Prevalence, Course, and Risk Factors for
Antenatal Anxiety and Depression
Antoinette M. Lee, PhD, Siu Keung Lam, MD, Stephanie Marie Sze Mun Lau, BsocSc,
Catherine Shiu Yin Chong, MBBS, Hang Wai Chui, MPH, and Daniel Yee Tak Fong,
OBJECTIVE: To estimate the prevalence and course of
antenatal anxiety and depression across different stages
Downloaded from http://journals.lww.com/greenjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 07/08/2021
of pregnancy, risk factors at each stage, and the relation-
ship between antenatal anxiety and depression and post-
partum depression.
METHODS: A consecutive sample of 357 pregnant women
in an antenatal clinic in a regional hospital was assessed
longitudinally at four stages of pregnancy: first trimester,
second trimester, third trimester, and 6 weeks postpartum.
The antenatal questionnaire assessed anxiety and depres-
sion (using the Hospital Anxiety and Depression Scale) and
demographic and psychosocial risk factors. The postpartum
questionnaire assessed postpartum depression with the
Edinburgh Postnatal Depression Scale.
RESULTS: More than one half (54%) and more than one
third (37.1%) of the women had antenatal anxiety and
depressive symptoms, respectively, in at least one ante-
natal assessment. Anxiety was more prevalent than de-
pression at all stages. A mixed-effects model showed that
both conditions had a nonlinear changing course (P<.05
for both), with both being more prevalent and severe in
the first and third trimesters. Risk factors were slightly
different at different stages. Both antenatal anxiety (ad-
justed odds ratio [OR] 2.66, Pⴝ.004 in the first trimester;
adjusted OR 3.65, P<.001 in the second trimester; ad-
justed OR 3.84, P<.001 in the third trimester) and de-
pression (adjusted OR 4.16, P<.001 in the first trimester;
adjusted OR 3.35, Pⴝ.001 in the second trimester; ad-
From the Departments of Psychiatry and Nursing Studies, LKS Faculty of
Medicine, the University of Hong Kong, Hong Kong; Department of Obstetrics
and Gynaecology, Kwong Wah Hospital, Hong Kong; and Pamela Youde
Nethersole Eastern Hospital, Hong Kong.
This study was supported by a grant (Seed Funding for New Staff) from the
University of Hong Kong (No. 10204464) to Antoinette M. Lee.
Corresponding author: Dr. Antoinette M. Lee, Department of Psychiatry, the
University of Hong Kong, 2/F New Clinical Building, Queen Mary Hospital,
Pokfulam, Hong Kong; e-mail: amlee@hkucc.hku.hk.
Financial Disclosure
The authors have no potential conflicts of interest to disclose.
© 2007 by The American College of Obstetricians and Gynecologists. Published
by Lippincott Williams & Wilkins.
ISSN: 0029-7844/07
1102 VOL. 110, NO. 5, NOVEMBER 2007
PhD
justed OR 2.67, Pⴝ.009 in the third trimester) increased
the risk of postpartum depression.
CONCLUSION: Antenatal anxiety and depression are
prevalent and serious problems with changing courses.
Continuous assessment over the course of pregnancy is
warranted. Identifying and treating these problems is
important in preventing postpartum depression.
(Obstet Gynecol 2007;110:1102–12)
LEVEL OF EVIDENCE: II
P ostpartum depression received burgeoning atten-
tion in the past decades. In contrast, studies on
antenatal psychological morbidity are only beginning
to flourish. This might partially be due to the miscon-
ception that women were hormonally protected from
psychological disturbances during pregnancy.1 Iden-
tification of patients with antenatal anxiety and de-
pression is further made difficult by the similarity
between somatic symptoms of anxiety and depressive
disorders and somatic complaints commonly found in
the normal course of pregnancy, such as fatigue and
appetite change.2 However, recent studies revealed
that rates of anxiety and depressive symptoms were
actually higher during pregnancy than in the postpar-
tum period,3,4 highlighting the significance of antena-
tal mental health problems.
It has been reported that 7–20% of pregnant
women suffer from antenatal depression.4,5–7 Data on
the prevalence of antenatal anxiety is more limited,
although a study of second-trimester pregnant women
found that 6.6% had antenatal anxiety disorders.5 The
same study identified 14.1% of pregnant women
having one or more psychiatric disorders but reported
that only 5.5% received some form of treatment,
showing that antenatal psychiatric problems are
largely underdiagnosed and undertreated. This is a
serious issue because mental health problems during
pregnancy are known to exert adverse influences on
both women and their offspring.5,8 –10 Antenatal anxi-
OBSTETRICS & GYNECOLOGY
ety and depression are also powerful predictors of
postpartum depression,3,11 although it is not clear at
which point during pregnancy these psychological
states are most predictive of postpartum depression.
Indeed, antenatal psychological states appeared to be
dynamic and changing in nature, with most studies
demonstrating a generally increasing trend of depres-
sive symptoms during pregnancy followed by a de-
cline after childbirth.4,12 Unfortunately, good data on
prevalence and course of other mental health prob-
lems during pregnancy are inadequate, particularly
anxiety problems.
Finally, although information on risk factors for
antenatal anxiety and depression is available in the
literature,13,14 most studies focused only on one stage
of pregnancy. A longitudinal analysis of risk factors
across different stages of pregnancy is of value for
examining how the effect of various risk factors
changes across pregnancy. This has implications for
developing effective prevention and early interven-
tion strategies.
Our study aimed at estimating the prevalence and
course of antenatal anxiety and depression across
different stages of pregnancy and identifying demo-
graphic and psychological risk factors for antenatal
anxiety and depression at each stage. We further
examined whether antenatal anxiety and depression
were associated with increased risk of postpartum
depression.
MATERIALS AND METHODS
The study was approved by the institutional review
boards of both the University of Hong Kong and the
Kwong Wah Hospital. Written informed consent was
sought from all participants. Based on power calcula-
tions,15 for a power of 80% for detecting a medium
effect size of 0.5015 at a 5% level of significance and an
attrition rate of 15%, a total of 345 pregnant women
were needed. Four hundred twenty-three consecutive
Chinese pregnant women at first presentation in the
antenatal clinic of a regional hospital in Hong Kong
were approached for participation in the study. The
regional hospital serves a population of half a million
persons of diverse socioeconomic background. Eligi-
ble women included all pregnant women of Chinese
ethnicity above 18 years of age. Women having
significant medical diseases, considering termination
of pregnancy, or having conceived through in vitro
fertilization were excluded.
A prospective longitudinal design was used. Par-
ticipating women were assessed a total of four times.
They were first assessed in the first trimester and were
reassessed at the second and third trimesters and at 6
VOL. 110, NO. 5, NOVEMBER 2007
weeks postpartum. The first three assessment ques-
tionnaires were completed in the clinic, whereas the
6-week postpartum questionnaire was mailed to the
participants, who were asked to send them back in a
self-addressed and stamped envelope.
A number of questionnaires were administered at
each assessment time point to assess antenatal anxi-
ety, antenatal depression, postpartum depression, and
risk factors. The Hospital Anxiety and Depression
Scale16 was used to assess antenatal anxiety and
depression. It is a 14-item self-report instrument with
two subscales providing separate measures of anxiety
and depression. Higher scores indicate higher levels
of anxiety and depression. The validated Chinese
version17 was used. The suggested cutoff of 7/8 was
used for each subscale to identify probable cases of
clinically significant anxiety or depression.16 The Hos-
pital Anxiety and Depression Scale was chosen be-
cause it was specifically developed to measure anxiety
and depression among medical patients by focusing
on affective symptoms and excluding somatic symp-
toms such as dizziness, insomnia, and fatigue that are
also related to physical disorders. As such, it is
suitable for use among pregnant women and is supe-
rior to other instruments that may inflate the rates of
anxiety and depression because many somatic symp-
toms are common experiences during pregnancy
rather than a reflection of psychological disturbances.
The Hospital Anxiety and Depression Scale has been
validated among pregnant women with a sensitivity
and specificity of 93% and 90%, respectively, for the
anxiety subscale and 90% and 91%, respectively, for
the depression subscale.18 The sensitivity and speci-
ficity of the Edinburgh Postnatal Depression Scale in
identifying antenatal depression was only 64% and
90%, respectively.19
Postpartum depression was assessed with the
Edinburgh Postnatal Depression Scale, a commonly
used 10-item measure of depression in the postpartum
period.20 The Chinese version has been validated
among pregnant women in Hong Kong, with good
psychometric properties.21 The recommended cutoff
of 12/13 was used for screening probable cases of
postpartum depression.20 This cutoff yields a sensitiv-
ity of 88% and a specificity of 93%.19
Risk factors examined included demographic and
psychosocial risk factors. Information on age, marital
status, parity, family income, past and current smok-
ing behavior, and past and current alcohol use were
sought.
Attitudes toward pregnancy was measured by
two items: whether the pregnancy was planned (yes/
no) and whether the pregnancy was wanted (yes/no).
Lee et al Antenatal Anxiety and Depression 1103
The Rosenberg Self Esteem Scale22 was used as a
measure of global self-esteem. Perceived social sup-
port was measured by a 9-item self-constructed scale.
It measures perceived support from significant others
including husband, parents, parents-in-law, siblings,
children, other relatives, friends, bosses, and col-
leagues on a 5-point Likert scale, from 1 (very unsat-
isfactory) to 5 (very satisfactory). The mean score was
used as an index of perceived social support. Finally,
respondents were asked to rate the quality of their
marital relationships on a 5-point Likert scale, from 1
(very poor) to 5 (very good).
Information on demographic risk factors was
obtained only in the first assessment (first trimester),
whereas the Edinburgh Postnatal Depression Scale
was included only in the last assessment (6 weeks
postpartum). All the other instruments were adminis-
tered in all assessments.
Statistical analysis was performed with SPSS 13.0
(SPSS Inc, Chicago, IL) and R 2.3.1 (The R Develop-

