Content uploaded by Lone Schmidt
Author content
All content in this area was uploaded by Lone Schmidt on Jan 13, 2016
Content may be subject to copyright.
ORIGINAL ARTICLE Reproductive epidemiology
Incidence of depression and influence
of depression on the number of
treatment cycles and births in a
national cohort of 42 880 women
treated with ART
C.S. Sejbaek1,*, I. Hageman2, A. Pinborg3, C.O. Hougaard1, and
L. Schmidt1
1
Department of Public Health, Section of Social Medicine, University of Copenhagen, 5 Oester Farimagsgade, P.O. Box 2099,
Copenhagen K DK-1014, Denmark
2
Psychiatric Center Copenhagen, Copenhagen University Hospital, Copenhagen DK-2100, Denmark
3
Fertility Clinic, Copenhagen University Hospital, Rigshospitalet, Copenhagen DK-2100, Denmark
*Correspondence address. Tel: +45-3532-7534; E-mail: camillasandal.sejbaek@sund.ku.dk
Submitted on July 12, 2012; resubmitted on November 27, 2012; accepted on November 30, 2012
study question: Does prior depression in women treated with assisted reproduction technology (ART) influence the number of
treatment cycles and ART live births?
summary answer: Women with a depression diagnosis prior to ART treatment initiated statistically significantly fewer ART treat-
ment cycles and had a lower mean number of ART live births compared with women with no history of depression.
what is known already: Previous studies have shown an increased prevalence of depressive symptoms in fertility patients than in
the comparison groups.
study design, size, duration: A register-based national cohort study, including all women (n¼42 915) treated with IVF, ICSI,
frozen embryo transfer and oocyte recipient cycle in Denmark from 1 January 1994 to 30 September 2009 extracted from the IVF register
(ART cohort). Data on births and depression diagnoses were obtained by linking to the Danish Medical Birth Register (1994– 2010) and the
Danish Psychiatric Central Research Register (1969– 2010).
participants/materials, setting, methods: For each woman in the ART cohort, we included five age-matched women
from the female background population not having received ART treatment. This comparison group was cross-linked with identical register data
as the ART cohort. Women with incomplete ART information or a depression diagnosis before 18 years of age were excluded; remaining n¼
42 880. The ART cohort was grouped into (i) women with a depression diagnosis and (ii) women never diagnosed with depression. In the ART
group with depression, analyses were specified on women with their first depression prior to ART treatment. In total, 2.6% of the women in the
ART cohort had a depression diagnosis. For the incidence rate ratio (IRR) 39 194 women from the ART cohort (3686 women were excluded
due to migration) were compared with 206 005 women from the age-matched comparison group who did not receive ART treatment.
main results and the role of chance: Of the women in the ART cohort with a depression diagnosis, 34.7% had their first
depression diagnosis prior to ART treatment, 4.7% during ART treatment and 60.7% after ART treatment. The mean number of initiated ART
cycles was significantly lower in the ART group of women having a depression diagnosis prior to ART treatment [2.55 (+1.78)] compared with
the ART group of women without a depression diagnosis [3.22 (+2.31); P,0.001; P,0.001]. Women having a depression diagnosis prior to
ART treatment had a lower mean number of ART live births [0.82 (+0.73)] compared with women without a depression diagnosis [1.03
(+0.81); P,0.001]. The incidence rate of first and recurrent depression diagnoses in the ART cohort was significantly lower compared
with the age-matched background population group; IRR ¼0.80 (P,0.001) and IRR ¼0.77 (P,0.001).
limitations, reasons for caution: Only clinical depression diagnoses treated in a psychiatric hospital setting are included. The
age-matched comparison group from the background population is heterogeneous as it consists of women differing in fertility status (both
mothers and childless women).
&The Author 2013. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
Human Reproduction, Vol.28, No.4 pp. 1100– 1109, 2013
Advanced Access publication on January 8, 2013 doi:10.1093/humrep/des442
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from
wider implications of the findings: Fewer women in the ART cohort developed depression over time compared with the
age-matched background population, which might reflect a healthy patient effect of the women seeking ART treatment. Women with a depres-
sion diagnosis before ART treatment receive fewer ART treatments and are less likely to achieve an ART live birth. These women might be
more vulnerable and we recommend that they be offered more psychiatric attention before starting, as well as during and after ART treatment.
study funding/competing interest(s): Research grants are funded by the Danish Health Insurance Foundation and Merck
Sharp & Dohme. The funders had no influence on the data collection, analyses or conclusions of the study. No conflict of interests to declare.
trial registration number: N/A.
Key words: ART / psychiatric morbidity / depression / epidemiology / live birth
Introduction
Infertility is a disease of the reproductive system defined by the failure to
achieve a clinical pregnancy after 12 months or more of regular unpro-
tected sexual intercourse (Zegers-Hochschild et al.,2009). Population-
based studies from European countries have reported a life-time preva-
lence of infertility ranging from 17 to 26% for women trying to achieve
at least one live birth (Schmidt, 2006;Bhattacharya et al.,2009). In
developed countries an average of 56% of infertile couples seek fertility
treatment (Boivin et al.,2007). In Denmark, 8–10% of the Danish na-
tional birth cohort are born after medically assisted reproduction (MAR)
treatment (Danish Fertility Society) and about half of these children are
born after treatment with assisted reproduction technology (ART)
methods, such as in vitro fertilization (IVF) or intracytoplasmic sperm in-
jection (ICSI). The Danish public health-care system is tax-financed, and
about 50% of all MAR treatments are conducted at public fertility clinics.
