History of depression and risk of hyperemesis gravidarum: a population-based cohort study

Abstract

Hyperemesis gravidarum (HG) is a pregnancy condition characterised by debilitating nausea and vomiting. HG has been associated with depression during pregnancy but the direction of the association remains unclear. The aim of this study was to assess whether previous depression is associated with HG. This is a population-based pregnancy cohort study using data from The Norwegian Mother and Child Cohort Study. The study reviewed 731 pregnancies with HG and 81,055 pregnancies without. Logistic regression analyses were performed to examine the association between a lifetime history of depression and hyperemesis gravidarum. Odds ratios were adjusted for symptoms of current depression, maternal age, parity, body mass index, smoking, sex of the child, education and pelvic girdle pain. A lifetime history of depression was associated with higher odds for hyperemesis gravidarum (aOR = 1.49, 95% CI (1.23; 1.79)). Two thirds of women with hyperemesis gravidarum had neither a history of depression nor symptoms of current depression, and 1.2% of women with a history of depression developed HG. A lifetime history of depression increased the risk of HG. However, given the fact that only 1.2% of women with a history of depression developed HG and that the majority of women with HG had no symptoms of depression, depression does not seem to be a main driver in the aetiology of HG.

Full text

ORIGINAL ARTICLE
History of depression and risk of hyperemesis gravidarum:
a population-based cohort study
Helena Kames Kjeldgaard
1,2
&Malin Eberhard-Gran
1,2,3
&JūratėŠaltytėBenth
1,2
&
Hedvig Nordeng
3,4
&Åse Vigdis Vikanes
5
Received: 21 December 2016 /Accepted: 26 December 2016 /Published online: 7 January 2017
#The Author(s) 2017. This article is published with open access at Springerlink.com
Abstract Hyperemesis gravidarum (HG) is a pregnancy con-
dition characterised by debilitating nausea and vomiting. HG
has been associated with depression during pregnancy but the
direction of the association remains unclear. The aim of this
study was to assess whether previous depression is associated
with HG. This is a population-based pregnancy cohort study
using data from The Norwegian Mother and Child Cohort
Study. The study reviewed 731 pregnancies with HG and
81,055 pregnancies without. Logistic regression analyses
were performed to examine the association between a lifetime
history of depression and hyperemesis gravidarum. Odds ra-
tios were adjusted for symptoms of current depression, mater-
nal age, parity, body mass index, smoking, sex of the child,
education and pelvic girdle pain. A lifetime history of depres-
sion was associated with higher odds for hyperemesis
gravidarum (aOR = 1.49, 95% CI (1.23; 1.79)). Two thirds
of women with hyperemesis gravidarum had neither a history
of depression nor symptoms of current depression, and 1.2%
of women with a history of depression developed HG. A life-
time history of depression increased the risk of HG. However,
given the fact that only 1.2% of women with a history of
depression developed HG and that the majority of women
with HG had no symptoms of depression, depression does
not seem to be a main driver in the aetiology of HG.
Keywords Depression .Hyperemesis gravidarum .Mental
health .Nausea and vomiting .Norwegian Mother and Child
Cohort Study
Introduction
Nausea and vomiting in pregnancy (NVP) is common and
affects up to 80% of all pregnancies (Gadsby et al. 1993).
Unlike NVP, hyperemesis gravidarum (HG) is characterised
by severe, debilitating symptoms. The International
Classification of Diseases (ICD-10) describes HG as exces-
sive vomiting starting before the 22nd week of gestation with
(severe HG) or without (mild HG) metabolic disturbances
(World Health Organization 2004). Although estimated to af-
fect 0.3 to 2% of all pregnancies (Eliakim et al. 2000), HG is a
primary reason for sick leave (Dorheim et al. 2013)and
hospitalisation during pregnancy (Gazmararian et al. 2002).
The aetiology and the pathogenesis of HG are unclear, and it
remains unknown whether NVP and HG are independent con-
ditions or if HG represents the extreme of a continuum of
NVP.
HG has historically been explained by a variety of psycho-
logical mechanisms that have been subjected to stigma
(Fairweather 1968). Other hypotheses have been proposed,
including genetic components (Corey et al. 1992; Fejzo
et al. 2008), endocrine factors and Helicobacter pylori infec-
tion, but none of these have proven sufficient to explain HG
(Verberg et al. 2005). Although, HG is today considered a
disease of unclear pathophysiology (Grooten et al. 2015),
*Helena Kames Kjeldgaard
Helena.Kames.Kjeldgaard@ahus.no
1
Health Services Research Unit, Akershus University Hospital, Post
Box 1000, 1478 Lørenskog, Norway
2
Institute of Clinical Medicine, Campus Ahus, University of Oslo,
Lørenskog, Norway
3
Domain for Mental and Physical Health, Norwegian Institute of
Public Health, Oslo, Norway
4
PharmacoEpidemiology & Drug Safety Research Group,
Department of Pharmacy, School of Pharmacy, University of Oslo,
Oslo, Norway
5
The Intervention Centre, Oslo University Hospital, Oslo, Norway
Arch Womens Ment Health (2017) 20:397–404
DOI 10.1007/s00737-016-0713-6
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

clinical practice still includes evaluation of hyperemetic wom-
en for psychiatric disease (Kim et al. 2009). Women with HG
report lack of support from their healthcare providers
(Heitmann et al. 2016; Poursharif et al. 2008), which may
have severe consequences such as termination of pregnancy
and psychological sequelae (Poursharif et al. 2008;Poursharif
et al. 2007).
