
countries, possibly due to social and behavioural factors along with
environmental exposures (Skakkebaek et al., 2006). Infertility is commonly defined as failure to conceive after at least 1 year of attempting
to achieve a pregnancy with unprotected intercourse (Schmidt, 2006;
Dansk Fertilitetsselskab, 2007). Couples suffering fertility problems
frequently turn to artificial reproductive technology (ART), and it is
assumed that 50% of all Danish couples experiencing infertility
seek ART treatment (Schmidt et al., 1995). In Denmark, the
number of initiated treatments with in-vitro fertilization (IVF) and
intra cytoplasmatic sperm injection (ICSI) performed at public and
private fertility clinics has increased by 83%, from approximately
6000 per year to more than 11 000 per year, within the last
10 years (National Board of Health D, 2006; The Danish Fertility
Society, 2009).
There is evidence to suggest that psychological factors may influence the chances of obtaining a pregnancy with IVF (Eugster and
Vingerhoets, 1999; Klonoff-Cohen, 2005; Boivin and Schmidt, 2005).
A number of studies have shown associations between distress,
mostly anxiety and depression, and IVF-outcome, although a few
studies have been unable to confirm this link (Milad et al., 1998;
Lovely et al., 2003; de Klerk et al., 2008). Such findings are often
reported as studies on effects of ‘stress’ on IVF-outcome (Harlow
et al., 1996; Csemiczky et al., 2000; Lovely et al., 2003; Smeenk
et al., 2005). However, the instruments used in these studies are
generally measures of various types of distress which may be
& The Author 2009. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: [email protected]
2174
unrelated to external factors. The majority of the available studies
therefore do not qualify as studies of stress.
The term ‘stress’ is, in fact, often confused with distress, which,
although considered part of the same process, is a different phenomenon. The stress process is generally considered to consist of three
aspects: (1) a stressor, i.e. an environmental demand, e.g. a life
event or a series of life events, followed by (2) a set of appraisals
and the perception of stress, which then may lead to (3) affective, behavioural and/or biological stress responses (i.e. distress) (Cohen
et al., 1995). Perception of stress results when the individual finds
the environmental demands taxing and/or threatening, and at the
same time feels insufficiently able to cope with the demands due to
lack of personal or environmental resources (Cohen et al., 1995).
According to the original conceptualization of a stressor, it is
defined as ‘that which produces stress’ (Selye, 1976).
However, there is a challenge in assessing stressors in terms of life
events, as the interpretation of and capacity to handle an experienced
event both emotionally and practically is highly individual (Cohen et al.,
1995).
For our purpose of examining the possible association between
stress and reproductive outcomes, stress is defined as events or
environmental demands (i.e. stressors), rated by the individual as
having had a negative influence on her quality of life. These events
will from here on be referred to as ‘negative life events’. The psychobiological effects of stress may potentially be long-lasting and persist
even after the stressor has ceased to act (Selye, 1976).
Only very few studies of stress in the strict sense of the term and
IVF-outcome have been conducted, focusing primarily on perceived
marital stress and perceived stress caused by infertility and IVFtreatment. Infertility can indeed be very stressful for the affected
couples (Berg et al., 1991; Eugster and Vingerhoets, 1999; Oddens
et al., 1999; Webb and Daniluk, 1999; Peterson et al., 2003, 2006;
Throsby and Gill, 2004; Schmidt et al., 2005a, b), and two large
studies conducted with IVF-populations have found some support for
infertility-related stress as a predictor of IVF-outcome (Klonoff-Cohen
et al., 2001; Boivin and Schmidt, 2005). Another study, however,
found the opposite result, i.e. higher infertility-related stress in the
group that subsequently achieved a pregnancy (Cooper et al., 2007).
One study measured IVF-treatment related stress (Klonoff-Cohen and
Natarajan, 2004) and found this source of stress to be a predictor of
the number of retrieved and fertilized oocytes. Women who prior to
treatment had severe concerns about the medical aspects of the procedure, about missing work, or about finances, had up to 20% fewer
oocytes retrieved and fertilized, which affected their pregnancy
chances negatively. The third type of stressor studied in relation to IVFtreatment outcome, perceived marital stress, has also been associated
with treatment outcome, in that the number of cycles-to-pregnancy was
higher in women reporting high-marital stress compared with women
reporting low-marital stress (Boivin and Schmidt, 2005). Preliminary evidence for an association between perceived marital stress and IVF
outcome has also found partial support in two other studies (Stoleru
et al., 1997; Verhaak et al., 2001). Finally, two studies (Facchinetti
et al., 1997; Gallinelli et al., 2001) have provided evidence for an
effect of experimentally induced stress on IVF-outcome using the
Stroop Color Word Test (Stroop, 1935).
Taken together the available evidence is limited, yet suggestive of a
stress effect on IVF-outcome. The majority of the available research
Ebbesen et al.
has focused on perceived stress in relation to infertility and IVFtreatment, which may be confounded by the couples’ knowledge
about their prognosis, and it is still unclear whether naturally occurring
stressors, which are unrelated to infertility, may influence pregnancy
chances following IVF. The aim of our study was therefore to investigate the possible association between negative life events and pregnancy chances in a sample of women in IVF-treatment at a large
regional university hospital fertility clinic. We hypothesized that an
increase in the number of negative life events would be associated
with reduced chances of obtaining IVF-induced pregnancy. Stress
may impact fertility possibly by interrupting menstrual cycles
(Chrousos and Gold, 1992; Ferin, 1999) or through a poorer
ovarian functional response to IVF-treatment (Klonoff-Cohen and
Natarajan, 2004). We therefore planned to explore the possible
role of the number of retrieved oocytes as a possible mediator of
this association, if an association between negative life events and
pregnancy chances was found.
Materials and Methods
Patients
A total of 837 women participated in this study from a cohort of approximately 1578 couples undergoing their first IVF-treatment cycle at the fertility clinic. The selection criteria were: (1) first IVF-cycle, (2) no previous
attempts with IVF-treatment and (3) ability to read and understand Danish
well enough to complete the questionnaires. Couples were excluded from
the study in cases of PGD (Preimplantation Genetic Diagnosis) as the indication for these treatments in the clinic is usually genetic and not infertility.
Another exclusion criterion was unplanned change of treatment type, e.g.
from insemination to IVF due to too many follicles.
Procedure
As a part of a larger study, couples referred to IVF-treatment at the Fertility Clinic, University Hospital, Skejby between October 2001 and
September 2006 received oral and written information about the study
and an invitation to participate during their first treatment cycle.