Table 1. Sample Characteristics of Women at First Trimester With Comparison Between (A) Women
Who Did Not Complete All the Antenatal Assessments and Women Who Completed All
Antenatal Assessments and Between (B) Women Who Completed All Antenatal and Postpartum
Assessments and Women Who Did Not Complete the Postpartum Assessment
(A) (B)

Women Who Women Who Women Who

Did Not Women Who Completed All Did Not
Complete All Completed Antenatal and Complete the
Antenatal All Antenatal Postpartum Postpartum
Assessments Assessments Assessments Assessment
(nⴝ22) (nⴝ335) (nⴝ244) (nⴝ91)

Characteristics n % n % P n % n % P
Age (y) .415 .746
25 or less 3 13.6 38 11.3 26 10.7 12 13.2
26–34 12 54.5 222 66.3 164 67.2 58 63.7
35 or more 7 31.8 75 22.4 54 22.1 21 23.1
Marital status (2) 1.000 (1) (1) .246
Married/cohabiting 21 95.5 318 95.5 234 96.3 84 93.3
Single/divorced 1 4.5 15 4.5 9 3.7 6 6.7
Education level (3) .044 (2) (1) .362
Primary 1 4.5 6 1.8 3 1.2 3 3.3
Secondary 12 54.5 254 76.5 188 77.7 66 73.3
Tertiary or above 9 40.9 72 21.7 51 21.1 21 23.3
Family income (monthly)* (1) (43) .167 (27) (16) .656
Less than 20,000 8 38.1 138 47.3 101 46.5 37 49.3
20,000–30,000 3 14.3 72 24.7 52 24 20 26.7
More than 30,000 10 47.6 82 28.1 64 29.5 18 24.0
Parity (2) 1.000 (1) (1) .378
Primigravida 13 59.1 200 60.1 142 58.4 58 64.4
Planned pregnancy (4) 1.000 (3) (1) .053
Yes 17 77.3 257 77.6 194 80.5 63 70
Wanted pregnancy (3) 1.000 (2) (1) .142
Yes 22 100 322 97 237 97.9 85 94.4
History of smoking (1) (4) .548 (3) (1) .410
Yes 2 9.5 56 16.9 38 15.8 18 20
History of drinking (5) 1.000 (4) (1) 1.000
Yes 4 18.2 70 21.2 51 21.3 19 21.1
Smoking (2) (9) 1.000 (7) (2) .353
Yes 0 0 13 4.0 8 3.4 5 5.6
Drinking (2) (22) 1.000 (18) (4) 1.000
Yes 0 0 13 4.2 10 4.4 3 3.4
Past psychiatric disorders (3) .062 (2) (1) 1.000
Yes 1 4.5 0 0 0 0 0 0
Numbers in parentheses refer to number of missing values.
* In Hong Kong dollars (1 U.S. dollar equals approximately 7.8 Hong Kong dollars).