HG has consistently been associated with mental distress
such as depression and anxiety. Previous studies are, however,
often small with a medium to high risk of bias (Mitchell-Jones
et al. 2016) or have limited availability of co-variates (Fell
et al. 2006;Sengetal.2007). Prior research has mainly fo-
cused on the association between anxiety/depression and HG
during pregnancy, whereas the effect of anxiety/depression
prior to pregnancy remains to be elucidated. Furthermore,
few studies have used reliable psychometric instruments to
assess anxiety/depression before pregnancy, rendering causal
inferences difficult (Fell et al. 2006;Sengetal.2007). Thus, a
key question remains of whether mental distress leads to HG
or HG leads to mental distress.
The aim of the present study was to assess whether a life-
time history of depression is associated with HG. The
Norwegian Mother and Child Cohort Study, comprising more
than 100,000 pregnancies, provides a unique opportunity to
explore this association.
Materials and methods
Study design and study population
From 1998 to 2008, all pregnant women scheduled to give
birth at 50 of Norway’s 52 hospitals with maternity units re-
ceived a postal invitation to participate in The Norwegian
Mother and Child Cohort Study (MoBa) together with ap-
pointments for routine ultrasound examination at around week
17 of pregnancy. All participants signed an informed consent
form (Magnus et al. 2016;Magnusetal.2006). MoBa was
approved by the Regional Committee for Medical Research
Ethics and by the Norwegian Data Protection Authority. The
protocol for the current study was submitted to the Norwegian
Institute of Public Health, who, upon approval, supplied the
researchers of this study with anonymised data through con-
tract (PDB 1527, www.fhi.no/moba).
The current study is based on version 8 of the quality-assured
data files linked to the Medical Birth Registry of Norway
(MBRN). The MBRN is based on the compulsory notification
of every birth or late abortion in Norway from the 16th week of
gestation, including information regarding pregnancy-related
complications (Irgens 2000). Approximately 40% of the invited
women participated, and each pregnancy was registered with a
unique identification number (Magnus et al. 2006).
The analyses of the current study are based on two ques-
tionnaires distributed in pregnancy week 17 (Q1) and week 30
(Q2). Q1 covers background factors including previous preg-
nancies, medical history before and during pregnancy, medi-
cation; occupation, lifestyle habits and mental health. Q2 pro-
vides information about the mental and physical health at this
stage of pregnancy as well as changes in work situation and
habits. English translations of the questionnaires can be found
at http://www.fhi.no/moba.
We included all singleton pregnancies (n= 112,288). We
excluded women with missing information on history of de-
pression (n= 3605), symptoms of depression at the 17th ges-
tational week, hospitalisation (n= 19,275), sex of the child
(n= 207) and education (n= 15,707). Some women had
missing values on more than one variable. The final sample
comprised 81.786, 72.8% of the total sample.
Variables
In accordance with previous studies on MoBa data (Vikanes
et al. 2010,2013), HG was defined as prolonged nausea and
vomiting leading to hospitalisation before the 25th gestational
week as reported in Q2 (week 30). This definition was chosen
in order to clearly separate HG from normal levels of NVP.
The main predictor was a lifetime history of depression,
measured by the Kendler’s lifetime major depression scale
(KLTDS). The KLTDS was defined using five of the nine
symptomatic criteria for major depression in DSM-III-R:
Have you ever experienced the following for a continuous
period of 2 weeks or more: (1) felt depressed, sad; (2) had
problems with appetite or eaten too much; (3) been bothered
by feeling weaker or a lack of energy; (4) really blamed your-
self and felt worthless and (5) had problems with concentra-
tion or had problems making decisions. The response to each
question was yes or no. A history of depression was defined as
present if a minimum of three of the five symptoms and sad
mood were reported to occur simultaneously for more than
2 weeks (Kendler et al. 1993).
A five-item short version (SCL-5) of the Hopkins
Symptom Checklist-25 (SCL-25) was used as a proxy for
current depression in pregnancy week 17. The SCL-5 is high-
ly correlated with the SCL-25 (correlation coefficient of 0.92)
(Tambs and Moum 1993) and consists of the following ques-
tions: Have you been bothered by any of the following during
the last 2 weeks: (1) feeling fearful, (2) nervousness or shak-
iness inside, (3) feeling hopeless about the future, (4) feeling
blue and (5) worrying too much about things. The response
categories ranged from ‘not bothered’to ‘very bothered’
(range 1–4), with a maximum total score of 20. Symptoms
of current depression were defined as a mean score >2
(Strand et al. 2003), which has been shown to provide the
same prevalence estimate of a depressive disorder as the
Composite International Diagnostic Interview (Robins et al.
398 Kjeldgaard H.K. et al.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

1988; Sandanger et al. 1998).Missingvaluesinthe
dichotomised version of the SCL-5 were handled as follows.
First, the average score on existing items was calculated for
each case if at least three of five questions were answered. If
the average of the existing items was clearly above or below
the cut-off and could not be affected by imputation of missing
values, it was dichotomised to zero or one, as appropriate.