Women, who agreed to participate, received a baseline questionnaire
package when presenting themselves at their first scheduled appointment
with a fertility clinic nurse, taking place approximately on the 21th day of
the cycle. At this meeting the patients were also instructed on how to
administer hormonal down-regulation with a GnRH-agonist for 2 weeks
followed by 8 days of stimulation with Follicle Stimulating Hormone
(FSH). The baseline questionnaire was completed at home during this
period and handed in by the patient approximately 2 weeks later when
presenting herself for a follicle scan.
The physician registered medical and technical details in relation to the
oocyte retrieval procedure and results of pregnancy tests were added
when available. The study was accepted by the Regional Ethics Committee
and approved by the Danish Data Protection Agency.
Measurements
Independent variables: stress measures
Negative life events. An adjusted version of The List of Recent Events
(LRE) (Henderson et al., 1981) was used to measure the presence and
psychological impact of major life events. The original version of the
scale consists of an event list of 59 items describing various potentially
stressful environmental events within the following categories: illness,
injury and accident, bereavement, pregnancy/children, changes in
2175
Stressful life events and IVF outcome
relationships, living conditions, study/schooling, work/employment,
finances and legal difficulties. For each event, the participants are to
answer yes or no to the question ‘Did you within the past 12 months
experience [this event]?’. For each event, the participants were asked to
indicate whether the event had an impact on subjective well-being by
answering the question: ‘Did this event have a negative impact on your
life quality’ with yes or no. The LRE has been shown to adequately
assess the presence of negative life events in the general population
(Henderson et al., 1981), and has shown predictive strength in relation
to physical health symptoms (Cohen et al., 1991). In the present study a
shortened 37-item version of the list was used. We excluded events
such as death of spouse, divorce, new partner, giving birth, since these
events were irrelevant to the sample due to the hospital’s admission
criteria for IVF-treatment. Two total scores are derived from the LRE:
the total number of experienced events, which in the following is
termed ‘number of events’, and total number of events ‘perceived as
having had a negative impact on quality of life’, from here on termed
‘negative life events’. Thus from the total number of events, ‘negative
life events’ refer to the (sum of) events evaluated by the person as
having a negative influence on her quality of life.
Perceived stress. Perceived stress within the past month was assessed with
the Perceived Stress Scale (PSS-10) (Cohen and Williamson, 1988), consisting of 10 items measuring the present level of self rated stress. The
PSS-10 is a widely used measure that has been shown to be reliable and
valid. The internal consistency of the PSS-10 was considered acceptable
within this sample (Cronbachs a ¼ 0.85).
Independent variables: distress measures
Depressive symptoms. The Beck Depression Inventory (BDI-II) (Beck et al.,
1996) was used to measure depressive symptoms. The BDI consists of 21
items measuring the severity of both somatic and cognitive-affective
depressive symptoms. It is a widely used instrument for assessing depressive symptoms in various populations and has shown good reliability and
validity. The internal consistency of the BDI in this sample was acceptable
(a ¼ 0.71).
Dependent variables
Pregnancy. A positive pregnancy test was defined by measurement of a
serum HCG-level of .20 IU 2 weeks following embryo transfer. Pregnant
patients were scanned routinely 5 weeks after embryo transfer. Clinical
pregnancy was defined as a live intrauterine pregnancy detected by ultrasound. Where nothing else is specified, pregnancy and IVF-outcome refers
to clinical pregnancy.
In cases of multiple comparisons, Bonferroni adjustments were made in
order to reduce the risk of type-1 error by dividing the alpha level with
the number of comparisons, resulting in accordingly adjusted p-levels
(Tabachnick and Fidell, 2007). It has been argued that in large samples,
normal distribution is not a prerequisite for the t-test (Lumley et al.,
2002), thus in the analyses t-tests are chosen over non-parametric test
in order to retain statistical power.
Univariate analyses were conducted, comparing patients who agreed to
participate to the eligible non-participating patients in the cohort. Independent samples t-test was applied in order to explore differences between
the two groups concerning female age, duration of infertility, total
hormone dose before aspiration, number of previous births, number of
previous provoked and spontaneous miscarriages, number of previous
ectopic pregnancies. x2-analysis was applied to assess differences in pregnancy outcomes, smoking habits and coffee intake.
Means and standard deviations were calculated for the independent
variables (IVs) BDI, PSS-10 and LRE for women who subsequently did
or did not achieve pregnancy following IVF-procedure. A series of independent samples t-test was carried out to assess differences between
these two groups.
If differences in the number of negative life events between IVFsuccessful and non-successful women were found, the data were initially
inspected for possible patient- or treatment-related confounders and
mediators of this association: correlation analyses/linear regression techniques were applied to investigate associations with dependent variables
(DVs): number of oocytes and negative life events, and logistic regression
was used for detection of possible associations with pregnancy chances as
a DV. Statistical information is provided for any significant results.
Next, a series of direct regression analyses were planned in order to
examine the main hypothesis of an association between negative life
events and IVF pregnancy chances, although controlling for relevant biological and medical factors relating to infertility and IVF-treatment procedure. Subsequently, to test the possible mediating role of number of
oocytes for the association between negative life events and IVF pregnancy, we used the method described by Baron and Kenny (Baron and
Kenny, 1986). Mediation is proposed to be established when (1) the IV
significantly predicts the DV, (2) the IV significantly predicts the mediator,
(3) the mediator predicts the DV when controlling for the IV and (4)
when controlling for the mediator, the effect of the IV on the DV
ceases (complete mediation) or is reduced (partial mediation) (Baron
and Kenny, 1986). Finally, x2 analyses were conducted for each
possible life event on the LRE in order to explore whether some
types of events posed a higher risk for an adverse IVF-outcome than
others.
Additional measures
Missing values
The number of oocytes retrieved at oocyte retrieval, the cause of infertility
(male, tubal, endometriosis, ovulation factor, idiopathic) and the type of
treatment (ICSI or IVF) were registered by the physician. In case of
more than one cause of infertility, the most clinically significant cause
was recorded as the primary cause. In the following, cause of infertility
refers to the primary cause of infertility. Medical and treatment-related
information and information on life style factors was extracted from the
hospital register database and used for comparison between sample and
non-participants in the cohort. BMI was calculated from self-reported
weight and height data. Basal FSH was measured on cycle day 2 – 4.