Access to MAR in the public health-care system is restricted to women
,40 years of age and to couples who have no children together. A
maximum of three IVF or ICSI transfer cycles with fresh embryos are
reimbursed and if the couple succeeds to achieve a childbirth after
the first or second treatment cycle, no further IVF/ICSI treatment
cycles with fresh embryos are provided by the public health-care
system. Couples are provided an unlimited number of intrauterine in-
semination (IUI) cycles, in practice three to six cycles, and an unlimited
number of frozen embryo transfer (FET) cycles. The private health-care
system has full self-payment for fertility treatment and, according to
Danish law, women older than 45 years of age are not allowed to be
treated.
The Danish legislation includes no lower age limit for fertility treat-
ment. However, MAR treatment is not offered to women or men
younger than 18 years. If the fertility doctor has any concerns (e.g.
due to a mental disorder) regarding the couple’s ability to parent,
the fertility doctor is obligated to contact the social authorities to
conduct an assessment of the couple’s parenting ability. If the assess-
ment is negative, fertility treatment is denied.
Depression has a major impact on life and according to the World
Health Organisation (WHO) depression is expected to be among the
top three causes of the global burden of disease (disability-adjusted
life years) in 2030 (WHO, 2012). In Denmark, the prevalence of
major depression in the general population is 3–4% (Olsen et al.,
2004). A Swedish study reported incidence rates (IRs) of depression
of 4.1 per 1000 person-years [95% confidence interval (CI) 2.3– 5.8]
in women aged 15– 39 years and 3.9 per 1000 person-years (95% CI
2.8– 5.0) in women aged 40– 69 years (Mattisson et al.,2005). An
American study reported an IR for depression of 2.6 per 1000 person-
years (95% CI 1.5– 4.1) for women being 30 years of age or older
(Eaton et al., 2007).
The association between infertility and depression is complex as in-
fertility is a potential risk factor for developing depression, and depres-
sion could be a potential risk factor for infertility. Women who
experience unsuccessful IVF/ICSI treatment are at risk of developing
depressive symptoms (Verhaak et al., 2005;Pasch et al., 2012). But
a history of depression might also influence the chance of achieving
a pregnancy. A review by Williams et al. (2007) showed that
women with a history of major depression had lower fertility rates
compared with the background population. A significantly lower fertil-
ity rate was also found in women with a previous admission to a psy-
chiatric hospital due to severe mental disorders in a Danish
prospective cohort study (Laursen and Munk-Olsen, 2010).
Whether lower fertility rates are due to infertility in women with pre-
vious depression is not clear. Lapane et al. (1995) hypothesized that
direct mechanisms between depression and infertility could include
(i) psychoendocrinological mechanisms (elevated prolactin and cortisol
levels), (ii) psychoimmunological mechanisms (impaired immune
defence) or (iii) behavioural mechanisms (e.g. reduced libido,
increased smoking). On the basis of a cross-sectional study, it was
found that young people with a depressive disorder were more
likely to engage in risky sexual intercourse (more partners, never
using or using condoms only sometimes) and had a higher risk of
having sexually transmitted diseases, which is a potential cause for
later tubal pathology and infertility (Ramrakha et al., 2000). Women
with previous depression were found to have an increased risk of in-
fertility (Lapane et al., 1995) and women suffering from polycystic
ovary syndrome (PCOS) were found to have a high risk for depression
(Bishop et al., 2009). Furthermore, smaller studies have shown a
higher prevalence of depressive symptoms in fertility patients than in
the comparison groups (Domar et al., 1992;Thiering et al., 1993;
Lukse and Vacc, 1999;Volgsten et al., 2008). These studies were
based on questionnaires and/or interviews, and depressive symptoms
were measured by standardized scales. A large register-based Finnish
cohort study (n¼9175) found less hospitalization due to depression
both before and after ART treatment when comparing women with
fertility treatment with their controls, but the findings were not statis-
tically significant (Yli-Kuha et al., 2010). They included women pur-
chasing drugs (1996–1998) used in fertility treatment and compared
with a similar sized control group. Another study among women
trying to conceive reported no association between depressive symp-
toms and fecundity (Lynch et al., 2012); however, all depression
Influence of depression and ART treatment 1101
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from
scores in this study were within the normal range, which limited the
interpretation of the results.
The aims of this study were: (i) to investigate potential differences in
the number of treatment cycles and live births after ART for all
women having a depression diagnosis compared with women with
no depression diagnosis; (ii) to investigate this in women having a de-
pression diagnosis prior to ART; (iii) to assess the prevalence of de-
pression (before, during or after ART); and (iv) to calculate the
crude IR of depression diagnosis in women who received ART treat-
ment compared with an age-matched group without ART treatment
randomly selected from the female background population.
Materials and Methods
This study is a part of the Copenhagen Multi-Centre Psychosocial Infertility
(COMPI) Research Programme initiated in 2000 (Schmidt, 2006). The
overall aims of the programme are to investigate psychosocial and
medical aspects of infertility and MAR. The COMPI Research Programme
is based on several longitudinal data sets.
Study populations
The compulsory Danish IVF register was established in 1994 and includes
records on all ART treatment cycles from both public and private fertility
clinics; i.e. IVF, ICSI, FET and oocyte recipient cycles (Andersen et al.,
1999;Henningsen et al., 2011).
The Danish National ART-Couple cohort
All women registered with at least one ART treatment in the IVF register,
between 1 January 1994 and 30 September 2009, were included in the
Danish National ART-Couple (DANAC) cohort (n¼42 880 women;
Fig. 1). We had excluded 35 women because they were born before 1
January 1946, or born after 31 December 1990, had missing information
of the first/last ART treatment date or were ,18 years of age at the
onset of their first depression (Fig. 1). By cross-linking the unique personal
identification number of each woman in the DANAC cohort with the reg-
isters listed below, we were able to obtain information on depression diag-
noses, socio-demographic background characteristics and all deliveries.