Imputation was not performed in cases where the average
score was not uniquely defining the value above or below
cut-off. Altogether, N= 18 cases were imputed.
Co-variates and possible confounders obtained from the
MBRN included sex of the child (Rashid et al. 2012), maternal
age and parity. Co-variates and possible confounders obtained
from MoBa Q1 were socio-economic status, BMI and
smoking (Vikanes et al. 2010). Pelvic girdle pain was obtained
from MoBa Q2 (Bjelland et al. 2013; Chortatos et al. 2015).
Regarding parity, women were dichotomised as either primip-
arous or multiparous. Education was used as a proxy for
socio-economic status, and length of education (in years)
was divided into three categories. Pre-pregnancy body mass
index (BMI) was calculated as weight/height
2
. Women shorter
than 120 cm (n= 199) and women weighing more than150 kg
or less than 40 kg were excluded (n= 58). Also, those
reporting reduction in weight by more than 20 kg or increase
in weight by more than 50 kgsince the start of pregnancy were
excluded (n= 65). Smoking was assessed as a yes/no response
to the question ‘did you smoke 3 months before pregnancy’
(Vikanes et al. 2010). Pelvic girdle pain was defined as pain in
the anterior pelvis and on both sides in the posterior pelvis
(Bjelland et al. 2013).
Other co-variates including H. pylori infection (Li et al.
2015), gastrointestinal disorders, rheumatoid arthritis, pre-
eclampsia, chronic hypertension, type 1 diabetes, asthma
(Bolin et al. 2013; Fell et al. 2006; Jorgensen et al. 2012),
eating disorders (Torgersen et al. 2008) and ethnicity
(Vikanes et al. 2008) were considered but not included in the
final analysis due to a small number of women with these
disorders in the HG group. Thyroid disease was not included
in the analysis as the questionnaire form does not allow dif-
ferentiation between hypothyroid and hyperthyroid disease.
Statistical analysis
Demographic and clinical characteristics among women with
and without HG and for the entire sample were presented as
frequencies and percentages or means and standard deviations
(SD).
To assess the association between a lifetime history of de-
pression and HG, a logistic regression model was estimated.
Due to multiple births, some women had several recordings in
the data set. According to the intra-women correlation coeffi-
cient, there was some degree of clustering detected. Thus, the
generalised estimating equations (GEE) model correctly
adjusting the estimates for intra-women correlations was
fitted.
A number of potential predictors and confounders were
considered. In order to test our hypotheses, a data splitting
approach was applied (Dahl et al. 2008). According to this
approach, the data set was split into two random parts contain-
ing approximately 30% (part I) and 70% (part II) of observa-
tions. Splitting was performed within stratas defined by sev-
eral key variables. Part I (pilot) was used to construct a model
for HG. Only predictors significant at the 5% level or those
otherwise considered important were left in the model estimat-
ed on pilot data. The hypothesis testing was then performed on
part II (test) data. Only the results with Pvalues below 0.05 in
the test data analyses were accepted as significant, regardless
of significance level in the pilot part. Once the hypotheses
were tested, the model was estimated on the entire data set
to achieve most accurate estimates for the model parameters.
Due to the numerous predictors considered, the level of sig-
nificance was set to 0.005 when interpreting the results in the
entire data set.
The interaction between BMI and smoking status was
assessed and kept in the model if significant.
All analyses were performed by SPSS v 22.
Results
Characteristics for the HG group and comparison group are
presented in Table 1. The mean age of pregnant women was
30.3 years (15–47 years; SD 4.5 years) and 45% were primip-
ara. A total of 731 (0.9%) women reported hospital admission
due to HG. More than 20% (17,351/81,786) of the women
reported a lifetime history of depression, whereas 6.1%
(4981/81,786) reported symptoms of current depression at
the 17th gestational week.
In the binary logistic regression model, a lifetime history of
depression was associated with higher odds for HG (unadjust-
ed OR = 1.53, 95% CI (1.29; 1.83)). Adjusting for potential
confounders including symptoms of depression in gestational
week 17 did not influence our results (adjusted OR = 1.49,
95% CI (1.23; 1.79)).
Symptoms of depression at the 17th gestational week was
independently associated with HG in the multivariate model
(OR = 1.71, 95% CI (1.31; 2.23)). As shown on Table 2,other
factors positively associated with HG included short educa-
tion, female sex of the child, multiparity, younger age of the
mother and pelvic girdle pain. Pre-pregnancy BMI did not
differ between women with and without HG, and smoking
was negatively associated with HG.
We also assessed whether women with a history of depres-
sion were more likely to be hospitalised during pregnancy in
general. Among women with previous depression, 7.7% were
hospitalised during pregnancy compared to 5.2% without;
History of depression and risk of hyperemesis gravidarum 399
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

history of depression was associated with higher odds for
hospitalisation (OR = 1.52, 95% CI (1.42; 1.62)).
Although HG was positively associated with depression,
the majority of women with HG (66%, 489/740) neither had
a lifetime history of depression nor symptoms of depression in
the 17th gestational week as shown in Fig. 1. Furthermore,
only 1.2% of women with previous depression developed HG.
Discussion
The main finding of the present study was that having a life-
time history of depression was associated with 50% higher
odds for HG. The majority of women with HG did not, how-
ever, have a history of depression, and less than 2% of women
with previous depression developed HG.