For all psychometric scales (excl. LRE), the proportion of missing values
was computed, and missing values were substituted with the mean
values for the remaining scale items. In cases where more than 50% of
items in the respective scales were unanswered, the case was omitted
from the analysis. Substituting missing values with the mean for the remaining scale items is recommended, provided that the internal reliability of the
scales is high (i.e. a . 0.70) (Schafer and Graham, 2002).
Some variables from the hospital register contained a proportion of
unubstituted missing data, and are primarily used for comparative analyses
between participants and non-participants. N is reported for these analyses. In addition, exploratory analyses were conducted on the basis of
hospital register data with FSH, BMI, coffee, smoking and duration of
infertility. All other analyses are based on data collected as a part of this
study.
Statistical analyses
The statistical analyses were conducted with the Statistics Package for the
Social Sciences (SPSS) version 15.0.
2176
Ebbesen et al.
Results
Study sample
Of the 837 couples, 18 cases had not completed questionnaires at
baseline and were excluded from the data analyses. Another 10
cases were excluded for various reasons (puncture of cysts, spontaneous pregnancy, ovulation prior to oocyte retrieval, treatment cancellation or moving away). This resulted in a total of 809 couples in the
final sample. Sociodemographic and reproductive characteristics of the
sample can be seen in Table I.
Minor inconsistencies between numbers given in Table I and in the
following comparative analyses between sample and non-participants
are due to data being drawn from two different sources, i.e. study questionnaires and hospital register. Study questionnaire data are used to
describe sample characteristics, as these data are more detailed.
Table I Socio-demographic and reproductive
characteristics of study participants
Characteristic
n
Mean (SD)
or %
Range
........................................................................................
Age, mean (SD)
796
31.2 (3.9)
[21, 40]
Marriage/Living together
804
4 –6 years
(2 years)
[2, 8þ]
Education (ISCED-based, (UNESCO, 2009))
Lower secondary general (7 years)
4
0.5%
Lower secondary general (8–
10 years)
46
5.7%
Upper secondary (11– 13 years)
245
30.3%
Tertiary , master degree (14–
17 years)
366
45.2%
Tertiary master degree
(18 years)
137
16.9%
Missing data
11
1.4%
Duration of infertility in months
668
31.5 (22.9)
284
35.1%
Comparing pregnant and non-pregnant
women
[0– 228]
Type of procedure
ICSI long down-regulation
IVF long down-regulation
385
47.6%
IVF/ICSI mild hormone stimulation
51
6.3%
Missing data
89
11%
Cause of infertility
Ovulation factor
26
3.2%
Tubal
135
16.7%
Endometriosis
79
9.8%
Male
351
43.4%
Idiopathic
145
17.9%
Missing data
73
9%
Positive pregnancy test
280
34.6%
Clinical pregnancy
217
26.8%
Missing data
14
1.7%
Number of oocytes retrieved
776
6.93 (4.4)
Missing data
33
4%
No differences were found between patients who agreed to participate and the 522 eligible non-participating patients in the cohort (total
n ¼ 1359) with respect to age (n ¼ 1359), duration of infertility (n ¼
1259) and daily amount of smoking (n ¼ 1001), number of previous
abortions (spontaneous or provoked), number of previous ectopic
pregnancies and number of previous births (n ¼ 729). Also, there
was no statistically significant difference between the study group
and the rest of the cohort as to pregnancy outcome from the first
treatment cycle (n ¼ 336) (data not shown).
The group of study participants represented a larger proportion of
male factor infertility cases (46% male factor, 33% female factor and
21% idiopathic) compared with the non-participants (37% male factor,
31% female factor and 32% idiopathic) who, in return represented
slightly more cases of idiopathic infertility (x2 (2) ¼ 20.193, P ¼
0.0005) (n ¼ 1359). In accordance, study participants represented a
larger proportion of patients receiving ICSI treatment for male factor
infertility (40% ICSI, 60% IVF) than the rest of the cohort (23% ICSI,
77% IVF) (x2(1) ¼ 39.220, P ¼ 0.0005) (n ¼ 1359). Study participants
received a higher total hormone dose before aspiration (MEAN:
1650.60 IU, SD: 1592.36 IU) compared with the rest of the cohort
sample (MEAN: 1252.03 IU, SD: 1795.29 IU) (t(1358) ¼ 24.272,
P ¼ 0.0005) (n ¼ 1358). Total hormone dose before aspiration
showed a moderately skewed distribution. Among study participants
there was a larger coffee intake compared with the non-participants in
the cohort (x2(3) ¼ 9.082, P ¼ 0.028) (n ¼ 986). When correcting
for multiple comparisons, using a Bonferroni adjustment of the alpha
level (0.05/13 ¼ 0.004), differences between participants and nonparticipants’ coffee habits ceased to be significant.
Means and standard deviations for women who subsequently did or
did not achieve pregnancy following the IVF-procedure on the IVs
BDI, PSS-10 and LRE are shown in Table II. The BDI and negative
life events showed moderately skewed distributions. A series of independent samples t-tests showed no significant differences between
successful and unsuccessful IVF-patients for depressive symptoms,
perceived stress and total number of life events (with an adjusted
(0.05/3) alpha level of 0.02) (data not shown).
Initial testing of the main hypothesis showed that women who
obtained a clinical pregnancy following the IVF-procedure reported a
statistically significantly smaller proportion of negative life events
than women who did not obtain pregnancy. These results are
shown in Table II.
Correlations with reproductive outcomes
IVF-treatment outcome
[0– 25]
The number of oocytes retrieved during oocyte retrieval showed a
small, but statistically significant negative correlation with number of
negative life events (r ¼ 20.10, n ¼ 772, P , 0.01 (2-tailed)), and
was also a statistically significant univariate predictor of pregnancy
(OR ¼ 1.072, n ¼ 733, P ¼ 0.0005). FSH is an indicator of potential
ovarian response and the ovarian response is a predictor of
IVF-outcome (Yih et al., 2005). Although FSH was a statistically significant univariate predictor of number of oocytes (b ¼ 20.197, n ¼
495, P ¼ 0.0005), we found no association between FSH and pregnancy, or between negative life events and FSH (no data shown).
2177
Stressful life events and IVF outcome
Table II Differences between pregnant and non-pregnant women
Measures
Pregnant
Mean (SD)
Non-pregnant
Mean (SD)
T
P
.............................................................................................................................................................................................