The Danish Psychiatric Central Research register. It provides information on
all psychiatric admissions from 1969 and thereafter. From 1995, all out-
patient treatments and emergency room contacts have also been regis-
tered (Mors et al., 2011). There are no private psychiatric hospitals in
Denmark and, thus, all psychiatric admissions and contacts were at
public hospitals. The included codes for depression diagnosis were
ICD-8 codes (296.0, 296.2) and ICD-10 codes (F32, F33). Depression
diagnoses registered from 1969 to 31 December 2010 were included
and only diagnoses from the date of the participant’s’ 18th birthday and
onwards. The latter cut-off was chosen to only include depression diagno-
ses in adulthood. We used the date of admission or outpatient contact as
the date of the depression diagnosis.
The Danish Medical Birth register. This was established in 1960 and
includes information on all deliveries in Denmark (Knudsen and Olsen,
1998). We divided the deliveries into two groups: (i) ART deliveries,
identified as a delivery with a date of birth from 140 to 308 days (20 –
44 gestational weeks) after the initiation of an ART cycle. For women
where more than one ART treatment cycle matched the criteria of a
subsequentbirth,theARTcyclewiththedateclosesttothedateofde-
livery was chosen; (ii) non-ART deliveries; all other deliveries, i.e. chil-
dren born after a spontaneous, non-treatment-related conception and
children born after fertility treatment other than ART such as ovulation
induction and/or IUI.
National socio-demographic registers. They were established in 1995 at
Statistics Denmark and include information on, e.g. age, education,
income, immigration and emigration.
The Danish Register of Causes of Death. It was established in 1875 and since
1994 has been based on the ICD-10 classification (Helweg-Larsen, 2011).
Women without ART treatment: the age-matched comparison
group
An age-matched comparison group based on the background female popu-
lation was established. For each woman in the DANAC cohort, five age-
matched women from the background population were randomly selected.
Women were excluded from the age-matched comparison group if they
had a record in the IVF register at any time. Cross-linkage with the registers
was performed similarly as described for the DANAC cohort.
The study was approved by the Danish Data Protection Agency (J.nr.
2008-41-2076), the National Board of Health (J.nr. 7-505-29-1658/1)
and Statistics Denmark (J.nr. 703481). As this study was register-based,
the scientific ethical committee did not require notification according to
the Danish Committee Legislation.
Data analyses
All the statistical analyses were conducted using SAS version 9.2 (SAS
Institute Inc., Cary, NY, USA).
DANAC cohort: ART treatment, delivery and depression
We identified all women with at least one depression diagnosis in the
DANAC cohort. The time interval between the date of depression diag-
nosis and the date of initiating and ending ART treatment was re-coded
into three time intervals: Date of first depression diagnosis (i) before,
(ii) during or (iii) after ART treatment. For the women registered with a
depression diagnosis before or after ART treatment, we also assessed
the time interval between the date of the most recent depression diagno-
sis and the date of initiating ART treatment (years) and between the date
of the first depression diagnosis and the date of terminating ART treat-
ment. Of the women in the DANAC cohort, we have a 5-year follow-up
for 64.6% and a 10-year follow-up for 31.4%.
The main outcome measures, highest educational level, ART treatment
parameters (ART treatment cycles, type of ART treatment and infertility
diagnosis), and live births per woman were compared for all women
with a depression diagnosis with women without a depression diagnosis
in the DANAC cohort. Furthermore, we compared the same outcome
measures for women with a depression diagnosis prior to ART treatment
with all women in ART treatment without a depression diagnosis.
Incidence rate and incidence rate ratio
To calculate the incidence rate ratio (IRR) of depression diagnosis, we cal-
culated the IR in the DANAC cohort and in the age-matched comparison
group. For each woman we calculated her risk time in the cohort, thereby
excluding the time periods where she lived outside Denmark and hence
not exposed to registration in the Danish Psychiatric Central Research
Register. The following exclusion criteria were applied: Women who (i)
had incomplete information about their migration; (ii) came to Denmark
after their 18th birthday; (iii) left Denmark before turning 26 years old,
did not return before their 26th birthday and the sojourn outside
Denmark was more than 6 months; and (iv) left Denmark before their
26th birthday and did not return (Fig. 1). After these exclusions the
DANAC cohort comprised 39 194 women, who were included in the cal-
culation of the IR of depression.
1102 Sejbaek et al.
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from
In the age-matched comparison group of women with no history of
ART treatment the following exclusion criteria were applied to calculate
the incidence rates of depression: (i) women born before 1 January
1946, or born after 31 December 1990; (ii) women who died before
the age of 18 years; (iii) women with their first depression diagnosis
before the age of 18 years; and (iv) women meeting the migration exclu-
sion criteria used for the DANAC cohort (Fig. 1). In total, we included
206 005 women from the age-matched background population, women
with no history of ART treatment, to calculate the IR of depression.
The following rules were applied to calculate the risk time per woman for
the IR of depression: (i) journeys undertaken before a woman’s 26th birth-
day were ignored; and (ii) after the woman turned 26 years, we counted the
number of journeys and the length of each journey in months. In the
DANAC cohort 25.4% and in the age-matched comparison group 21.5%
of the women had one or more journeys at some point in their life.
Many of these journeys were undertaken either after finishing high school
or in relation to the woman’s further vocational training before the age of
26 years. Thus, all women were included with their full risk time until 26
years of age. If a woman left Denmark (emigrating) after turning 26 years
for the first and/or second time, the following rules were applied: if the
journey was 6 months or less, the woman was included with the total
risk time, otherwise censored at the date the woman left Denmark.