The results are in line with previous research. Using health
insurance data from the Midwestern USA between 2000 and
2004, Seng et al. (2007) found that a diagnosis of depression
before pregnancy was positively associated with HG in a pop-
ulation of 11,016 women, including 208 HG pregnancies
(OR = 3.2, 95% CI (2.0; 5.2)). Additionally, they found that
the burden of illness increased the likelihood of HG. Having
had a psychiatric or somatic condition before pregnancy in-
creased the odds for HG twofold, while having had both a
psychiatric and somatic condition increased the odds fourfold.
The study design permitted the identification of psychiatric
diagnoses occurring before pregnancy, but information about
other co-variates was limited.
Another large cohort study comprising 157,922 women, of
whom 1301 had HG, was extracted from a population-based
healthcare database covering all deliveries to residents of Nova
Scotia, Canada, between 1988 and 2002 (Fell et al. 2006). The
Tabl e 1 Characteristics of the sample according to HG status among
81,786 women
HG n(%) No HG n(%) Total n(%)
History of depression
No 520 (71.1) 63,915 (78.9) 64,435 (78.8)
Yes 211 (28.9) 17,140 (21.1) 17,351 (21.2)
Symptoms of current depression
Low score 650 (88.9) 76,155 (94.0) 76,805 (93.9)
High score 81 (11.1) 4900 (6.0) 4981 (6.1)
Parity
Primipara 287 (39.3) 36,480 (45.0) 36,767 (45.0)
Multipara 444 (60.7) 44,575 (55.0) 45,019 (55.0)
Length of education (years)
<12 79 (10.8) 5599 (6.9) 5678 (6.9)
13–16 536 (73.3) 56,034 (69.1) 56,570 (69.2)
>16 116 (15.9) 19,422 (24.0) 19,538 (23.9)
Smoking
No 495 (79.5) 49,153 (69.3) 49,648 (69.4)
Yes 128 (20.5) 21,811 (30.7) 21,939 (30.6)
Sex of the child
Boy 307 (42.0) 41,571 (51.3) 41,878 (51.2)
Girl 424 (58.0) 39,484 (48.7) 39,908 (48.8)
Pelvic girdle pain
No 583 (79.8) 69,145 (85.3) 69,728 (85.3)
Yes 148 (20.2) 11,910(14.7) 12,058 (14.7)
HG mean (SD) No HG mean (SD) Total mean (SD)
Maternal age 29.3 (4.9) 30.3 (4.5) 30.3 (4.5)
Pre-pregnancy BMI 24.5 (4.2) 24.1 (4.3) 24.1 (4.3)
Tabl e 2 Unadjusted and adjusted
odds ratios (OR) with 95% confi-
dence intervals (CI) for
hyperemesis gravidarum
(n= 611, 0.9%) among 69,864
pregnancies
Unadjusted OR
(95% CI)
Pvalue Adjusted OR
(95% CI)
Pvalue
History of depression
No 1 –1–
Yes 1.53 (1.29; 1.83) < 0.001 1.49 (1.23; 1.79) <0.001
Symptoms of current depression
Low score 1 –1–
High score 2.11 (1.65; 2.69) < 0.001 1.71 (1.31; 2.23) <0.001
Maternal age 0.94 (0.93; 0.96) < 0.001 0.93 (0.91; 0.95) <0.001
Parity
Primipara 1 –1–
Multipara 1.24 (1.05; 1.45) 0.010 1.43 (1.20; 1.69) <0.001
Length of education (years)
<12 2.42 (1.76; 3.32) < 0.001 1.91 (1.36; 2.69) <0.001
13–16 1.64 (1.31; 2.06) < 0.001 1.44 (1.13; 1.82) 0.003
>16 1 –1–
Pre-pregnancy BMI 1.03 (1.01; 1.04) 0.002 1.02 (1.00; 1.04) 0.030
Smoking
No 1 –1–
Yes 0.60 (0.49; 0.72) <0.001 0.46 (0.37; 0.56) <0.001
Sex of the child
Boy 1 –1–
Girl 1.49 (1.27; 1.75) <0.001 1.50 (1.28; 1.76) <0.001
Pelvic girdle pain
No 1 –1–
Yes 1.52 (1.25; 1.85) <0.001 1.30 (1.06; 1.59) 0.011
400 Kjeldgaard H.K. et al.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

study revealed a fourfold higher risk of HG in women with
psychiatric disease (RR = 4.1, 95% CI (3.0; 5.7)). The timing
or types of psychiatric disease were not specified, but a crude
RR of 2.5 with 95% CI (1.5; 4.2) was reported for HG in women
with depression compared to women without depression.
On the other hand, Magtira et al. (2014) found no statisti-
cally significant differences in the prevalence of psychiatric
conditions prior to the first pregnancy when comparing 84
women with recurrence of HG with 34 women with no recur-
rence. The authors predicted that if psychiatric symptoms pos-
itively correlate with HG, then psychiatric symptoms would
correlate positively with recurrence risk. As the study was
based on data from an online survey, the participating women
were not randomly selected among women with HG and may
therefore have had a different risk profile from the women
who did not participate (Bornehag et al. 2012). Additionally,
only women who had had at least two pregnancies lasting
beyond the second trimester were included, which may intro-
duce recall bias, e.g. whether psychiatric symptoms preceded
pregnancy, or selection biasas women with poor psychosocial
health may have been less likely to continue participation, as
were women terminating their pregnancies due to HG
(McDonald et al. 2013).