1104 Lee et al Antenatal Anxiety and Depression OBSTETRICS & GYNECOLOGY

ment Core Team, Auckland). Descriptive statistics
were used to summarize the demographic character-
istics of the sample. Attrition analyses comparing
those who participated in all time points of the study
and those who dropped out of the study were per-
formed with Fisher exact test for categorical variables,
t tests for continuous variables when the normality
assumption was satisfied, and Mann-Whitney U test
for continuous variables when the normality assump-
tion as required for the t test was violated. Because the
two groups were compared on a large number of
characteristics, a more conservative threshold for
significance (P⬍.01) was used instead of the conven-
tional threshold of P⬍.05.
A mixed-effects model23 was used to examine the
relationship between antenatal anxiety and depres-
sion subscale scores and gestational week to charac-
terize the course of antenatal anxiety and depression
across the whole period of pregnancy. This method
was chosen because it accommodates subjects with
unequal numbers of measurements at irregular time
intervals.
Risk factors for antenatal anxiety or depression
(defined as an anxiety or depression subscale score on
the Hospital Anxiety and Depression Scale above the
suggested threshold for a probable case of anxiety or
depression) at each trimester were examined in two
phases. In the first phase, demographic variables were
considered in a logistic regression analysis with a
forward stepwise variables selection procedure. In the
second phase, a logistic regression was performed
with the stepwise procedure performed on the psy-
chosocial characteristics after force-entering those de-
mographics significant in the first phase. The same
analysis was performed to identify predictors of post-
partum depression, but with antenatal anxiety and
depression in each trimester instead of the psychoso-
cial variables entered in the second phase. All esti-
mates were accompanied by a 95% confidence inter-
val (CI).
RESULTS
Of the 423 women invited to participate in the study,
357 agreed, yielding a response rate of 84.4%. Of
these, 335 pregnant women (93.8%) completed all
three antenatal time points, yielding an attrition rate
of 6.2%. More participants dropped out after child-
birth. The total number of participants who com-
pleted all the antenatal assessments and the 6-week
postpartum assessment was 244, rendering a postpar-
tum attrition rate of 27.2%. For analyses involving the
antenatal time points (first trimester to third trimester)
only, the final sample was 335. For analyses involving
VOL. 110, NO. 5, NOVEMBER 2007
the postpartum time point, the final sample was 244.
No significant differences on core demographic vari-
ables were found between those who participated in
all the antenatal assessments and those who dropped
out of the study before giving birth (Table 1). The two
groups also did not differ on psychosocial character-
istics at first trimester in terms of attitudes toward
pregnancy, self-esteem, perceived social support, and
marital satisfaction. Those who participated in all four
time points and those who dropped out after giving
birth were also not significantly different on core
demographic variables examined or psychosocial
characteristics at first, second, and third trimester.
The two groups also did not differ on mean anxiety
and depression scores at first, second, and third
trimester.
The mean age of the antenatal sample (n⫽335)
was 31 (standard deviation [SD] 4.8). Almost all of the
women (95.5%) were either married or cohabiting,
and 98.2% had attained at least a secondary level of
education. More than half (60.1%) were primiparous,
and 77.6% had planned for the pregnancy. Whether
or not the pregnancy was planned, 97.0% of the
pregnancies were wanted. Most (83.1%) of the women
were never smokers, and 78.8% had no history of
drinking. Nearly all were not smoking (96%) or
drinking (95.8%) at the time of baseline assessment.
None reported a history of past psychiatric disorders.
Because too few participants reported current smok-
ing, current drinking, and a history of psychiatric
disorders, these were not further examined in subse-
quent analyses. The sample characteristics are sum-
marized in Table 1. The mean weeks of gestation at
which assessments were carried out were 12.5 (SD
1.2), 19.5 (SD 1.8), and 34.4 (SD 1.7) for first, second,
and third trimester, respectively.
Table 1 also shows the characteristics of the
sample with available antenatal and postpartum data
(n⫽244). In general, the demographic characteristics
of this sample were very similar to those of the
antenatal sample.
The prevalence of antenatal anxiety and depres-
sion was characterized by a U-shaped curve, with
both decreasing from first trimester to second trimes-
ter and then increasing again in the third trimester.
Antenatal anxiety was more prevalent than depres-
sion. The prevalence of antenatal anxiety was 36.3%
(95% CI 33.7–38.9%) at first trimester. The rate
dropped to 32.3% (95% CI 29.7–34.9%) at second
trimester but increased again to 35.8% (95% CI
33.2–38.4%) at third trimester. For antenatal depres-
sion, the prevalence was 22.1% (95% CI 19.9 –24.4%)
at first trimester. A slight drop in prevalence to 18.9%
Lee et al Antenatal Anxiety and Depression 1105
Table 2. Univariable Logistic Regression Analysis of Demographic Risk Factors for Antenatal Anxiety at
Each Trimester
First Trimester

Anxiety No Anxiety
Characteristics (nⴝ121) (nⴝ212) OR (95% CI) P
Age 0.94 (0.90–0.99) .012
Parity
Primigravida 85 (70.2) 115 (54.2) 1.99 (1.24–3.20) .004
Marital status
Single/divorced 8 (6.6) 7 (3.3) 2.07 (0.73–5.87) .169
Education level [1]
Primary 5 (4.1) 1 (0.5) 9.40 (1.04–84.93) .46
Secondary 91 (75.2) 163 (77.3) 1.05 (0.61–1.82) .863
Tertiary or above 25 (20.7) 47 (22.3)
Family income (monthly)* [13] [28]
Less than 20,000 55 (50.9) 83 (45.1) 1.35 (0.76–2.39) .305
20,000–30,000 26 (24.1) 46 (25.0) 1.15 (0.59–2.24) .678
More than 30,000 27 (25.0) 55 (29.9)
History of smoking [2]
Yes 30 (24.8) 26 (12.4) 2.33 (1.30–4.18) .004
History of drinking [3]
Yes 38 (31.4) 32 (15.3) 2.53 (1.48–4.34) .001
OR, odds ratio; CI, confidence interval.
Data are expressed as n (%). Numbers in brackets refer to number of missing values.
* In Hong Kong dollars (1 U.S. dollar equals approximately 7.8 Hong Kong dollars).