Upon leaving Denmark for the third time, the woman was censored at
the date of emigration independently of the length of the journey. All
other women were censored either at the date they died or at the end-
point of the study period, 31 December 2010. Following these restrictions,
the IRs for first depression diagnosis and for recurrent depression diagnosis
were calculated for both the DANAC cohort and the age-matched com-
parison group. The IRR was calculated both for first depression diagnosis
and for recurrent depression diagnosis.
Results
DANAC cohort
In total, 1096 (2.6% of 42 880) of women treated with ART had a
depression diagnosis as adults. Of those, 380 women (34.7%) had a
Figure 1 Flow diagram; the exclusion criteria for the women in the DANAC cohort.
a
The DANAC cohort is used for all the analyses comparing women
with a depression diagnosis and women without a depression diagnosis.
b
The subpopulation of the DANAC cohort is used for the analyses of IR and IRR.
Influence of depression and ART treatment 1103
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from
depression diagnosis before their first ART treatment, 51 women (4.7%)
were diagnosed during ART treatment and 665 women (60.7%) had their
first depression diagnosis after their last ART treatment.
Of the 380 women with a depression diagnosis before their first
ART treatment, 11.8% had their most recent depression diagnosis
within a 1-year period before their ART treatment, and 26.1% had
their most recent depression diagnosis within a 2-year period before
ART treatment. Within a 5-year period prior to their first ART treat-
ment, 65.3% had their most recent depression diagnosis. The vast ma-
jority of women with depression diagnoses before ART treatment
were diagnosed within a 10-year period prior to ART treatment
(93.2%; Fig. 2).
Of the 665 women with their first depression diagnosis after ART
treatment, 12.8% had their first depression diagnosis within the first
year after ending ART treatment, and 28.0% had their first depression
diagnosis within the first 2 years following ART treatment. Within a
5-year period after ending ART treatment 55.3% had their first de-
pression diagnosis. The majority of the women (87.8%) were diag-
nosed within 10 years after terminating ART treatment (Fig. 2).
Comparison of women with a history of a
depression diagnosis (n51096) with women
without a depression diagnosis
The mean age when initiating ART treatment was similar in the cohort
of women with a depression diagnosis compared with women without
a depression diagnosis. Women with a depression diagnosis had less
education compared with women without a depression diagnosis
(P,0.001; Table I). Women with a diagnosis of depression initiated
significantly fewer ART treatment cycles [2.79 (+2.03)] than women
without a depression diagnosis [3.22 (+2.31), P,0.001]. Tubal
pathology (32.8%) was significantly more common as the infertility
diagnosis in women with a depression diagnosis compared with
women without a depression diagnosis (27.6%). The proportion of
women with live births (total) was similar in the group with a depres-
sion diagnosis compared with those without, and no difference was
found regarding the mean number of ART live births per woman
with a live birth in women with a depression diagnosis [0.98
(+0.78)] compared with women without a depression diagnosis
[1.03 (+0.81), P¼0.069; Table II].
Comparison of women with their first
depression diagnosis prior to ART treatment
(n5380) with women without a depression
diagnosis
Women with their first depression diagnosis before ART treatment
were on average 8.5 months older at their first ART treatment
(P¼0.007) and had a lower educational level (P,0.001) compared
with women without a depression diagnosis (Table I). A significantly
lower number of initiated ART treatment cycles per women was
found for women with their first depression diagnosis before ART
treatment [2.55 (+1.78)] compared with women without a
Figure 2 Time interval of depression diagnosis in relation to ART treatment. Women with their first depression diagnosis before ART treatment
(left/dark-blue): the most recent depression diagnosis before ART treatment; up to 17 years before starting ART treatment [n¼379; percentage
calculated from n¼380 (One woman with a depression diagnosis 26 years before initiating ART treatment; the case is not included in the
figure.)]. Women with their first depression diagnosis after ART treatment (right/green): first depression diagnosis after terminating ART treatment;
up to 17 years after ART treatment (n¼665; percentage calculated from n¼665).
1104 Sejbaek et al.
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from
depression diagnosis [3.22 (+2.31), P,0.001]. Regarding infertility
diagnosis more women with a depression diagnosis prior to ART
had ‘other female factor’ (20.5%; cervical factor, PCOS, endometriosis
and female sterility) and ‘mixed female-male factor’ (16.3%) than
women without a depression diagnosis (12.2%; 9.2%, respectively;
Table II). The results showed a tendency towards a reduced chance
of a live birth after ART treatment for women with their first depres-
sion diagnosis before ART treatment (47.1%) compared with women
without a depression diagnosis (51.9%, P¼0.062). In the group with
their first depression diagnosis before ART treatment the mean
number of ART live births per woman with a live birth was significantly
lower [0.82 (+0.73)] compared with women without a depression
diagnosis [1.03 (+0.81), P,0.001].
Incidence rate of depression
In the DANAC cohort, excluding women who emigrated (n¼
39 194), 952 women had a depression diagnosis (2.4% of 39 194). In
the age-matched comparison group, 6257 women had a depression diag-
nosis (3.0% of 206 005). As shown in Table III,wefoundanIRRofafirst
depression diagnosis of 0.80 (P,0.001) when comparing the DANAC
cohort with the age-matched comparison group. Thus, women in the
DANAC cohort were less likely to have a first depression diagnosis com-
pared with the age-matched comparison group.
When including recurrent depression diagnoses, we found 2205
separate depression diagnoses for the 952 women with a depression
diagnosis in the DANAC cohort and 15 076 separate depression diag-
noses for the 6257 women with a depression diagnosis in the age-
matched comparison group. Women in the DANAC cohort were
also less likely to experience recurrence of depression diagnoses
compared with the age-matched comparison group; IRR ¼0.77
(P,0.001; Table III).
Discussion
Fewer women in ART treatment experienced a depression diagnosis
as adults compared with women in the age-matched background
population with no history of ART treatment. This was true both
for first time depression as well as recurrent depression diagnoses.