Perceived stress
14.8 (6.0)
14.4 (6.0)
ns
Depressive symptoms
7.2 (6.1)
7.2 (6.5)
ns
Total number of life events
5.4 (3.0)
5.9 (3.4)
Negative life events
2.5 (2.5)
3.0 (3.0)
We were therefore unable to establish FSH as a possible mediator,
either of the association between negative life events and the
number of oocytes or the association between negative life events
and pregnancy. Similarly, although age was shown to be significantly
associated with pregnancy (OR ¼ 0.948, n ¼ 783, P ¼ 0.01), this variable was unrelated to negative life events (no data shown), and can
thus likewise be excluded as a possible mediator. There were no significant differences in the number of oocytes retrieved as a function of
treatment type (IVF or ICSI) (no data shown). BMI (n ¼ 651), daily
smoking habits (n ¼ 521), and daily coffee intake (n ¼ 518) were
unrelated to clinical pregnancy (data not shown).
Effects of negative life events on pregnancy
chances
To test whether negative life events predicted the likelihood of obtaining a positive IVF-treatment outcome, a series of logistic regression
analyses were carried out with pregnancy versus no pregnancy as
the DV and negative life events as the predictor, while controlling
for total number of life events experienced, current perceived
stress, depressive symptoms, and age.
Moderate to strong positive correlations were seen between the
number of negative life events (LRE) and depressive symptoms (BDI),
and between negative life events and perceived stress within the last
month (PSS-10). Whether life events are negatively evaluated may be
ns
T(465.28) ¼ 2.390
P ¼ 0.017
confounded by the person’s current level of perceived stress and
depressive symptoms, and since the correlations were not sufficiently
large to suggest problems related to multicollinearity (Tabachnick and
Fidell, 2007), these two factors were included in the regression model.
The models provided an acceptable fit to the data as indicated by
the omnibus test of model coefficients (Positive pregnancy test:
x2(5) ¼ 17.03, P ¼ 0.004, n ¼ 764, Clinical pregnancy: x2(5) ¼
16.35), P ¼ 0.006, n ¼ 764) explaining 2–3% of the variance in pregnancy rate (pseudo R 2-statistics). The IV, negative life events,
remained a statistically significant predictor of a positive pregnancy
test following IVF, when controlling for depressive symptoms,
perceived stress, age and total number of events experienced (OR:
0.909, P ¼ 0.018). Concerning clinical pregnancy, there was a trend
toward negative life events being a predictor IVF-outcome when controlling for the same variables (OR: 0.922, P ¼ 0.063).
When using pregnancy as a DV, it is important to control for
selected medical factors related to the IVF-patient and the treatment.
We therefore conducted a second logistic regression analysis,
additionally controlling for number of oocytes, duration of infertility
in months and cause of infertility, categorized into female, male or
idiopathic factor. Female factor was used as a simple contrast (reference group) in the analysis. The results are presented in Table III.
Negative life events continued to be a predictor of clinical pregnancy
chances when controlling for number of oocytes, cause of infertility
and duration of infertility in the model (OR: 0.889, P ¼ 0.02). The
Table III Effects of stress measures on IVF outcome, when controlling for medical variables
Measures
Coefficient (B)
S.E.
Wald
d.f.
Odds Ratio
95% CI
P
.............................................................................................................................................................................................
Negative life eventsa
20.118
0.050
5.423
1
0.889
[0.805, 0.982]
0.02
Number of eventsb
0.009
0.040
0.058
1
1.010
[0.934, 1.091]
0.81
Depressive symptoms
0.005
0.020
0.070
1
1.005
[0.967, 1.045]
0.79
Perceived stress
0.032
0.022
2.181
1
1.032
[0.990, 1.077]
0.14
20.074
0.026
8.349
1
0.928
[0.882, 0.976]
0.00
0.061
0.022
7.839
1
1.063
[1.018, 1.109]
0.01
1.586
2
Male cause of infertility
20.269
0.215
1.570
1
0.764
[0.501, 1.164]
0.21
Idiopathic infertility
20.179
0.269
0.444
1
0.836
[0.493, 1.417]
0.50
0.004
0.004
0.936
1
1.004
[0.996, 1.012]
0.33
Age
Number of oocytes
Female cause of infertility
Duration of infertility
All variables used in the analysis are shown.
N ¼ 588 (Missing cases: 221). Pseudo R 2-statistics: Cox and Snell, R 2 ¼ 0.048, Nagelkerke R 2 ¼ 0.069.
a
LRE, events rated as having had negative impact on quality of life.
b
LRE, number of events.
0.45
2178
Ebbesen et al.
results indicated that for every negative life event a patient had experienced within the past 12 months, there was approximately a 10%
decrease in the chances of obtaining a clinical pregnancy.
Age and the number of oocytes retrieved were both significantly
associated with chances of obtaining a clinical pregnancy, even when
controlling for other medical factors related to the patient or the IVFtreatment. Duration of infertility in months and cause of infertility did
not predict clinical pregnancy. As in the previous analyses, perceived
recent stress and depressive symptoms did not emerge as significant
independent predictors of the DV.
Mediation analysis
As a small correlation was observed between the number of negative
life events and number of oocytes retrieved, we further examined
whether the association between number of negative life events and
pregnancy rates could possibly be mediated by number of oocytes
harvested during oocyte retrieval.
A series of linear regression analyses showed that the number of
negative life events significantly predicted both IVF-outcome and
number of oocytes harvested during the oocyte retrieval procedure.
Furthermore, when controlling for negative life events, the number
of oocytes harvested predicted pregnancy. When controlling for
oocytes harvested, the first association between negative life events
and pregnancy ceased to be significant. The results of the mediation
analysis are shown in Table IV. Applying the criteria for mediation
(Baron and Kenny, 1986), the analysis showed the number of
oocytes harvested during oocyte retrieval to be a partial mediator
of the association between negative life events and pregnancy.
Impact of specific types of life events
Finally, analyses were conducted to explore whether failure to obtain a
pregnancy was associated with specific life events. x2-tests revealed a
trend as to whether the woman reported having experienced violence
or threats of violence within the past 12 months or not: 16 women
reported experiences of violence or threats of violence and none of
these women achieved a viable pregnancy following their first cycle
of IVF (X2(1) ¼ 6.130, P , 0.009 (Fishers Exact Test), total n ¼
795). Furthermore, a trend was seen with respect to whether
couples had experienced prolonged or serious incidents of illness
Table IV Mediation analysis
Testing for mediation of the association between number of
negative life events and IVF outcome
........................................................................................