(95% CI 16.8 –21.1%) was observed at second trimes-
ter, but the rate increased again to 21.6% (95% CI
19.4 –28.9%) at third trimester. A total of 14.2% (95%
CI 12.3–16.1%) had both antenatal anxiety and ante-
natal depression at first trimester. The corresponding
comorbidity rates at second and third trimester were
12.6% (95% CI 10.8 –14.4%) and 16.9% (95% CI
14.8 –19%), respectively.
A lot of intraindividual instability in anxiety and
depression across the three trimesters was observed.
For anxiety, 17.8% of all the pregnant women had
antenatal anxiety at all three time points, 15.4% had
antenatal anxiety at two time points, and 20.8%
had antenatal anxiety at one time point only. For
depression, 6.9%, 11.2%, 19% had depression at three,
two, and one time point, respectively. More than half
(54%) of the women had anxiety at least once, and more
than a third (37.1%) were depressed at least once.
The pregnant women were further categorized
into the consistent/deteriorating problem group or
the fleeting/no problem group. Women in the
consistent/deteriorating group consisted of those
who 1) screened negative in all three time points, 2)
screened negative in the first trimester but positive
in the second and third trimesters, and 3) screened
negative in the first and second trimesters but
positive in the third trimester. Those in the fleet-
ing/no symptom group consisted of other women.
For anxiety, 31.4% of the pregnant women were in
the consistent/deteriorating group, whereas for de-
pression, only 17.5% were in the consistent/deteri-
orating group. Women who were younger (OR
0.95, 95% CI 0.90 – 0.99, P⬍.05) and had a history
of drinking (OR 2.09, 95% CI 1.13–3.87, P⬍.05)
were more likely to have consistent/deteriorating
anxiety symptoms. Those with a history of drinking
(OR 2.37, 95% CI 1.20 – 4.70, P⬍.05) were more
likely to have consistent/deteriorating depressive
symptoms. In addition, being in the middle
monthly family income category (20,000 –30,000
Hong Kong dollars, which amount to approxi-
mately 2,564 –3,046 U.S. dollars) was a protective
factor. Being in a lower (below 20,000 Hong Kong
dollars) income category increased the risk of hav-
ing consistent/deteriorating anxiety symptoms (OR
2.39, 95% CI 1.18 – 4.81, P⬍.05), whereas being in
an either lower or a higher income group (above
30,000 Hong Kong dollars) increased the risk of
having consistent/deteriorating depressive symp-
toms (OR 3.14, 95% CI 1.22– 8.06, P⬍.05, and OR
3.11, 95% CI 1.13– 8.52, P⬍.05, respectively).
Mixed-effects model analyses were conducted
to examine whether anxiety and depression scores
changed significantly over the period of pregnancy.
Results showed significant time effect with nonlin-
ear association between time and anxiety/depres-

1106 Lee et al Antenatal Anxiety and Depression OBSTETRICS & GYNECOLOGY

 Second Trimester Third Trimester

Anxiety No Anxiety Anxiety No Anxiety

(nⴝ108) (nⴝ226) OR (95% CI) P (nⴝ119) (nⴝ213) OR (95% CI) P
0.957 (0.91–1.00) .073 0.92 (0.88–0.97) .001
[1] [1]
70 (64.8) 130 (57.8) 1.35 (0.84–2.17) .220 79 (66.4) 120 (56.6) 1.51 (0.95–2.42) .082
[1] [1]
6 (5.6) 9 (4.0) 1.41 (0.49–4.07) .524 7 (5.9) 8 (3.8) 1.59 (0.56–4.51) .380
[2] [2]
1 (0.9) 5 (2.2) 0.38 (0.04–3.40) .384 3 (2.5) 3 (1.4) 1.63 (0.31–8.66) .567
82 (75.9) 172 (76.8) 0.90 (0.52–1.56) .697 89 (74.8) 164 (77.7) 0.88 (0.51–1.52) .658
25 (23.1) 47 (21.0) 27 (22.7) 44 (20.9)
[14] [28] [16] [25]
47 (50.0) 91 (46.0) 1.00 (0.56–1.77) .989 59 (57.3) 79 (42.0) 1.44 (0.82–2.54) .208
19 (20.2) 53 (26.8) 0.69 (0.35–1.39) .298 16 (15.5) 55 (29.3) 0.56 (0.27–1.15) .116
28 (29.8) 54 (27.3) 28 (27.2) 54 (28.7)
[3] [1] [2]
25 (23.1) 31 (13.9) 1.87 (1.04–3.35) .037 27 (22.9) 29 (13.7) 1.86 (1.04–3.33) .036
[4] [1] [3]
32 (29.6) 38 (17.1) 2.04 (1.19–3.50) .010 36 (30.5) 34 (16.2) 2.27 (1.33–3.89) .003