More than 60% of the women with depression in the ART population
were diagnosed after terminating ART treatment. Women with a de-
pression diagnosis prior to ART treatment differed from women
without a depression diagnosis in ART treatment as they had signifi-
cantly fewer ART treatment cycles and fewer ART live births. Of
the women with depression before ART, 1 in 10 had their most
recent depression diagnosis within 1 year prior to initiating ART treat-
ment and more than half of the women had their most recent depres-
sion diagnosis within the 5-year period before initiating ART treatment.
Our finding of fewer women in the DANAC cohort with a depres-
sion diagnosis compared with women in the age-matched comparison
group is in line with the findings from the Finnish study by (Yli-Kuha
.............................................................................................................................................................................................
Table I Socio-demographic characteristics for the DANAC cohort: women with their first depression diagnosis before
ART treatment, women with a history of depression diagnosis, and women without a depression diagnosis in the DANAC
cohort (comparison group).
Women with their first
depression diagnosis
before ART treatment (I)
All women with
depression
diagnosis (II)
Comparison group,
women without a
depression diagnosis
(III)
P-value
a
:
comparing I
versus III
P-value
b
:
comparing II
versus III
Women, n380 1096 41 784 NA NA
Age at onset of
depression (years),
mean (+SD)
28.5 (+5.2) 35.2 (+7.6) NA NA NA
Age at first ART
treatment (years),
mean (+SD)
33.7 (+5.0) 33.1 (+4.8) 33.0 (+4.5) 0.007 0.280
Age range at ART
treatment (years)
21.7– 45.7 19.4–45.8 18.5 – 45.9 NA NA
Educational level
c
,
n(%)
0.001 ,0.001
Low I 82 (21.6) 293 (26.7) 6537 (15.6)
Medium II 179 (47.1) 508 (46.4) 19 753 (47.3)
High III 81 (21.3) 196 (17.9) 8714 (20.9)
Highest IV 32 (8.4) 79 (7.2) 5498 (13.2)
Unknown 6 (1.6) 20 (1.8) 1282 (3.1)
SD, standard deviation.
a
Comparing women with their first depression diagnosis before ART treatment in the DANAC cohort (I) with the comparison group of women without a depression diagnosis in the
DANAC cohort (III).
b
Comparing all women with a depression diagnosis in the DANAC cohort (II) with the comparison group of women without a depression diagnosis in the DANAC cohort (III).
c
Highest educational level: Low I: up to 10 years education; Medium II: upper secondary education, vocational education and academy profession; High III: professional bachelor
programmes; Highest VI: bachelor and master’s programmes and PhD (Andersen et al., 2009).
Influence of depression and ART treatment 1105
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from
.............................................................................................................................................................................................
Table II ART treatment characteristics and deliveries for the DANAC cohort: women with their first depression
diagnosis before ART treatment, women with a history of a depression diagnosis, and women without a depression
diagnosis in the DANAC cohort (comparison group).
Women with their first
depression diagnosis
before ART treatment
(I)
All women with a
depression
diagnosis (II)
Comparison group,
women without a
depression diagnosis
(III)
P-value
a
:
comparing I
versus III
P-value
b
:
comparing II
versus III
Women, n380 1096 41 784 NA NA
Total ART treatment
cycles, N
970 3057 134 704 NA NA
Initiated ART treatment
cycles per woman,
mean (+SD)
2.55 (+1.78) 2.79 (+2.03) 3.22 (+2.31) ,0.001 ,0.001
ART treatment cycles,
women n(%)
,0.001 ,0.001
1–3 289 (76.1) 797 (72.7) 27 237 (65.2)
4–6 81 (21.3) 241 (22.0) 10 951 (26.2)
≥7 10 (2.6) 58 (5.3) 3596 (8.6)
Type of ART treatment
cycles per woman,
mean (+SD)
IVF/ICSI 2.14 (+1.49) 2.30 (+1.68) 2.64 (+1.89) ,0.001 ,0.001
FET 0.31 ( +0.71) 0.33 (+0.74) 0.45 ( +0.87) ,0.001 ,0.001
Oocyte recipient
cycle
0.03 (+0.22) 0.03 (+0.34) 0.03 (+0.27) 0.772 0.539
Infertility diagnosis at
first registered ART
treatment, n(%)
,0.001 0.002
Ovulatory disorder 10 (2.6) 43 (3.9) 2387 (5.7)
Tubal pathology 103 (27.1) 359 (32.8) 11 540 (27.6)
Other female factor
c
78 (20.5) 133 (12.1) 5111 (12.2)
Male factor 84 (22.1) 317 (28.9) 12 550 (30.0)
Mixed female-male
factor
62 (16.3) 95 (8.7) 3860 (9.2)
Unexplained 43 (11.3) 149 (13.6) 6336 (15.2)
Total live births
d
,n(%) NA NA
ART
e
226 (48.5) 750 (58.5) 30 644 (60.2)
Non-ART
f
240 (51.5) 533 (41.5) 20 299 (39.8)
Women with live birth
(total), n(%)
277 (72.9) 765 (69.8) 29 634 (70.9) 0.399 0.419
Women with live birth
after ART treatment,
n(%)
179 (47.1) 554 (50.5) 21 692 (51.9) 0.062 0.371
Total live births per
woman with a live birth,
mean (+SD)
1.68 (+0.75) 1.68 (+0.76) 1.72 (+0.76) 0.420 0.130
ART live births per
woman with a live birth,
mean (+SD)
0.82 (+0.73) 0.98 (+0.78) 1.03 (+0.81) ,0.001 0.069
SD, standard deviation; IVF, in vitro fertilization; ICSI, intracytoplasmic sperm injection; FET, frozen embryo transfer.
a
Comparing the women with their first depression diagnosis before ART treatment (I) with the comparison group of women without a depression diagnosis in the DANAC cohort (III).
b
Comparing all women with a depression diagnosis (II) with the comparison group of women without a depression diagnosis in the DANAC cohort (III).
c
Other female factor: cervical factor, PCOS, endometriosis and female sterility.
d
Live birth: singletons and multiple birth.
e
ART: live births after ART (IVF, ICSI, FET and oocyte recipient cycle).
f
Non-ART: live births after other MAR treatment and non-treatment-related pregnancies.