Step
Coeff. (B)
P
........................................................................................
1: Negative life events predict IVF-outcome
20.012
0.028
2: Negative life events predict number of
oocytes
20.154
0.005
0.014
0.000
3: Number of oocytes predict IVF-outcome,
controlling for negative life events
4: Negative life events—IVF-outcome
association reduced, when controlling for
the number of oocytes
N ¼ 791 (step 1), 772 (step 2, 4), 773 (step 3).
20.010
0.072 (ns)
among close relatives. A larger proportion of the women who
reported having had a parent with serious illness did not achieve IVFpregnancy (21% pregnant) in the year following this event compared
with those who did not report this type of event (29% pregnant)
(X2(1) ¼ 3.691, P , 0.055, total n ¼ 795) (Bonferroni adjusted
p-level of 0.001 (0.05/37)).
Discussion
Our results showed that women who became pregnant following IVFtreatment reported fewer negative life events than women who did
not obtain pregnancy. The data in this study indicate that the lower
pregnancy rate associated with negative life events is partly mediated
by a lower number of oocytes harvested. Our results thus extend previous findings linking the presence of infertility and relationship stress
with impaired chances of achieving pregnancy with an IVF-procedure
(Facchinetti et al., 1997; Stoleru et al., 1997; Gallinelli et al., 2001;
Klonoff-Cohen et al., 2001; Verhaak et al., 2001; Boivin and
Schmidt, 2005).
Effects of negative life events
on IVF-outcome
Our results are in line with two of the three other studies exploring
associations between environmental stressors and reduced chances
of pregnancy following IVF-treatment. In a 12-month prospective
study by Boivin and Schmidt (2005) of a cohort of 818 Danish
couples in IVF-treatment, an association was found between baseline
perceived infertility-related stress and pregnancy obtained within the
following 12 months of treatment. Among couples reporting that fertility problems had disrupted their lives, were difficult to cope with, or
had caused severe strain on their physical and mental health, fewer
conceived within the subsequent 12 months of treatment when compared with couples not reporting such stressors. The authors also
explored another potential environmental stressor, marital problems,
and found that the number of cycles-to pregnancy was significantly
higher in women reporting high perceived marital stress compared
with women reporting low levels of this stressor. Women who
reported that infertility had caused a crisis in their relationship, given
rise to thoughts of divorce, or placed great strain on their sexual
relationship thus had reduced chances of a successful IVF-outcome
(Boivin and Schmidt, 2005). The preliminary evidence for an association between perceived marital stress and IVF-outcome has also
found support in one other study (Stoleru et al., 1997). One study,
however, has reported the opposite result (Cooper et al., 2007),
but this study was potentially confounded by access to psychological
counselling offered by the fertility clinic, which was not controlled
for statistically. Together with our results, the limited available evidence is suggestive of an association between stress and fertility,
with stress having a detrimental impact on chances of achieving a
viable pregnancy through IVF-treatment.
Depression and perceived stress within the past month were not
significant predictors of IVF-outcome in multivariate analysis. This is
not an uncommon finding, and previous studies have shown low predictive quality of the particular measures used (BDI and PSS-10) concerning IVF-outcomes (Klonoff-Cohen et al., 2001; Smeenk et al.,
2001). The BDI may be too exclusively designed for populations of
2179
Stressful life events and IVF outcome
clinically depressed individuals, failing to capture ongoing mood shifts
in psychologically healthy samples, and several studies have stated
that IVF-patients are not clinically depressed (Greil, 1997; Eugster
and Vingerhoets, 1999).
The PSS-10 measures current experiences of stress in that the timeframe is limited to feelings of stress during the past month, and we
suggest that the fluctuations in current stress captured by the
PSS-10 may not be potent enough to harm the reproductive
system. Consistent with this, perceived stress did not emerge as a significant predictor of IVF-outcome. In contrast to the lack of effect of
perceived current stress, we did find an effect of having experienced
traumatic or negative life events within the past 12 months on pregnancy chances. Major life stressors may be viewed as a more
chronic source of stress. The fact that the effect of negative life
events on IVF-outcome was independent of effects of distress, i.e.
depression and of perceived stress, warrants further discussion.
Negative life events may exert influence on the female reproductive
system in two ways, either directly or indirectly (i.e. through life style
behaviours). It is possible that the experience of negative life events
may result in a subtle stress reaction, that may only be partially perceived (or even suppressed) by the individual. Although thus not
necessarily being considered by the individual as a source of daily (perceived) stress, the negative life events or series of negative life events
may over time lead to an increase in maladaptive coping strategies and
life style behaviours (i.e. diet, drinking, smoking habits etc.) and thus
indirectly impact pregnancy chances. The experience of highly stressful
life events is generally known to be associated with behavioural health
problems (i.e. maladaptive life style behaviours in terms of eating
behaviours, alcohol and cigarette use) (Veenstra et al., 2007;
Bacharach et al., 2008; Hooper et al., 2008; Ogden et al., 2009),
and behavioural health problems are known to be associated with
reduced fertilization chances (Klonoff-Cohen, 2005). Specifically, in
one study the authors investigated female alcohol consumption and
found this related to pregnancy chances (Klonoff-Cohen et al., 2003).
This possibility is however in need of further testing, as our study
only used relatively simple measures of life-style behaviours and had
no appropriate measure of alcohol consumption, which may be an
important life-style variable in this context. BMI was unrelated to
probability of a clinical pregnancy. However, as BMI was based upon
self-reported measures, this may have given rise to bias.
A direct effect of negative life events on the female reproductive
system may take place through release of stress hormones such as
cortisol, which may cause estradiol inhibition indicated by impaired
granulosa cell function and possibly compromised follicular maturation
and lower number of oocytes to be harvested (Lancastle and Boivin,
2005). The more severe the stress, the more cycle disturbance is
likely to be produced, with stress responses inhibiting the reproductive
system through activation of the hypothalamic-pituitary-adrenal (HPA)
axis (Chrousos and Gold, 1992).
In terms of psychological interventions, it seems premature to put
forth suggestions, before direct and/or indirect pathways of a stress
effect on pregnancy chances have been clarified. Should further
research point to mediating effects of maladaptive coping strategies
and life style behaviours (i.e. diet, drinking, smoking habits etc.) in
the association between stress and IVF-outcome, cognitivebehavioural interventions could provide a useful tool for addressing
these maladaptive ways of handling the psychological consequences
of negative life events as well as for implementing alternative health
behaviours. If the effect is unrelated to health behaviour, interventions
targeted at conscious processing of psychological consequences of
experiencing negative life events may alleviate psychobiological stress
reactions and thereby possibly optimize pregnancy chances following
IVF through a reduction of stress-related HPA-axis activity.