Given the nature of the MoBa data, we were able to explore
whether symptoms of depression in the current pregnancy
were independently associated with HG. Consistent with a
recent meta-analysis (Mitchell-Jones et al. 2016), we found
an association between HG and depression during pregnancy.
Two prospective studies, both excluding women with a histo-
ry of psychiatric disease, also reported that women with HG
were more likely to suffer from symptoms of anxiety and
depression during pregnancy compared to asymptomatic
pregnant women (Aksoy et al. 2015; Pirimoglu et al. 2010).
It was therefore argued that psychological distress was a con-
sequence of HG rather than the cause (Aksoy et al. 2015).
Since hospitalisation due to prolonged NVP was a require-
ment for having HG in the current study, our results may have
been biased by a greater likelihood of being hospitalised
among women with a history of depression (Atanackovic
et al. 2001). A relationship between depression and severity
of NVP has previously been suggested (Kelly et al. 2001;
Mazzotta et al. 2000) although other studies do not support
this finding (Swallow et al. 2004; Tan et al. 2010). We there-
fore assessed whether women with previous depression were
more likely to be hospitalised in general during pregnancy.
Previous depression was associated with hospitalisation
(OR = 1.52, 95% CI (1.42; 1.62)), which may have contrib-
uted to overestimating the effect of previous depression on the
risk of HG. However, in Norway, only women with severe
symptoms of HG, including metabolic disturbances, are
hospitalised. Additionally, there is no tradition for outpatient
treatment for these patients. This indicates that our sample is
restricted to severe HG cases corresponding to ICD 10 code
O21.1, and it is therefore unlikely that the women have been
hospitalised due to depression. Given that hospital care in
Norway is free of charge, it is furthermore unlikely that more
socially disadvantaged women are less likely to be
hospitalised.
Several studies show that a variety of somatic diseases such
as pelvic girdle pain, H. pylori infection, thyroid disease, gas-
trointestinal disorders, rheumatoid arthritis, pre-eclampsia,
chronic hypertension, type 1 diabetes, asthma and eating disor-
ders are associated with higher risk of HG (Bolin et al. 2013;
Fell et al. 2006; Jorgensen et al. 2012;Lietal.2015;Sengetal.
2007; Torgersen et al. 2008). However, in the present study, the
number of HG cases with these conditions was too small to
Fig. 1 The total number of
women with aHG (n=731),ba
historyofdepression(n=17,351)
and csymptoms of current
depression (n = 4981) among
81,786 women in the Norwegian
Mother and Child Cohort Study.
The number of women with HG
and a history of depression (ab,
n= 211); with HG and symptoms
of current depression (ac,n=81)
and with HG, a history of
depression and symptoms of
current depression (abc,n=50).
The number of women with no
HG and a history of depression
and symptoms of current
depression (bc,n= 2910)
History of depression and risk of hyperemesis gravidarum 401
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

explore possible influences of these conditions. Our results
should be interpreted with these limitations in mind.
HG is a diagnosis by exclusion and an international con-
sensus on the definition of HG is yet to be established, limiting
comparison of previous research (Mitchell-Jones et al. 2016).
The lack of consensus is a challenge for clinicians who may
need to distinguish milder forms of HG from more common
nausea and vomiting in pregnancy (Grooten et al. 2015).
Inadequate care of women with HG may have severe conse-
quences including therapeutic abortions, Wernicke’senceph-
alopathy and even death (Eliakim et al. 2000; Poursharif et al.
2007). Adverse pregnancy outcomes such as low birth weight
and preterm delivery may in particular affect HG women with
poor pregnancy weight gain (<7 kg) (Dodds et al. 2006).
Adequate care of women with HG is thus of the utmost
importance.
To our knowledge, this is the first time a large, high-quality
data set enables the study of the associations between a history
of depression and HG and symptoms of depression during
pregnancy and HG. Our results advocate that routine psychi-
atric consultations of HG women may be unnecessary.
Treatment should focus on relief of somatic complaints and
ensure the health of the mother and child.
The large number of HG pregnancies is a major
strength of the current study. Furthermore, the study cov-
ered all regions of Norway, and the prospective nature of
data collection minimises the risk of recall bias. To date,
more than 400 articles have been published based on
MoBa data. Around 40% of the invited women participat-
ed in the study, introducing a possibility of self-selection
bias. However, a recent study looking into potential bias
by skewed selection of participants in MoBa found that
the participant selection influenced the prevalence esti-
mates but not the exposure outcome associations (Nilsen
et al. 2009). Women known to be underrepresented in MoBa
include single women, those with shorter education, those
under 25 years of age, immigrants and smokers (Nilsen et al.
2009; Vikanes et al. 2010). Hospitalisation for HG was
assessed retrospectively; however, recall bias is highly unlike-
ly due to the relatively short interval between hospitalisation
and reporting of HG in week 32 of pregnancy (Vikanes et al.
2010). The comparison group comprised all other pregnant
women in the study, including those with complications other
than HG, reducing the risk of overestimating the association
between previous depression and HG.