1106 Sejbaek et al.
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from
et al., 2010). They found that women in IVF treatment had a non-
significantly lower probability of depression both before [adjusted
odds ratio (aOR) ¼0.73; 95% CI 0.49 – 1.10] and after IVF treatment
compared with their controls (aOR ¼0.84; 95%CI 0.64 – 1.16). In
contrast, another study on self-reported depression in infertile
women versus fertile women found a higher probability of depression
in the infertile women (OR ¼1.20; 95%CI 1.01– 1.43); but infertile
women with self-reported depression also sought less fertility treatment
(Herbert et al., 2010). These findings suggest that infertile women who
seek fertility treatment are less depressed. However, patients seeking
infertility treatment may also underreport depressive symptoms out of
fear that it might have consequences for the treatment.
To our knowledge this is the first study investigating the IR of de-
pression in an ART cohort with comparison with an age-matched
background population with no history of ART treatment. We
cannot exclude a healthy patient effect as women who choose to ini-
tiate fertility treatment may be more robust and resistant before,
during and after the fertility treatment than infertile women who do
not undergo ART treatment. An earlier study has shown that
women seeking fertility treatment are more likely to have well-
adjusted relationships with their partners (Edelmann et al., 1994),
which may be the reason why fewer women with a depression diag-
nosis initiate fertility treatment in our ART population compared
with the age-matched comparison group. On the other hand, we
only included the most severe cases of depression, as only women ad-
mitted to hospital or having outpatient contact in hospital settings for
their depression were included in the DANAC cohort as well as the
age-matched comparison group. Women receiving ART treatment
with mild to moderate depression are in general treated by their
general practitioner or at a psychiatric setting outside the hospital.
The general practitioners prescribe close to 85% of all the dispensed
antidepressant medicine in Denmark. This fact underlines that only
severe major depression is included in our study. Furthermore, most
cases of depression found in a Swedish study among fertility patients
were not diagnosed nor treated (Volgsten et al., 2008).
The IRR of depression was calculated among all participants irre-
spectively of whether they had achieved motherhood or not both in
the DANAC cohort and in the age-matched comparison group.
Even though the comparison group was age-matched to the
DANAC cohort, the comparison group is heterogeneous. The com-
parison group includes women who are (yet) voluntarily childless,
woman who are involuntarily childless without seeking ART
treatment, fertile women having achieved motherhood without ART
treatment and infertile women having achieved motherhood after
other types of MAR treatments than ART.
A recent meta-analysis including 14 prospective studies of women
in ART treatment (Boivin et al., 2011) found no association
between pretreatment emotional distress (self-reported anxiety symp-
toms and/or depressive symptoms) and treatment outcome after one
cycle of ART, which is also supported by a recent study by Pasch et al.
(2012).Matthiesen et al. (2011) reported a non-significant trend for
the association between self-reported depressive symptoms and
lower chance of a clinical pregnancy in their meta-analysis. In a pro-
spective study including 400 women Volgsten et al. (2010) found
that women with major depression during IVF/ICSI treatment had a
significantly lower pregnancy rate as well as a lower live birth rate
compared with women with no psychiatric disorder. Our findings
partly support these findings. The association of a lower number of
ART live births in women with a depression diagnosis was statistically
significant for women with a depression diagnosis before initiating ART
treatment. The majority of previous studies are surveys with depres-
sion diagnostics based on self-reported questionnaires and/or inter-
views with the ability of identifying patients with depressive
symptoms and hence less severe depression. Surveys may also over-
estimate the true prevalence of depression and are limited by the
risk of selection bias due to the non-responders.
Our data showed that women with a depression diagnosis before
ART treatment had a higher prevalence of ‘other female factor’,
which includes PCOS as a cause of infertility. In a meta-analysis,
Veltman-Verhulst et al. (2012) found that women with PCOS
showed higher emotional distress (depressive symptoms and
anxiety) compared with their controls. However, infertility could
not, together with hirsutism and obesity, explain the whole association
between PCOS and emotional distress. One explanation for our find-
ings could be that women with PCOS are more likely to develop a
depression even before entering ART treatment.
In general, women with a depression diagnosis terminated ART
treatment before the completion of their three fully reimbursed
ART treatment cycles as the average number of treatment cycles
was below three. One reason for the lower number of ART treatment
cycles in women with depression may be that women with a depres-
sion diagnosis have a lower educational level compared with women
without a depression diagnosis in this cohort. A lower educational
level is associated with a lower income, which may hinder women
.............................................................................................................................................................................................
Table III IRs for the DANAC cohort (n539 194) and for the age-matched comparison group (n5206 005).
DANAC cohort, IR per
1000 person-years
Age-matched comparison group,
IR per 1000 person-years
DANAC cohort versus
age-matched comparison group,
IRR
P-value
a
First depression
diagnosis
1.07 1.33 0.80 ,0.001
Recurrent
depression
diagnoses
2.47 3.21 0.77 ,0.001
IRR comparing the DANAC cohort and the age-matched comparison group after exclusion of women who emigrated.
a
P-value for the IRR.