Mediating mechanisms
Stress (Facchinetti et al., 1997; Klonoff-Cohen et al., 2001; Gallinelli
et al., 2001) and distress (Sanders and Bruce, 1999; Smeenk et al.,
2001) have been found to be associated with a poorer ovarian functional response to IVF-treatment, and reduced fertilization, implantation and live birth rates. It has been suggested that stress may
impact fertility by interrupting menstrual cycles (Demyttenaere et al.,
1992) or by interacting with aspects of the IVF-treatment (Boivin
and Takefman, 1995); our study has provided preliminary evidence
for the latter.
The number of oocytes retrieved during oocyte retrieval is a critical
characteristic of the ovarian response to IVF-treatment. The quality of
the ovarian response is a potent predictor of IVF-outcome and has
been shown to predict increased clinical pregnancy rates relatively
independent of other factors such as maternal age (Yih et al., 2005).
It has been reported that the biological response of the ovaries
during IVF-treatment (e.g. number of follicles and number of
oocytes) can be predicted by psychological variables (Lancastle and
Boivin, 2005). This suggests a possible association between stress,
number of oocytes, and pregnancy. Wishing to gain a more detailed
knowledge about the mechanisms at work, we used mediation analysis
to explore oocytes as a potential link in the stress-fertility association,
and found that the number of oocytes retrieved was a partial mediator
between negative life events and IVF-pregnancy.
Evidence for a link between stress and number of oocytes has been
reported in a study by Klonoff-Cohen and colleagues, who prospectively
studied infertility-related stress and IVF-treatment-related stress in 151
female IVF-patients (Klonoff-Cohen et al., 2001; Klonoff-Cohen and
Natarajan, 2004). Although they did not find infertility-related stress
to be a significant predictor of pregnancy, trends were found for this
source of stress as a predictor of number of oocytes retrieved (KlonoffCohen et al., 2001). Concerning treatment-related stress, it appeared
that women who prior to treatment had severe concerns about the
medical aspects of the procedure, about missing work, or about
finances, had up to 20% fewer oocytes retrieved and fertilized (KlonoffCohen and Natarajan, 2004).
Together with our findings, the results by Klonoff-Cohen et al. thus
suggest a detrimental effect of stress on the number of oocytes that
can be retrieved during procedure. Our results pointed to further consequences in terms of a reduction in pregnancy chances.
Impact of specific types of life events
Psychological stress refers to an individual’s psychological reaction
when faced with challenges (stressors) exceeding his or her personal
resources, i.e. coping abilities (Lazarus and Folkman, 1984). The
total amount of psychological stress experienced is known to be influenced by the level of subjective control over the situation, which again
is closely related to perceived coping possibilities. According to the
theoretical and empirical works on stress, it is especially the prolonged
2180
exposure to an environmental stressor and/or uncontrollable situations that lead to human psychological and psychobiological stress
with potentially adverse consequences for health (Bloom and Lazerson, 1988). Trends were found in our study that long lasting illness
within the nearest family and experiences with violence were more
highly related to IVF-outcome than other stressors. Long lasting
illness in the family and experienced violence are highly uncontrollable
and therefore likely to produce a very high amount of psychological
stress, which, without proper resolution, may also be of a long
lasting character. Due to the relatively small number of participants
having experienced such life events, these results must however be
considered preliminary, and more studies are needed to confirm
this finding.
Strengths and limitations
This is the first study to examine a possible association between
IVF-outcome and stress following negative life experiences, even
though taking into account procedural stress related to the immediate
treatment situation. Investigating life events within the year before
commencing treatment provides knowledge about stressors of potential damage to the reproductive system that take place at a very early
point in time, compared with the time frame set in the existing studies
on stress and IVF-outcome. This study also includes a sufficiently large
sample to explore mechanisms while at the same time controlling for
relevant demographic, clinical and treatment-related factors. Existing
studies exploring psychological factors in IVF-treatment rarely
include more than 100 participants, and most studies include less
than 50 participants (e.g. Demyttenaere et al., 1991; Boivin and Takefman, 1995; Facchinetti et al., 1997; Csemiczky et al., 2000; Lovely
et al., 2003), which limits the number of IVs in the statistical analyses
and enhances the risk of type-2 error (Tabachnick and Fidell, 2007).
However, our study is not without limitations. We had no precise
monitoring of non-response per se as questionnaires unfortunately
were handed out at the clinic without complete registration of those
who chose to participate and those who declined. Therefore a detailed
non-responder analysis was not possible as some of the women may not
have been offered participation. The problem was addressed by careful
comparison between participants and non-participants in the cohort on
the basis of hospital registry data on relevant variables. Comparisons of
participants with non-participating women in the same cohort showed
that participants included more cases of male factor infertility and ICSI
treatment than non-participants, whereas the latter group consisted
of more cases of idiopathic infertility. Treatment modality and infertility
etiology were however unrelated to pregnancy in our study, and general
differences for these two parameters are therefore less likely to affect
the external validity of our findings concerning stress and IVF-outcome.
In addition, differences were found concerning hormone doses, with
study participants receiving a higher total hormone dose before aspiration compared with the rest of the cohort sample. This could have been
due to a smaller proportion of patients in mild hormone stimulation
protocol among participants, but, unfortunately, we have no registration
of this particular treatment factor for the non-participating part of
the cohort. However, the fact that participants completed the LRE
prior to beginning hormone treatment, together with the similarity of
pregnancy outcomes in the participant and non-participant groups,
excludes hormone treatment as a possible confounder in the association
Ebbesen et al.
between negative life events and IVF-outcome. Participants and nonparticipants were also generally similar with respect to age as well as
life style factors, medical variables, and reproductive history, suggesting
that our results can be considered relatively generalizable.
Use of variables from the hospital register meant acceptance of an
appreciable proportion of missing values on some variables, thus
results from these analyses must be interpreted with some caution,
e.g. comparisons between participants and non-participants. In the
analyses, we were unable to investigate any effects of alcohol consumption as we did not have an appropriate measure of this potential
confounder. Future studies should include measures of alcohol consumption in order to assess this factor as a possible mediator of
stress-infertility associations. Concerning the non-appearance of a statistically significant association of FSH, BMI and other life style habits
with IVF-outcome, we cannot exclude lack of statistical power as a
possible explanation.