sion scores. More specifically, significant changes
during the gestational period for both anxiety and
depression, with a U-shaped relationship between
gestational week and anxiety (P⬍.05) and depres-
sion (P⬍.05) scores, were observed. The lowest
point of anxiety occurred in 23.72 weeks, with an
anxiety subscale score of 5.86, whereas the lowest
point of depression occurred in 24.48 weeks, with a
depression subscale score of 4.93.
The results of univariable logistic regression anal-
yses examining the association of maternal demo-
graphic factors with anxiety at each time point are
presented in Table 2. Being younger (OR 0.94,
P⬍.05), bearing the first child (OR 1.99, P⬍.01),
history of smoking (OR 2.33, P⬍.01), and history of
drinking (OR 2.53, P⬍.01) were significantly associ-
ated with an increased likelihood of anxiety at first
trimester, whereas history of smoking (OR 1.87,
P⬍.05) and drinking (OR 2.04, P⬍.01) were signifi-
cantly associated with anxiety at second trimester. For
the third trimester, younger age (OR 0.92, P⬍.01),
history of smoking (OR 1.86, P⬍.05) and drinking
(OR 2.27, P⬍.01) were associated with an increased
risk of anxiety. In the multiple logistic regression
analyses, history of drinking was associated with a
nearly twofold increased risk of anxiety at first trimes-
ter (OR 2.30, 95% CI 1.29 – 4.10, P⫽.005) and second
trimester (OR 2.16, 95% CI 1.21–3.87, P⫽.009),
whereas being younger (OR 0.93, 95% CI 0.88 – 0.98,
P⫽.007) and history of drinking (OR 1.86, 95% CI
1.03–3.37, P⫽.041) were associated with anxiety at
third trimester.
Similar analyses were conducted for antenatal
depression. The findings are summarized in Table 3.
Of all the demographic risk factors examined, being
single or divorced (OR 3.27, P⬍.05) and having a
history of drinking (OR 2.00, P⬍.05) were signifi-
cantly associated with an increased risk of depression
at first trimester. Younger age (OR 0.92, P⬍.01) was
associated with increased risk of depression at second
trimester, whereas past drinking (OR 2.15, P⬍.05)
significantly predicted depression at third trimester.
For the multiple regression analyses, past drinking
(OR 2.15, 95% CI 1.15– 4.04, P⫽.017) was associated
with a nearly twofold increased risk of depression at
first trimester. Being younger (OR 0.92, 95% CI
0.87– 0.98, P⫽.010) was associated with depression at
second trimester, whereas both past drinking (OR
1.86, 95% CI 1.03–3.38, P⫽.041) and being younger
(OR 0.93, 95% CI 0.88 – 0.98, P⫽.007) were associ-
ated with depression at third trimester.
To examine whether psychosocial factors could
provide additional predictive power of antenatal anx-
iety and depression over and above that of demo-
graphic risk factors, a series of hierarchical multiple
logistic regression analyses were conducted, with the
effects of identified demographic risk factors con-
trolled for. The results are summarized in Table 4.
When a forward stepwise multiple logistic regression
was applied to these psychosocial risk factors, low

VOL. 110, NO. 5, NOVEMBER 2007 Lee et al Antenatal Anxiety and Depression 1107
Table 3. Univariable Logistic Regression Analysis of Demographic Risk Factors for Antenatal Depression
at Each Trimester
First Trimester

Depression No Depression
(nⴝ74) (nⴝ258) OR (95% CI) P
Age 0.98 (0.92–1.03) .351
Parity
Primigravida 48 (64.9) 151 (58.5) 1.31 (0.76–2.24) .327
Marital status
Single/divorced 7 (9.5) 8 (3.1) 3.27 (1.14–9.33) .027
Education level [1]
Primary 3 (4.1) 3 (1.2) 5.00 (0.90–27.82) .066
Secondary 59 (79.7) 194 (75.5) 1.52 (0.77–3.02) .230
Tertiary or above 12 (16.2) 60 (23.3)
Family income (monthly)* [11] [30]
Less than 20,000 34 (54.0) 104 (45.6) 1.59 (0.80–3.18) .191
20,000–30,000 15 (23.8) 56 (24.6) 1.30 (0.58–2.92) .524
More than 30,000 14 (22.2) 68 (29.8)
History of smoking [2]
Yes 16 (21.6) 39 (15.2) 1.54 (0.80–2.90) .196
History of drinking [3]
Yes 23 (31.1) 47 (18.4) 2.00 (1.11–3.58) .021
OR, odds ratio; CI, confidence interval.
Data are expressed as n (%). Numbers in brackets refer to number of missing values.
* In Hong Kong dollars (1 U.S. dollar equals approximately 7.8 Hong Kong dollars).

self-esteem was associated with an increased risk of
anxiety at first trimester (adjusted OR 0.80, P⬍.001),
second trimester (adjusted OR 0.82, P⬍.001), and
third trimester (adjusted OR 0.81, P⬍.001). Low
perceived social support was associated with an in-
creased risk of anxiety at second trimester (adjusted
OR 0.58, P⫽.030), and low marital satisfaction was
associated with an increased risk of anxiety at third
trimester (adjusted OR 0.62, P⫽.009). For depression,
the results of multiple logistic regression indicated
that unwanted pregnancy was associated with a more
than sixfold increased risk of depression at first tri-
mester (adjusted OR 6.51, P⫽.011). Low self-esteem
was associated with increased risk of depression at
first trimester (adjusted OR 0.89, P⫽.005), second
trimester (adjusted OR 0.82, P⬍.001), and third tri-
mester (adjusted OR 0.79, P⬍.001). Increased risk of
depression was predicted by low perceived social
support at first trimester (adjusted OR 0.59, P⫽.044)
and third trimester (adjusted OR 0.46, P⫽.002). Low
martial satisfaction was associated with increased risk
of depression at second trimester (adjusted OR 0.60,
P⫽.017).
A total of 244 women completed the 6-week
postpartum questionnaire. Of them, 24.2% (59 of 244)
scored above threshold on the Edinburgh Postnatal
Depression Scale, indicating a high potential of hav-
ing clinically significant postpartum depression. No
demographic variables were needed to adjust for the
postpartum depression analysis. Antenatal anxiety
and depression in all three trimesters were associated
with postpartum depression (antenatal anxiety: ad-
justed OR 2.66, 95% CI 1.36 –5.20, P⫽.004 in first
trimester; adjusted OR 3.65, 95% CI 1.89 –7.07,
P⬍.001 in second trimester; adjusted OR 3.84, 95%
CI 1.92–7.65, P⬍.001 in third trimester; antenatal
depression: adjusted OR 4.16, 95% CI 2.05– 8.46,
P⬍.001 in first trimester; adjusted OR 3.35, 95% CI
1.62– 6.91, P⫽.001 in second trimester; adjusted OR
2.67, 95% CI 1.27–5.58, P⫽.009 in third trimester).
Antenatal anxiety in late pregnancy and antenatal
depression in early pregnancy were particularly pow-
erful predictors of postpartum depression, increasing
the odds by more than three times and more than four
times, respectively. Those with consistent/deteriorat-
ing antenatal anxiety (adjusted OR 4.12, 95% CI
2.10 – 8.08, P⬍.001) or depression (adjusted OR 3.01,
95% CI 1.38 – 6.60, P⬍.01) had increased risk of
postpartum depression.
DISCUSSION
Mental health problems in the antenatal period are
much less recognized than those in the postpartum
period. Our findings, however, showed that antenatal
mental health problems are prevalent, with antenatal
anxiety symptoms being even more prevalent than