Influence of depression and ART treatment 1107
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from
with depression in having self-funded ART treatment. Thus, the lower
number of ART treatments may be explained by the lower income
and not by the depression diagnosis. Furthermore, women with a de-
pression prior to ART treatment also had a lower number of live
births conceived after ART treatment compared with women
without a depression diagnosis. The lower number of live births
could be a result of the lower number of treatment cycles among
women with a depression diagnosis before ART treatment. Addition-
ally, both women who had a depression diagnosis before initiating
ART treatment and women who developed a depression at any
point in their life may have a lower threshold for stress related to fer-
tility treatment. Hence, they may terminate ART treatment earlier
than women without a depression diagnosis.
More women had their first depression diagnosis after ART treat-
ment rather than before/during ART treatment, which may partly
be explained by the change in the registration in the Psychiatric
Central Research Register. From 1995 outpatients and emergency
contacts were also included in the register. Hence, less severe cases
of depression are included in the register from 1995, which may
lead to more registered cases of depression after ART treatment.
However, this change in registration will not have any impact on the
IRR because the change occurred both for the DANAC cohort, as
well as for the age-matched comparison group. Furthermore, the
two populations were age-matched and the changes would therefore
occur similarly in the two populations. After terminating ART treat-
ment, women having mild or moderate depression may be more
willing to seek treatment because nothing is at stake in relation to
the ART treatment at this point. On the other hand, the burden of
the ART treatment may also amplify the depressive symptoms,
which would lead to more women being diagnosed with depression
within a certain time span after ART treatment. Women with
severe depression are most likely not able to hide their symptoms
from health professionals.
Strengths and limitations
The major strength of this study is the large national population and
the longitudinal design. Furthermore, another significant and important
strength is that the diagnosis of depression is made in a clinical setting
and by a psychiatrist, as opposed to studies where the depression
diagnosis is based on self-administered questionnaires. Finally, a clear
advantage was that we were able to categorize the deliveries in
ART and non-ART by cross-linkage of the IVF register and Medical
Birth Register.
A paper presenting the IVF register and the research possibilities of
the register has revealed the possibility of late or double entry of treat-
ment cycles, sometimes due to paper-based reports. However, the
size of this problem is unknown (Blenstrup and Knudsen, 2011).
Hence, some women who received ART treatment during the study
period are missing or recorded with too few treatments in the IVF
register, while others appear twice due to double entry. However,
this registration inaccuracy is similar for women with and without de-
pression in the ART cohort and should not bias the results. Despite
the problems with the data entry in the IVF register, the conclusion
is that the IVF register is considered usable for research purposes
(Blenstrup and Knudsen, 2011).
Conclusion
Women in ART treatment seem to have less risk of developing de-
pression and recurrent depression than an age-matched female popu-
lation. This is of course not due to a benefit of the ART treatment per
se but probably due to the fact that the women with severe and recur-
rent depression to a lesser extent seek fertility treatment or disclose
depressive symptoms before or during fertility treatment. It is reassur-
ing that moderate to severe depression is observed at a lower inci-
dence rate in the ART population compared with the age-matched
comparison group; however, ART professionals should be aware of
the strain ART treatment places on these women and the conse-
quences it might have.
Women with a depression diagnosis prior to ART treatment go
through less ART treatment cycles and have fewer live births after
ART, indicating that the fertility treatment puts greater strain on
them than on women with no history of depression. Thus, this
study emphasizes that ART professionals should take precautions
regarding this vulnerable group of patients.
Acknowledgements
This study is part of the COMPI Research Programme initiated by Dr
L. Schmidt, University of Copenhagen, 2000. We thank Gurli Pilgaard
Perto, the Danish Psychiatric Central Research Register; Jørn Korsbø
Petersen, Statistics Denmark; and Steen Rasmussen, the National
Board of Health for their efforts in extracting and linking register
data for this project. We thank Henrik Brønnum-Hansen, University
of Copenhagen, for providing guidance for the statistical analyses of
the incidence rates.
Authors’ roles
All authors contributed to the concept and the design of the study.
A.P., C.O.H. and L.S. obtained the data. C.S.S. and C.O.H. performed
the data analysis. C.S.S. drafted the article. All other co-authors con-
tributed towards the data interpretation, critical revision of the paper
and final approval of the manuscript.
Funding
C.S.S. has a PhD grant funded by the Danish Health Insurance Foun-
dation (J.nr. 2008B105) and Merck Sharp & Dohme (MSD). The spon-
sors had no influence on how data were retrieved and analysed or on
the conclusions of the study.
Conflict of interest
None declared.
References
Andersen AN, Westergaard HB, Olsen J. The Danish in vitro fertilisation
(IVF) register. Dan Med Bull 1999;46:357–360.
Andersen I, Thielen K, Nygaard E, Diderichsen F. Social inequality in the
prevalence of depressive disorders. J Epidemiol Community Health
2009;63:575–581.
1108 Sejbaek et al.
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from
Bhattacharya S, Porter M, Amalraj E, Templeton A, Hamilton M, Lee AJ,
Kurinczuk JJ. The epidemiology of infertility in the North East of
Scotland. Hum Reprod 2009;24:3096–3107.
Bishop SC, Basch S, Futterweit W. Polycystic ovary syndrome, depression,
and affective disorders. Endocr Pract 2009;15:475–482.
Blenstrup LT, Knudsen LB. Danish registers on aspects of reproduction.
Scand J Public Health 2011;39:79–82.
Boivin J, Bunting L, Collins JA, Nygren KG. International estimates of
infertility prevalence and treatment-seeking: potential need and
demand for infertility medical care. Hum Reprod 2007;22:1506 – 1512.
Boivin J, Griffiths E, Venetis CA. Emotional distress in infertile women and
failure of assisted reproductive technologies: meta-analysis of
prospective psychosocial studies. Br Med J 2011;342:d223.
Danish Fertility Society. Anual Reports. http://www.fertilitetsselskab.dk/
(17 December 2012, date last accessed).