The KLTDS and SCL-5 are the only available mea-
suresofmentalhealthintheMoBastudy.Unlikeclin-
ical interviews, the KLTDS and SCL-5 cannot be used
to diagnose depression. The scales have, however, been
developed and validated to measure symptoms of de-
pression in population studies. Extensive questionnaire
studies with a broad scope such as the MoBa study
often have a shortage of space for the original lengthy
psychometric instruments, and short versions may be
useful to improve response rates. While the short ver-
sions affect the measurement precision, the precision
remains sufficient for epidemiological purposes (Strand
et al. 2003; Tambs and Moum 1993;Tambsand
Røysamb 2014).
The fact that a history of depression was not measured
before pregnancy is a limitation of our study. Women
responded to the KLTDS in gestational week 17, which
for most women with HG is after the onset of severe
nausea and vomiting. This may have affected their re-
sponse. In our analyses, we therefore adjusted for symp-
toms of current depression at the 17th gestational week
to quantify the direct effect of a previous depression on
HG. The effect estimates changed only slightly in the
adjusted model indicating that KLTDS and SCL-5 cover
different aspects of women’s mental health in relation to
HG.
Conclusion
In conclusion, a lifetime history of depression increased the
odds for hospitalisation for HG by approximately 50%.
However, two thirds of women with HG had neither a history
of depression nor symptoms of depression at the 17th gesta-
tional week. Given the fact that only 1.2% of women with
previous depression developed HG, depression does not ap-
pear to be a main driver in the aetiology and pathogenesis of
HG. Our results advocate that routine psychiatric consulta-
tions may be unnecessary.
Acknowledgements We are grateful to all of the women and their
families for participating in this continuing cohort study.
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
interest.
Funding The Norwegian Mother and Child Cohort Study is supported
by the Norwegian Ministry of Health and the Ministry of Education and
Research,NIH/NIEHS (contract no. N01-ES-75558), NIH/NINDS (grant
no. 1 UO1 NS 047537–01 and grant no. 2 UO1 NS 047537-06A1). The
present study was supported by the South-Eastern Norway Regional
Health Authority (grant no. 2014003). The funding sources had no role
in the conduct of the study.
Open Access This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided you give appro-
priate credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made.
402 Kjeldgaard H.K. et al.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

References
Aksoy H, Aksoy U, Karadag OI, Hacimusalar Y, Acmaz G, Aykut G,
Cagli F, Yucel B, Aydin T, Babayigit MA (2015) Depression levels
in patients with hyperemesis gravidarum: a prospective case-control
study. SpringerPlus 4:34. doi:10.1186/s40064-015-0820-2
Atanackovic G, Wolpin J, Koren G (2001) Determinants of the need for
hospital care among women with nausea and vomiting of pregnancy.
Clin Invest Med 24:90–93
Bjelland EK, Stuge B, Engdahl B, Eberhard-Gran M (2013) The effect of
emotional distress on persistent pelvic girdle pain after delivery: a
longitudinal population study. BJOG 120:32–40. doi:10.1111/1471-
0528.12029
Bolin M, Akerud H, Cnattingius S, Stephansson O, Wikstrom AK (2013)
Hyperemesis gravidarum and risks of placental dysfunction disor-
ders: a population-based cohort study. BJOG 120:541–547.
doi:10.1111/1471-0528.12132
Bornehag CG, Moniruzzaman S, Larsson M, Lindstrom CB, Hasselgren
M, Bodin A, von Kobyletzkic LB, Carlstedt F, Lundin F, Nanberg E,
Jonsson BA, Sigsgaard T, Janson S (2012) The SELMA study: a
birth cohort study in Sweden following more than 2000 mother-
child pairs. Paediatr Perinat Epidemiol 26:456–467. doi:10.1111
/j.1365-3016.2012.01314.x
Chortatos A, Haugen M, Iversen PO, Vikanes A, Eberhard-Gran M,
Bjelland EK, Magnus P, Veierod MB (2015) Pregnancy complica-
tions and birth outcomes among women experiencing nausea only
or nausea and vomiting during pregnancy in the Norwegian Mother
and Child Cohort Study. BMC Pregnancy Childbirth 15:138.
doi:10.1186/s12884-015-0580-6
Corey LA, Berg K, Solaas MH, Nance WE (1992) The epidemiology of
pregnancy complications and outcome in a Norwegian twin popu-
lation. Obstet Gynecol 80:989–994
Dahl FA, Grotle M, Saltyte Benth J, Natvig B (2008) Data splitting as a
countermeasure against hypothesis fishing: with a case study of
predictors for low back pain. Eur J Epidemiol 23:237–242.