1108 Lee et al Antenatal Anxiety and Depression OBSTETRICS & GYNECOLOGY

 Second Trimester Third Trimester

Depression No Depression Depression No Depression

(nⴝ63) (nⴝ270) OR (95% CI) P (nⴝ72) (nⴝ261) OR (95% CI) P
0.92 (0.87–0.97) .003 0.96 (0.91–1.01) .141
[1] [1]
39 (61.9) 160 (59.3) 1.17 (0.66–2.06) .598 46 (64.8) 153 (58.6) 1.30 (0.75–2.24) .348
[1] [1]
5 (8.1) 10 (3.7) 2.28 (0.75–6.93) .146 4 (5.6) 11 (4.2) 1.36 (0.42–4.40) .611
[1] [1] [1] [1]
3 (4.8) 3 (1.1) 5.55 (0.99–31.10) .052 2 (2.8) 4 (1.5) 1.47 (0.25–8.73) .670
48 (77.4) 205 (76.2) 1.30 (0.64–2.65) .474 51 (71.8) 203 (78.1) 0.74 (0.40-1.37) .338
11 (17.7) 61 (22.7) 18 (25.4) 53 (20.4)
[7] [35] [7] [35]
30 (53.6) 108 (46.0) 1.06 (0.54–2.08) .860 37 (56.9) 101 (44.7) 1.22 (0.64–2.30) .549
9 (16.1) 62 (26.4) 0.56 (0.23–1.34) .190 9 (13.8) 62 (27.4) 0.48 (0.20–1.15) .098
17 (30.4) 65 (27.7) 19 (29.2) 63 (27.9)
[1] [2] [1] [2]
15 (24.2) 40 (147.9) 1.82 (0.93–3.56) .081 16 (22.5) 40 (15.4) 1.59 (0.83–3.05) .161
[1] [3] [1] [3]
17 (27.4) 53 (19.9) 1.53 (0.81–2.88) .192 23 (32.4) 47 (18.2) 2.15 (1.19–3.87) .011

antenatal depressive symptoms. Indeed, more than
half (54%) of our pregnant women had elevated
anxiety at some point during their pregnancies, and
more than a third of them (37.1%) had elevated
depression. This underscores the need for greater
attention to be paid to the mental health and well-
being of pregnant women. This is especially impor-
tant because we also found that comorbidity is com-
mon, with 12.6 –16.9% of pregnant women having
comorbid anxiety and depressive symptoms at vari-
ous stages of pregnancy.
The strength of our study lies in the longitudinal
assessment of both anxiety and depression across all
three trimesters. Our data shows that antenatal anxi-
ety and depression are not static but instead show a
changing course, both in prevalence rates and in
intra-individual anxiety and depression levels. Both
conditions are more prevalent in early pregnancy and
late pregnancy and less prevalent in the second
trimester. A previous cross-sectional study using the
Hospital Anxiety and Depression Scale also found a
similar trend.24 It was, however, difficult to make
conclusions from that study because the sample char-
acteristics of the pregnant women assessed during the
different trimesters were not compared and thus
could potentially be different. Our longitudinal study
provides evidence to confirm that such a U-shaped
relationship between stage of pregnancy and anxiety
and depression reflects the true state of affairs rather
than being an artifact of methodological or sampling
bias, especially given our extremely low attrition rate.
Intra-individual changes in both anxiety and de-
pressive symptoms across different stages of preg-
nancy are characterized by a U-shaped curve, with
the levels being lowest in the second trimester. Al-
though previous studies demonstrated a generally
increasing trend of psychological disturbances during
pregnancy,12 a recent meta-analysis7 found a similar,
but insignificant, trend for antenatal depression. We
used a powerful and robust analytical tool, the mixed-
effects model, and found that both anxiety and de-
pression levels decreased from early to mid-preg-
nancy but increased again in late pregnancy.
Quite a substantial proportion of pregnant
women had anxiety or depressive symptoms at only
one or two time points. This shows that new cases of
anxiety or depression or both can emerge in any
trimester, implying that a one-time screening at any
one antenatal visit only is not sufficient. Clinicians
should be vigilant of potential cases of anxiety and
depression emerging in different stages of pregnancy,
with ongoing screening being done throughout. It is
dangerous to exclude a case of probable antenatal
anxiety or depression with information from only one
antenatal visit.
With regard to identification of high-risk individ-
uals, young age was found to be associated with
antenatal anxiety and depressive symptoms. History
of drinking was found to be one of the most significant
risk factors, predicting anxiety symptoms in all three
trimesters and depression in the first and third trimes-
ters. It is not exactly clear why history of drinking is