Conclusion
Our results indicate that the number of negative life events influences
IVF outcome, an effect that is partly mediated by number of oocytes
harvested during treatment. A large number of negative life-events
may indicate chronic stress, which may only be partially acknowledged
by the individual, and which could reduce the chances of a successful
IVF-outcome through a detrimental impact on the number of oocytes
retrieved. Ovarian response is one of the most important predictors
of pregnancy and may prove to be an important link in the association
between psychological factors, e.g. stress, and pregnancy. More
studies are needed, focusing specifically on stress and preferably
including physiological measures, in order to properly clarify the mechanisms behind the stress-fertility association, both in terms of (1) indirect pathways of this association, i.e. health problem behaviour, and (2)
direct effects in terms of psychoneuroendocrine mechanisms.
Acknowledgements
The authors gratefully thank all the couples who willingly participated
in our study as well as the co-workers of Aarhus University Hospital.
Special thanks are extended to the nurses at the hospital for taking
great interest in our study and for providing invaluable help in
handing out/distributing the questionnaires to all the couples. Also,
the authors would like to thank Søren Christensen for valuable statistical assistance.
References
Bacharach SB, Bamberger PA, Doveh E. Firefighters, critical incidents, and
drinking to cope: the adequacy of unit-level performance resources as a
source of vulnerability and protection. J Appl Psychol 2008;93:155– 169.
Baron RM, Kenny DA. The moderator mediator variable distinction in
social psychological-research - conceptual, strategic, and statistical
considerations. J Pers Soc Psychol 1986;51:1173– 1182.
Beck AT, Steer RA, Brown GK. Beck Depression Inventory, 2nd edn. San
Antonio: The Psychological Corp, 1996.
Berg BJ, Wilson JF, Weingartner PJ. Psychological sequelae of infertility
treatment: the role of gender and sex-role identification. Soc Sci Med
1991;33:1071 – 1080.
Stressful life events and IVF outcome
Bloom FE, Lazerson A. Emotions: the highs and Lows of the brain. In:
Brain, Mind and Behavior. New York: W.H. Freeman and Company,
1988, 207 – 238.
Boivin J, Takefman JE. Stress level across stages of in vitro fertilization in
subsequently pregnant and nonpregnant women. Fertil Steril 1995;
64:802– 810.
Boivin J, Schmidt L. Infertility-related stress in men and women predicts
treatment outcome 1 year later. Fertil Steril 2005;83:1745 – 1752.
Chrousos GP, Gold PW. The concepts of stress and stress system
disorders. Overview of physical and behavioral homeostasis. JAMA
1992;267:1244 – 1252.
Cohen S, Tyrrell DA, Smith AP. Psychological stress and susceptibility to
the common cold. N Engl J Med 1991;325:606– 612.
Cohen S, Williamson GM. Perceived Stress in a Probability Sample of the
United States. In: Spaapan S, Oskamps (eds). The Social Psychology of
Health. California: Sage Publications Inc, 1988, 31 – 67.
Cohen S, Kessler RC, Gordon LU. Conceptualizing stress and its relation
to disease. In: Cohen S, Kessler RC, Gordon LU (eds). Measuring Stress A Guide for Health and Social Scientists. NY: Oxford University Press,
1995, 3 – 26.
Cooper BC, Gerber JR, McGettrick AL, Johnson JV. Perceived
infertility-related stress correlates with in vitro fertilization outcome.
Fertil Steril 2007;88:714– 717.
Csemiczky G, Landgren BM, Collins A. The influence of stress and state
anxiety on the outcome of IVF-treatment: psychological and
endocrinological assessment of Swedish women entering
IVF-treatment. Acta Obstet Gynecol Scand 2000;79:113– 118.
de Klerk C, Hunfeld JAM, Heijnen EMEW, Eijkemans MJC, Fauser BCJM,
Passchier J, Macklon NS. Low negative affect prior to treatment is
associated with a decreased chance of live birth from a first IVF cycle.
Hum Reprod 2008;23:112 – 116.
Demyttenaere K, Nijs P, Evers-Kiebooms G, Koninckx PR. Coping,
ineffectiveness of coping and the psychoendocrinological stress
responses during in-vitro fertilization. J Psychosom Res 1991;
35:231– 243.
Demyttenaere K, Nijs P, Evers-Kiebooms G, Koninckx PR. Coping and
the ineffectiveness of coping influence the outcome of in vitro
fertilization through stress responses. Psychoneuroendocrinology 1992;
17:655– 665.
Eugster A, Vingerhoets AJ. Psychological aspects of in vitro fertilization: a
review. Soc Sci Med 1999;48:575 – 589.
Facchinetti F, Matteo ML, Artini GP, Volpe A, Genazzani AR. An increased
vulnerability to stress is associated with a poor outcome of in vitro
fertilization-embryo transfer treatment. Fertil Steril 1997;67:309– 314.
Ferin M. Stress and the reproductive cycle. J Clin Endocrinol Metab 1999;
84:1768– 1774.
Gallinelli A, Roncaglia R, Matteo ML, Ciaccio I, Volpe A, Facchinetti F.
Immunological changes and stress are associated with different
implantation rates in patients undergoing in vitro fertilization-embryo
transfer. Fertil Steril 2001;76:85 – 91.
Greil AL. Infertility and psychological distress: a critical review of the
literature. Soc Sci Med 1997;45:1679 – 1704.
Harlow CR, Fahy UM, Talbot WM, Wardle PG, Hull MG. Stress and
stress-related hormones during in-vitro fertilization treatment. Hum
Reprod 1996;11:274– 279.
Henderson S, Byrne DG, Duncan-Jones P. Neurosis and the Social
Environment. Sydney: Academic Press, 1981.
Hooper R, Rona RJ, Jones M, Fear NT, Hull L, Wessely S. Cigarette and alcohol
use in the UK Armed Forces, and their association with combat exposures:
a prospective study. Addict Behav 2008;33:1067–1071.
2181
Juul S, Karmaus W, Olsen J. Regional differences in waiting time to
pregnancy: pregnancy-based surveys from Denmark, France,
Germany, Italy and Sweden. Hum Reprod 1999;14:1250– 1254.
Klonoff-Cohen H. Female and male lifestyle habits and IVF: what is known
and unknown. Hum Reprod Update 2005;11:179 – 203.