doi:10.1007/s10654-008-9230-x
DoddsL,FellDB,JosephKS,AllenVM,ButlerB(2006)Outcomesof
pregnancies complicated by hyperemesis gravidarum. Obstet Gynecol
107:285–292. doi:10.1097/01.AOG.0000195060.22832.cd
Dorheim SK, Bjorvatn B, Eberhard-Gran M (2013) Sick leave during
pregnancy: a longitudinal study of rates and risk factors in a
Norwegian population. BJOG 120:521–530. doi:10.1111/1471-
0528.12035
Eliakim R, Abulafia O, Sherer DM (2000) Hyperemesis gravidarum: a
current review. Am J Perinatol 17:207–218. doi:10.1055/s-2000-
9424
Fairweather DV (1968) Nausea and vomiting in pregnancy. Am J Obstet
Gynecol 102:135–175
Fejzo MS, Ingles SA, Wilson M, Wang W, MacGibbon K, Romero R,
Goodwin TM (2008) High prevalence of severe nausea and
vomiting of pregnancy and hyperemesis gravidarum among rela-
tives of affected individuals. Eur J Obstet Gynecol Reprod Biol
141:13–17. doi:10.1016/j.ejogrb.2008.07.003
Fell DB, Dodds L, Joseph KS, Allen VM, Butler B (2006) Risk factors
for hyperemesis gravidarum requiring hospital admission during
pregnancy. Obstet Gynecol 107:277–284. doi:10.1097/01.
aog.0000195059.82029.74
Gadsby R, Barnie-Adshead AM, Jagger C (1993) A prospective study of
nausea and vomiting during pregnancy. Br J Gen Pract 43:245–248
Gazmararian JA, Petersen R, Jamieson DJ, Schild L, Adams MM,
Deshpande AD, Franks AL (2002) Hospitalizations during
pregnancy among managed care enrolees. Obstet Gynecol
100:94–100
Grooten IJ, Roseboom TJ, Painter RC (2015) Barriers and challenges in
hyperemesis gravidarum research. Nutrition and metabolic insights
8:33–39. doi:10.4137/nmi.s29523
Heitmann K, Solheimsnes A, Havnen GC, Nordeng H, Holst L (2016)
Treatment of nausea and vomiting during pregnancy—across-
sectional study among 712 Norwegian women. Eur J Clin
Pharmacol. doi:10.1007/s00228-016-2012-6
Irgens LM (2000) The Medical Birth Registry of Norway.
Epidemiological research and surveillance throughout 30 years.
Acta Obstet Gynecol Scand 79:435–439
Jorgensen KT, Nielsen NM, Pedersen BV, Jacobsen S, Frisch M (2012)
Hyperemesis, gestational hypertensive disorders, pregnancy losses
and risk of autoimmune diseases in a Danish population-based co-
hort. J Autoimmun 38:J120–J128. doi:10.1016/j.jaut.2011.10.002
Kelly RH, Russo J, Katon W (2001) Somatic complaints among pregnant
women cared for in obstetrics: normal pregnancy or depressive and
anxiety symptom amplification revisited? Gen Hosp Psychiatry 23:
107–113
Kendler KS, Neale MC, Kessler RC, Heath AC, Eaves LJ (1993) The
lifetime history of major depression in women. Reliability of diag-
nosis and heritability. Arch Gen Psychiatry 50:863–870
Kim DR, Connolly KR, Cristancho P, Zappone M, Weinrieb RM (2009)
Psychiatric consultation of patients with hyperemesis gravidarum.
Arch Womens Ment Health 12:61–67. doi:10.1007/s00737-009-
0064-7
Li L, Li L, Zhou X, Xiao S, Gu H, Zhang G (2015) Helicobacter pylori
infection is associated with an increased risk of hyperemesis
gravidarum: a meta-analysis. Gastroenterol Res Pract 2015:
278905. doi:10.1155/2015/278905
Magnus P, Irgens LM, Haug K, Nystad W, Skjaerven R, Stoltenberg C
(2006) Cohort profile: the Norwegian Mother and Child Cohort
Study (MoBa). Int J Epidemiol 35:1146–1150. doi:10.1093
/ije/dyl170
Magnus P, Birke C, Vejrup K, Haugan A, Alsaker E, Daltveit AK, Handal
M, Haugen M, Hoiseth G, Knudsen GP, Paltiel L, Schreuder P,
Tambs K, Vold L, Stoltenberg C (2016) Cohort profile update: the
Norwegian Mother and Child Cohort Study (MoBa). Int J
Epidemiol 45:382–388. doi:10.1093/ije/dyw029
Magtira A, Paik Schoenberg F, MacGibbon K, Tabsh K, Fejzo MS (2014)
Psychiatric factors do not affect recurrence risk of hyperemesis
gravidarum. J Obstet Gynaecol Res. doi:10.1111/jog.12592
Mazzotta P, Stewart D, Atanackovic G, Koren G, Magee LA (2000)
Psychosocial morbidity among women with nausea and vomiting
of pregnancy: prevalence and association with anti-emetic therapy. J
Psychosom Obstet Gynaecol 21:129–136
McDonald SW, Lyon AW, Benzies KM, McNeil DA, Lye SJ, Dolan SM,
Pennell CE, Bocking AD, Tough SC (2013) The All Our Babies
pregnancy cohort: design, methods, and participant characteristics.