VOL. 110, NO. 5, NOVEMBER 2007 Lee et al Antenatal Anxiety and Depression 1109
Table 4. Psychosocial Characteristics Associated With Antenatal Anxiety and Depression at Each
Trimester After Controlling for Demographic Risk Factors
First Trimester
(nⴝ322, R2ⴝ19.2%)
aOR (95% CI)* P aOR (95% CI)*
Planned pregnancy — —
Wanted pregnancy — —
Social support — —
Marital satisfaction — —
Self-esteem 0.80 (0.74–0.87) ⬍.001
aOR, adjusted odds ratio; CI, confidence interval; R2, percentage of variance accounted for by the model.
* Demographic variables adjusted: Antenatal Anxiety—first and second trimester: drinking history; third trimester: age and drinking history;
Antenatal Depression—first trimester: drinking history; second trimester: age; third trimester: age and drinking history.
such a significant predictor, an area that definitely
requires further investigation. Because drinking is
generally not common among Chinese females,25,26
and a strong relationship exists between drinking
behavior and psychological distress,27 it is possible
that women with a history of drinking may have
varying degrees of unrecognized and untreated psy-
chological disturbance before pregnancy and, hence,
are more prone to developing anxiety and depressive
symptoms when faced with stressors associated with
pregnancy. Alternatively, their anxiety and depres-
sive symptoms may represent withdrawal symptoms
resulting from alcohol abstinence during pregnancy.
Among psychosocial variables, self-esteem was
the most predictive factor for anxiety and depression
in all three trimesters. Our findings add to the accu-
mulating body of evidence delineating self-esteem as
a risk factor for antenatal anxiety13 by demonstrating
that it is consistently associated with both anxiety and
depressive symptoms across all stages of pregnancy.
Pregnant women with low self-esteem are ill-
equipped to face the multitude of developmental
challenges and stressors of pregnancy and, thus, are
more prone to anxiety and depressive symptoms
throughout pregnancy. Given the salience of self-
esteem as a risk factor across all stages of pregnancy,
close monitoring of women with low self-esteem is
warranted. Development of intervention for enhanc-
ing self-esteem to prevent antenatal psychological
disturbances should also be encouraged.
Perceived social support and marital satisfaction
have been identified as protective factors for antenatal
anxiety and depression in the literature.13,14 Our study
showed that their significance vary across different
stages of pregnancy. Perceived social support pro-
tected against anxiety in the second trimester and
1110 Lee et al Antenatal Anxiety and Depression
Antenatal Anxiety
Second Trimester Third Trimester
(nⴝ326, R2ⴝ19.8%) (nⴝ322, R2ⴝ25.6%)
P aOR (95% CI)* P
— — — —
— — — —
0.58 (0.35–0.95) .030 — —
— — 0.62 (0.43–0.89) .009
0.82 (0.76–0.89) ⬍.001 0.81 (0.75–0.88) ⬍.001
against depression in the first and third trimesters.
Marital satisfaction, on the other hand, protected
against anxiety in the third trimester and against
depression in the second trimester. The exact reason
for this pattern of results requires further examination,
but the general picture reflects that external resources
are important, in addition to internal resources such
as self-esteem. Unfavorable external factors place
pregnant women at risk of psychological distress.
Unwanted pregnancy was found to be a signifi-
cant predictor of depression only in the first trimester,
increasing its risk by more than six times. Preliminary
evidence in the literature showed that unwanted
pregnancy places women under great psychological
risks,14 but these studies did not examine its varying
importance across different stages of pregnancy. Our
study revealed that its significance decreases over
time. It is possible that, initially, pregnant women
with unwanted pregnancies found it hard to accept
the reality, predisposing them to depression. How-
ever, as pregnancy progresses, the shock subsides and
the bonding with the growing fetus strengthens, thus
facilitating acceptance. Other factors, such as self-
esteem, assume greater importance as the women
grapple with ways to cope with the unwanted but real
event.
Several implications are evident from our find-
ings of risk factors for antenatal anxiety and depres-
sive symptoms in different stages of pregnancy. First,
we found that risk factors differ slightly in different
trimesters, suggesting that anxiety or depression
emerging in different stages of pregnancy might rep-
resent clinical problems with different pathogeneses.
It is also evident from our findings that psychosocial
factors play an important role, with significant predic-
tive power even when the influence of demographic
OBSTETRICS & GYNECOLOGY
 Antenatal Depression

First Trimester Second Trimester Third Trimester

(nⴝ322, R2ⴝ19.2%) (nⴝ326, R2ⴝ19.8%) (nⴝ322, R2ⴝ25.6%)

aOR (95% CI)* P aOR (95% CI)* P aOR (95% CI)* P

— — — — — —
6.51 (1.53–27.69) .011 — — — —
0.59 (0.35–0.99) .044 — — 0.46 (0.27–0.75) .002
— — 0.60 (0.39–0.91) .017 — —
0.89 (0.82–0.97) .005 0.82 (0.74–0.90) ⬍.001 0.79 (0.72–0.87) ⬍.001

risk factors was controlled for. This highlights the
need to attend to the pregnant woman’s psychosocial
makeup, to assess this in determining her risk status,
and to develop ways to improve her psychosocial
resources to prevent or treat antenatal anxiety and
depression.
Both antenatal anxiety and depressive symptoms
predicted postpartum depression in our sample of
pregnant women, further underscoring the impor-
tance of identifying and treating antenatal anxiety and
depression. Previous studies mainly identified antena-
tal depression as a risk factor for postpartum depres-
sion although some studies also recognized antenatal
anxiety as a risk factor.3,28 It is perhaps not surprising
that depressive symptoms during pregnancy continue
through to the postpartum period, but the salience of
antenatal anxiety symptoms in predicting postpartum
depression is somewhat unexpected. The association
between antenatal anxiety symptoms and postpartum
depression also increases as pregnancy progresses,
with anxiety symptoms in late pregnancy being most
strongly associated with postpartum depression. We
do not know the exact nature of the anxiety experi-
enced by pregnant women, but it is possible that their
anxiety is anticipatory in nature, arising out of per-
ceived inadequate resources to cope with the chal-
lenges of pregnancy and motherhood. Women who
are still anxious in the final trimester are those with
most difficulty in adjusting to the maternal role and
least confidence in meeting the demands of mother-
hood. In the postpartum period, the demands took on
real form. Instead of being anxious about impending
challenges, they became depressed by their inability
to grapple with the real demands of motherhood.
More qualitative studies of the exact nature of ante-
natal anxiety would better inform us about the mech-
anisms underlying its relationship with postpartum
depression. Nevertheless, preventing, identifying, and
treating both antenatal anxiety and antenatal depres-
sion should be made a priority in efforts to prevent
postpartum depression.
We attempted to characterize a group of pregnant
women who deserve particular clinical attention as
compared with those with consistent or deteriorating
anxiety or depressive symptoms. Our analyses
showed that this group tended to be younger, to be in
the middle income range, and to have a history of
drinking. Women with these risk factors certainly
require close clinical monitoring and frequent assess-
ment for anxiety and depressive symptoms, especially
since they were also found to be more prone to
postpartum depression.
Our study has several limitations. First, there is
no good local data concerning the prevalence of
anxiety and depression in women of childbearing age
for comparison. It is therefore difficult to tell whether
the rates of depression and anxiety are higher in
pregnant women than among non-childbearing
women of similar ages. Further studies should try to
include matched controls to detect whether there are
differences. Second, we used only self-report screen-
ing tools to assess antenatal anxiety and depressive
symptoms. Scoring above the cutoff score only indi-
cates that the individual is a “probable case of anxiety
or depression.” To get more accurate estimates of the
prevalence of these problems, follow-up clinical inter-
views with the identified cases should be performed in
future studies. However, because we used the Hospi-
tal Anxiety and Depression Scale, which has very
good sensitivity and specificity values, our estimates
should have considerable validity.
Notwithstanding these limitations, our study
serves as an important step toward recognizing the
dynamic nature of anxiety and depression across
different stages of pregnancy. We also identified
history of drinking, young age, low self-esteem, low
perceived social support and marital satisfaction, and
unwanted pregnancy as risk factors for antenatal