Domar AD, Broome A, Zuttermeister PC, Seibel M, Friedman R. The
prevalence and predictability of depression in infertile women. Fertil
Steril 1992;58:1158–1163.
Eaton WW, Kalaydjian A, Scharfstein DO, Mezuk B, Ding Y. Prevalence
and incidence of depressive disorder: the Baltimore ECA follow-up,
1981–2004. Acta Psychiatr Scand 2007;116:182–188.
Edelmann RJ, Connolly KJ, Bartlett H. Coping strategies and psychological
adjustment of couples presenting for IVF. J Psychosom Res 1994;
38:355–364.
Helweg-Larsen K. The Danish register of causes of death. Scand J Public
Health 2011;39:26–29.
Henningsen AK, Romundstad LB, Gissler M, Nygren KG, Lidegaard O,
Skjaerven R, Tiitinen A, Andersen AN, Wennerholm UB, Pinborg A.
Infant and maternal health monitoring using a combined Nordic database
on ART and safety. Acta Obstet Gynecol Scand 2011;90:683–691.
Herbert DL, Lucke JC, Dobson AJ. Depression: an emotional obstacle to
seeking medical advice for infertility. Fertil Steril 2010;94:1817 – 1821.
Knudsen LB, Olsen J. The Danish Medical Birth Registry. Dan Med Bull
1998;45:320–323.
Lapane KL, Zierler S, Lasater TM, Stein M, Barbour MM, Hume AL. Is a
history of depressive symptoms associated with an increased risk of
infertility in women? Psychosom Med 1995;57:509–513.
Laursen TM, Munk-Olsen T. Reproductive patterns in psychotic patients.
Schizophr Res 2010;121:234–240.
Lukse MP, Vacc NA. Grief, depression, and coping in women undergoing
infertility treatment. Obstet Gynecol 1999;93:245–251.
Lynch CD, Sundaram R, Buck Louis GM, Lum KJ, Pyper C. Are increased
levels of self-reported psychosocial stress, anxiety, and depression
associated with fecundity? Fertil Steril 2012;98:453 – 458.
Matthiesen SM, Frederiksen Y, Ingerslev HJ, Zachariae R. Stress, distress
and outcome of assisted reproductive technology (ART): a meta-
analysis. Hum Reprod 2011;26:2763– 2776.
Mattisson C, Bogren M, Nettelbladt P, Munk-Jorgensen P, Bhugra D. First
incidence depression in the Lundby Study: a comparison of the two
time periods 1947–1972 and 1972–1997. JAffectDisord2005;
87:151–160.
Mors O, Perto GP, Mortensen PB. The Danish psychiatric central research
register. Scand J Public Health 2011;39:54– 57.
Olsen LR, Mortensen EL, Bech P. Prevalence of major depression and
stress indicators in the Danish general population. Acta Psychiatr Scand
2004;109:96–103.
Pasch LA, Gregorich SE, Katz PK, Millstein SG, Nachtigall RD, Bleil ME,
Adler NE. Psychological distress and in vitro fertilization outcome.
Fertil Steril 2012;98:459–464.
Ramrakha S, Caspi A, Dickson N, Moffitt TE, Paul C. Psychiatric disorders
and risky sexual behaviour in young adulthood: cross sectional study in
birth cohort. Br Med J 2000;321:263–266.
Schmidt L. Infertility and assisted reproduction in Denmark. Epidemiology
and psychosocial consequences.Dan Med Bull 2006;53:390–417.
Thiering P, Beaurepaire J, Jones M, Saunders D, Tennant C. Mood state as
a predictor of treatment outcome after in vitro fertilization/embryo
transfer technology (IVF/ET). J Psychosom Res 1993;37:481 – 491.
Veltman-Verhulst SM, Boivin J, Eijkemans MJ, Fauser BJ. Emotional distress
is a common risk in women with polycystic ovary syndrome: a
systematic review and meta-analysis of 28 studies. Hum Reprod Update
2012;18:638–651.
Verhaak CM, Smeenk JM, van MA, Kremer JA, Kraaimaat FW. A
longitudinal, prospective study on emotional adjustment before, during
and after consecutive fertility treatment cycles. Hum Reprod 2005;
20:2253–2260.
Volgsten H, Skoog SA, Ekselius L, Lundkvist O, Sundstrom P I. Prevalence
of psychiatric disorders in infertile women and men undergoing in vitro
fertilization treatment. Hum Reprod 2008;23:2056– 2063.
Volgsten H, Skoog SA, Ekselius L, Lundkvist O, Sundstrom P I. Risk factors
for psychiatric disorders in infertile women and men undergoing in vitro
fertilization treatment. Fertil Steril 2010;93:1088– 1096.
WHO (World Health Organization). Depression. http://www.emro.who
.int/health-topics/depression/ (17 December 2012, date last accessed).
Williams KE, Marsh WK, Rasgon NL. Mood disorders and fertility in
women: a critical review of the literature and implications for future
research. Hum Reprod Update 2007;13:607– 616.
Yli-Kuha AN, Gissler M, Klemetti R, Luoto R, Koivisto E, Hemminki E.
Psychiatric disorders leading to hospitalization before and after
infertility treatments. Hum Reprod 2010;25:2018–2023.
Zegers-Hochschild F, Adamson GD, de Mouzon J, Ishihara O, Mansour R,
Nygren K, Sullivan E, van der Poel S. The International Committee for
Monitoring Assisted Reproductive Technology (ICMART) and the
World Health Organization (WHO) Revised Glossary on ART
Terminology, 2009. Hum Reprod 2009;24:2683– 2687.
Influence of depression and ART treatment 1109
by guest on October 27, 2015http://humrep.oxfordjournals.org/Downloaded from