Klonoff-Cohen H, Natarajan L. The concerns during assisted reproductive
technologies (CART) scale and pregnancy outcomes. Fertil Steril 2004;
81:982– 988.
Klonoff-Cohen H, Chu E, Natarajan L, Sieber W. A prospective study of
stress among women undergoing in vitro fertilization or gamete
intrafallopian transfer. Fertil Steril 2001;76:675– 687.
Klonoff-Cohen H, Lam-Kruglick P, Gonzalez C. Effects of maternal and
paternal alcohol consumption on the success rates of in vitro
fertilization and gamete intrafallopian transfer. Fertil Steril 2003;
79:330– 339.
Lancastle D, Boivin J. Dispositional optimism, trait anxiety, and coping:
unique or shared effects on biological response to fertility treatment?
Health Psychol 2005;24:171 – 178.
Lazarus R, Folkman S. Stress, Appraisal and Coping. Mass: Springer
Publications, 1984.
Lovely LP, Meyer WR, Ekstrom RD, Golden RN. Effect of stress on
pregnancy outcome among women undergoing assisted reproduction
procedures. South Med J 2003;96:548 – 551.
Lumley T, Diehr P, Emerson S, Chen L. The importance of the normality
assumption in large public health data sets. Annu Rev Public Health 2002;
23:151– 169.
Milad MP, Klock SC, Moses S, Chatterton R. Stress and anxiety do not
result in pregnancy wastage. Hum Reprod 1998;13:2296– 2300.
National Board of Health D. IVF-behandlinger i Danmark 2004.
Copenhagen, 2006, 1 –16.
Oddens BJ, den Tonkelaar I, Nieuwenhuyse H. Psychosocial experiences
in women facing fertility problems – a comparative survey. Hum Reprod
1999;14:255 – 261.
Ogden J, Stavrinaki M, Stubbs J. Understanding the role of life events in
weight loss and weight gain. Psychol Health Med 2009;14:239 – 249.
Peterson BD, Newton CR, Rosen KH. Examining congruence between
partners’ perceived infertility-related stress and its relationship to
marital adjustment and depression in infertile couples. Fam Process
2003;42:59 – 70.
Peterson BD, Newton CR, Rosen KH, Skaggs GE. Gender differences in
how men and women who are referred for IVF cope with infertility
stress. Hum Reprod 2006;21:2443 – 2449.
Sanders KA, Bruce NW. Psychosocial stress and treatment outcome following
assisted reproductive technology. Hum Reprod 1999;14:1656–1662.
Schafer JL, Graham JW. Missing data: our view of the state of the art.
Psychol Methods 2002;7:147 – 177.
Schmidt L. Infertility and Assisted Reproduction in Denmark: Epidemiology
and Psychosocial Consequences. Dan Med Bulletin 2006;53:390 – 417.
Schmidt L, Münster K, Helm P. Infertility and the seeking of infertility
treatment in a representative population. Br J Obstet Gynaecol 1995;
102:978– 984.
Schmidt L, Holstein B, Christensen U, Boivin J. Does infertility cause
marital benefit? An epidemiological study of 2250 women and men in
fertility treatment. Patient Educ Couns 2005a;59:244 – 251.
Schmidt L, Holstein BE, Christensen U, Boivin J. Communication and
coping as predictors of fertility problem stress: cohort study of 816
participants who did not achieve a delivery after 12 months of fertility
treatment. Hum Reprod 2005b;20:3248 – 3256.
Selye H. The Stress of Life (rev. ed.). New York: McGraw-Hill, 1976.
Skakkebaek NE, Jorgensen N, Main KM, Rajpert-De Meyts E, Leffers H,
Andersson AM, Juul A, Carlsen E, Mortensen GK, Jensen TK et al. Is
human fecundity declining? Int J Androl 2006;29:2 – 11.
2182
Smeenk JM, Verhaak CM, Eugster A, van Minnen A, Zielhuis GA,
Braat DD. The effect of anxiety and depression on the outcome of
in-vitro fertilization. Hum Reprod 2001;16:1420– 1423.
Smeenk JM, Verhaak CM, Vingerhoets AJ, Sweep CG, Merkus JM,
Willemsen SJ, van Minnen A, Straatman H, Braat DD. Stress and
outcome success in IVF: the role of self-reports and endocrine
variables. Hum Reprod 2005;20:991 – 996.
Stoleru S, Cornet D, Vaugeois P, Fermanian J, Magnin F, Zerah S, Spira A.
The influence of psychological factors on the outcome of the fertilization
step of in vitro fertilization. J Psychosom Obstet Gynaecol 1997;
18:189 – 202.
Stroop JR. Studies of interference in serial verbal reactions. J Exp Psychol
1935;18:643 – 662.
Tabachnick BG, Fidell LS. Using Multivariate Statistics, 5th edn. Boston:
Allyn & Bacon: Pearson Education Inc, 2007.
The Danish Fertility Society. Kliniske guidelines: Primær udredning
ved infertilitet. 2007; http://www.fertilitetsselskab.dk/images/2008_
dokumenter/3.Information18.06.2006.doc.
The Danish Fertility Society. Annual report 2007. 2009.
Ebbesen et al.
Throsby K, Gill R. It’s Different for Men: Masculinity and IVF. Men and
Masculinities 2004;6:330 – 348.
UNESCO. ISCED International Standard Classification of Education
(Re-edition). 2009.
Veenstra MY, Lemmens PH, Friesema IH, Tan FE, Garretsen HF,
Knottnerus JA, Zwietering PJ. Coping style mediates impact of stress
on alcohol use: a prospective population-based study. Addiction 2007;
102:1890– 1898.
Verhaak CM, Smeenk JM, Eugster A, van Minnen A, Kremer JA,
Kraaimaat FW. Stress and marital satisfaction among women before
and after their first cycle of in vitro fertilization and intracytoplasmic
sperm injection. Fertil Steril 2001;76:525– 531.
Webb RE, Daniluk JC. The end of the line: Infertile men’s experiences of
being unable to produce a child. Men and Masculinities 1999;2:6 – 25.
Yih MC, Spandorfer SD, Rosenwaks Z. Egg production predicts a doubling
of in vitro fertilization pregnancy rates even within defined age and
ovarian reserve categories. Fertil Steril 2005;83:24 – 29.
Submitted on December 30, 2008; resubmitted on April 17, 2009; accepted on
April 21, 2009