VOL. 110, NO. 5, NOVEMBER 2007 Lee et al Antenatal Anxiety and Depression 1111
anxiety and depressive symptoms. Finally, our study
also points to the need for greater research and
clinical attention to the much neglected but common
clinical problem of antenatal anxiety. It is time to shift
our emphasis from the postpartum period to the
antenatal period and to shift our focus from depres-
sion to anxiety.
REFERENCES
1. Spinelli MG. Interpersonal psychotherapy for depressed ante-
partum women: a pilot study. Am J Psychiatry 1997;154:
1028–30.
2. Klein MH, Essex MJ. Pregnant or depressed? The effect of
overlap between symptoms of depression and somatic com-
plaints of pregnancy on rates of major depression in second
trimester. Depression 1995;2:308–14.
3. Heron J, O’Connor TG, Evans J, Golding J, Glover V. The
course of anxiety and depression through pregnancy and the
postpartum in a community sample. J Affect Dis 2004;80:
65–73.
4. Evans J, Heron J, Francomb H, Oke S, Golding J. Cohort study
of depressed mood during pregnancy and after childbirth. BMJ
2001;323:257–60.
5. Andersson L, Sundstrom-Poromaa I, Bixo M, Wulff M, Bond-
estam K, Astrom M. Point prevalence of psychiatric disorders
during the second trimester of pregnancy: a population-based
study. Am J Obstet Gynecol 2003;189:148–54.
6. Bennett HA, Einarson A, Taddio A, Koren G, Einarson T.
Prevalence of depression during pregnancy: systematic review.
Obstet Gynecol 2004;103:698–709.
7. Gavin NI, Gaynes BN, Lohr KN, Meltzer-Brody S, Gartlehner
G, Swinson T. Perinatal depression: a systematic review of
prevalence and incidence. Obstet Gynecol 2005;106:1071–83.
8. Field T, Diego M, Hernandez-Reif M, Schanberg S, Kuhn C,
Yando R, et al. Pregnancy anxiety and comorbid depression
and anger: effects on the fetus and neonate. Depress Anxiety
2003;17:140–51.
9. Chung TK, Lau TK, Yip AS, Chiu HF, Lee DT. Antepartum
depressive symptomatology is associated with adverse obstet-
ric and neonatal outcomes. Psychosom Med 2001;63:830–4.
10. Kurki T, Hiilesmaa V, Raitasalo R, Mattila H, Ylikorkala O.
Depression and anxiety in early pregnancy and risk for
preeclampsia. Obstet Gynecol 2000;95:487–90.
11. Larsson C, Sydsjo G, Josefsson A. Health, sociodemographic
data, and pregnancy outcome in women with antepartum
depressive symptoms. Obstet Gynecol 2004;104:459–66.
12. Buesching DP, Glasser ML, Frate DA. Progression of depres-
sion in the prenatal and postpartum periods. Women Health
1986;11:61–78.
1112 Lee et al Antenatal Anxiety and Depression
13. Littleton HL, Breitkopf CR, Berenson AB. Correlates of anx-
iety symptoms during pregnancy and association with perina-
tal outcomes: a meta-analysis. Am J Obstet Gynecol 2007;196:
424–32.
14. Bowen A, Muhajarine N. Antenatal depression. Can Nurse
2006;102:27–30.
15. Cohen J. Statistical power analysis for the behavioral sciences.
2nd ed. Hillsdale (NJ): L. Erlbaum Associates; 1988.
16. Zigmond AS, Snaith RP. The Hospital Anxiety and Depres-
sion Scale. Acta Psychiatr Scand 1983;67:361–70.
17. Leung CM, Ho S, Kan CS, Hung CH, Chen CN. Evaluation of
the Chinese version of the Hospital Anxiety and Depression
Scale: a cross-cultural perspective. Int J Psychosom 1993;40:
29–34.
18. Abiodun OA. A validity study of the Hospital Anxiety and
Depression Scale in general hospital units and a community
sample in Nigeria. Br J Psychiatry 1994;165:669–72.
19. Murray D, Cox J. Screening for depression during pregnancy
with the Edinburgh Depression Scale (EPDS). J Reprod Infant
Psychol 1990;8:99–107.
20. Cox JL, Holden J, Sagovsky R. Detection of postnatal depres-
sion: development of the 10-item Edinburgh Postnatal Depres-
sion Scale. Br J Psychiatry 1987;150:782–6.
21. Lee DTS, Yip SK, Chiu HFK, Leung TYS, Chan KPM, Chau
IOL, et al. Detecting postnatal depression in Chinese women.
Validation of the Chinese version of the Edinburgh Postnatal
Depression Scale. Br J Psychiatry 1998;172:433–7.
22. Rosenberg M. Society and the adolescent self-image. Princeton
(NJ): Princeton University Press; 1965.
23. Verbeke G, Molenberghs G. Linear mixed models for longi-
tudinal data. New York (NY): Springer-Verlag; 2000.
24. Berle JO, Mykletun A, Daltveit AK, Rasmussen S, Holsten F,
Dahl AA. Neonatal outcomes in offspring of women with
anxiety and depression during pregnancy. Arch Womens
Ment Health 2005;8:181–9.
25. Janghorbani M, Ho SY, Lam TH, Janus ED. Prevalence and
correlates of alcohol use: a population-based study in Hong
Kong. Addiction 2003;98:215–24.
26. Lam SK. Alcoholism in pregnancy. J HK C Psych 1992;2:
37–9.
27. Dawson DA, Grant BF, Stinson FS, Chou PS. Psychopathology
associated with drinking and alcohol use disorders in the
college and general adult populations. Drug Alcohol Depend
2005;77:139–50.
28. Sutter-Dallay AL, Giaconne-Marcesche V, Glatigny-Dallay E,
Verdoux H. Women with anxiety disorders during pregnancy
are at increased risk of intense postnatal depression symptoms:
a prospective survey of the MATQUID cohort. Eur Psychiatry
2004;19:459–63.
OBSTETRICS & GYNECOLOGY