BMC Pregnancy Childbirth 13(Suppl 1):S2. doi:10.1186/1471-
2393-13-s1-s2
Mitchell-Jones N, Gallos I, Farren J, Tobias A, Bottomley C, Bourne T
(2016) Psychological morbidity associated with hyperemesis
gravidarum; a systematic review and meta-analysis. BJOG doi.
doi:10.1111/1471-0528.14180
Nilsen RM, Vollset SE, Gjessing HK, Skjaerven R, MelveKK, Schreuder
P, Alsaker ER, Haug K, Daltveit AK, Magnus P (2009) Self-
selection and bias in a large prospective pregnancy cohort in
Norway. Paediatr Perinat Epidemiol 23:597–608. doi:10.1111
/j.1365-3016.2009.01062.x
Pirimoglu ZM, Guzelmeric K, Alpay B, Balcik O, Unal O, Turan MC
(2010) Psychological factors of hyperemesis gravidarum by using
the SCL-90-R questionnaire. Clin Exp Obstet Gynecol 37:56–59
Poursharif B, Korst LM, Macgibbon KW, Fejzo MS, Romero R,
Goodwin TM (2007) Elective pregnancy termination in a large
History of depression and risk of hyperemesis gravidarum 403
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

cohort of women with hyperemesis gravidarum. Contraception 76:
451–455. doi:10.1016/j.contraception.2007.08.009
Poursharif B, Korst LM, Fejzo MS, MacGibbon KW, Romero R,
Goodwin TM (2008) The psychosocial burden of hyperemesis
gravidarum. J Perinatol 28:176–181. doi:10.1038/sj.jp.7211906
Rashid M, Rashid MH, Malik F, Herath RP (2012) Hyperemesis
gravidarum and fetal gender: a retrospective study. J Obstet
Gynaecol 32:475–478. doi:10.3109/01443615.2012.666580
Robins LN, Wing J, Wittchen HU, Helzer JE, Babor TF, Burke J,
Farmer A, Jablenski A, Pickens R, Regier DA et al (1988)
The Composite International Diagnostic Interview. An epide-
miologic instrument suitable for use in conjunction with dif-
ferent diagnostic systems and in different cultures. Arch Gen
Psychiatry 45:1069–1077
Sandanger I, Moum T, Ingebrigtsen G, Dalgard OS, Sorensen T,
Bruusgaard D (1998) Concordance between symptom screening
and diagnostic procedure: the Hopkins Symptom Checklist-25 and
the Composite International Diagnostic Interview I. Soc Psychiatry
Psychiatr Epidemiol 33:345–354
Seng JS, Schrot JA, van De Ven C, Liberzon I (2007) Service use data
analysis of pre-pregnancy psychiatric and somatic diagnoses in
women with hyperemesis gravidarum. J Psychosom Obstet
Gynaecol 28:209–217. doi:10.1080/01674820701262044
Strand BH, Dalgard OS, Tambs K, Rognerud M (2003) Measuring the
mental health status of the Norwegian population: a comparison of
the instruments SCL-25, SCL-10, SCL-5 and MHI-5 (SF-36). Nord
J Psychiatry 57:113–118. doi:10.1080/08039480310000932
Swallow BL, Lindow SW, Masson EA, Hay DM (2004)
Psychological health in early pregnancy: relationship with
nausea and vomiting. J Obstet Gynaecol 24:28–32.
doi:10.1080/01443610310001620251
Tambs K, Moum T (1993) How well can a few questionnaire items
indicate anxiety and depression? Acta Psychiatr Scand 87:364–367
Tambs K, Røysamb E (2014) Selection of questions to short-form ver-
sions of original psychometric instruments in MoBa. Norsk
epidemiologi 24:195–201
Tan PC, Vani S, Lim BK, Omar SZ (2010) Anxiety and depression in
hyperemesis gravidarum: prevalence, risk factors and correlation
with clinical severity. Eur J Obstet Gynecol Reprod Biol 149:153–
158. doi:10.1016/j.ejogrb.2009.12.031
Torgersen L, Von Holle A, Reichborn-Kjennerud T, Berg CK, Hamer R,
Sullivan P, Bulik CM (2008) Nausea and vomiting of pregnancy in
women with bulimia nervosa and eating disorders not otherwise
specified. Int J Eat Disord 41:722–727. doi:10.1002/eat.20564
Verberg MF, Gillott DJ, Al-Fardan N, Grudzinskas JG (2005)
Hyperemesis gravidarum, a literature review. Hum Reprod Update
11:5 27–539. doi:10.1093/humupd/dmi021
Vikanes A, Grjibovski AM, Vangen S, Magnus P (2008) Variations in
prevalence of hyperemesis gravidarum by country of birth: a study
of 900,074 pregnancies in Norway, 1967-2005. Scand J Public
Health 36:135–142. doi:10.1177/1403494807085189
Vikanes A, Grjibovski AM, Vangen S, Gunnes N, Samuelsen SO,
Magnus P (2010) Maternal body composition, smoking, and
hyperemesis gravidarum. Ann Epidemiol 20:592–598.
doi:10.1016/j.annepidem.2010.05.009
Vikanes AV, Stoer NC, Magnus P, Grjibovski AM (2013) Hyperemesis
gravidarum and pregnancy outcomes in the Norwegian Mother and
Child Cohort—a cohort study. BMC Pregnancy Childbirth 13:169.
doi:10.1186/1471-2393-13-169
World Health Organization (2004) ICD-10, Chapter XV, Pregnancy,
childbirth and the puerperium (O00-O99), Other maternal disorders
predominantly related to pregnancy (O20-O29). World Health
Organization. http://apps.who.int/classifications/apps/icd/icd10
online2004/fr-icd.htm?go20.htm+
404 Kjeldgaard H.K. et al.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com