ORIGINAL ARTICLE
The Effects of Neuroticism, Extraversion, and Positive and
Negative Life Events on a One-Year Course of Depressive
Symptoms in Euthymic Previously Depressed
Patients Versus Healthy Controls
Philip Spinhoven, PhD,*Þ Bernet Elzinga, PhD,* Karin Roelofs, PhD,* Jacqueline G.F.M. Hovens, MD,Þ
Patricia van Oppen, PhD,þ Frans G. Zitman, PhD,Þ and Brenda W.J.H. Penninx, PhDÞþ§
Abstract: We investigated a) the concurrent impact of positive and negative
life events on the course of depressive symptoms in persons remitted from
depression and healthy controls, b) whether the impact of life events on
symptom course is moderated by the history of depression and the personality
traits of neuroticism and extraversion, and c) whether life events mediate
possible relationships of history of depression and personality traits with
symptom course. Using data from the Netherlands Study of Depression and
Anxiety, we examined 239 euthymic participants with a previous depressive
disorder based on DSM-IV and 450 healthy controls who completed a) baseline assessments of personality dimensions (NEO Five-Factor Inventory) and
depression severity (Inventory of Depressive Symptoms [IDS]) and b) 1-year
follow-up assessments of depression severity and the occurrence of positive
and negative life events during the follow-up period (List of Threatening
Events Questionnaire). Remitted persons reported higher IDS scores at 1-year
follow-up than did the controls. Extraversion and positive and negative life
events independently predicted the course of depressive symptoms. The impact
of life events on symptom course was not moderated by history of depression
or personality traits. The effect of extraversion on symptom course was partly
caused by differential engagement in positive life events.
Key Words: Depression, life events, neuroticism, remission, extraversion.
(J Nerv Ment Dis 2011;199: 684Y689)
M
ajor depression is one of the most prevalent and recurrent forms
of psychopathology, with at least 50% of persons who recover
from a first episode experiencing one or more additional episodes and
about 80% of those with a history of two episodes having another
recurrence (Kupfer et al., 1996; Post, 1992). Because of its highly
recurrent nature, recent research focuses on identifying causes of recurrence instead of first episodes (Lewinsohn et al., 1999). Available
*Institute of Psychology, Leiden University, Leiden; †Department of Psychiatry,
Leiden University Medical Center, Leiden; ‡Department of Psychiatry/EMGO
Institute, VU University Medical Center, Amsterdam; and §Department of
Psychiatry, University Medical Center Groningen, Groningen, The Netherlands.
The infrastructure for the Netherlands Study of Depression and Anxiety
(www.nesda.nl) is funded through the Geestkracht program of the Netherlands
Organisation for Health Research and Development (Zon-Mw, grant number
10-000-1002) and is supported by the following participating universities
and mental health care organizations: VU University Medical Center, GGZ
inGeest, Arkin, Leiden University Medical Center, GGZ Rivierduinen,
University Medical Center Groningen, Lentis, GGZ Friesland, GGZ Drenthe,
Scientific Institute for Quality of Healthcare (IQ Healthcare), Netherlands
Institute for Health Services Research (NIVEL) and Netherlands Institute
of Mental Health and Addiction (Trimbos).
The authors have no conflicts of interest to declare.
Send reprint requests to Philip Spinhoven, PhD, Leiden University, Institute of
Psychology, Wassenaarseweg 52, 2333 AK Leiden, The Netherlands.
E-mail: [email protected].
Copyright * 2011 by Lippincott Williams & Wilkins
ISSN: 0022-3018/11/19909-0684
DOI: 10.1097/NMD.0b013e318229d21f
684
www.jonmd.com
studies show that risk factors for the incidence of a first episode differ
from those of recurrent depressive episodes (Burcusa and Iacono,
2007). In contrast with demographic factors, clinical characteristics
(e.g., age at first onset and number of previous depressive episodes)
are associated with recurrence of depression.
Relevant for the purpose of the present study are psychological and psychosocial risk factors. Neuroticism and stressful life
events have both been identified as risk factors for recurrent depression, whereas positive events may constitute a protective factor
(Berlanga et al., 1999; Monroe et al., 1996; Needles and Abramson,
1990). Because recurrence increases further risk for depression, it
follows that either depression increases vulnerability for recurrent
depression or that individuals at high risk for recurrence already
possess certain vulnerability characteristics. Scar theorists presume
that some characteristics change during a depressive episode, and
that this change is long-lasting and makes a future episode more
likely. Among others, neuroticism has been hypothesized as a scar
from depression, although the available evidence is equivocal (Burcusa
and Iacono, 2007). In the specific context of the recurrence of depression, the occurrence of negative life events can be conceptualized
based on two types of scar models: sensitization and stress generation.
The first model assumes that the history of depression moderates the
relationship of stress with depression recurrence. This moderation
model is based on both sensitization to stressors (i.e., less stress is
needed to elicit the same depressogenic response) and episode sensitization (i.e., less dysphoric mood is needed to produce a depressive
episode; Post, 1992). The second model assumes that stressful life
events mediate the relationship of history of depression with depression recurrence. According to this mediation model, depressed persons generate more stressful life events for themselves with an
increasing number of episodes, resulting in additional recurrences
(Hammen, 1991).
Besides neuroticism, extraversion may also be a relevant basic
personality dimension in the course of depressive symptoms. The
basic ‘‘temperamental cores’’ of neuroticism and extraversion (Clark
et al., 1994) may both be heritable (e.g., Fanous et al., 2002; Hettema
et al., 2004; Viken et al., 1994) and account not only for the onset
and course of depression but also for its recurrence. Although extraversion is not as strongly and consistently related to depression and
anxiety disorders across diagnostic categories as neuroticism is, in
cross-sectional studies, high extraversion has been found to be associated with lower levels of depressive symptoms in particular
(Watson et al., 2005). However, its role as a risk factor for depression
is more obscure than that of neuroticism (Enns and Cox, 1997).
Furthermore, previous studies found that persons scoring high on
extraversion experience more (subjectively) positive events than do
persons low in extraversion (Headey and Wearing, 1989; Magnus
et al., 1993). The question of whether a possible relationship between
extraversion and course of depressive symptoms may be caused by
differential exposure to positive life events, however, awaits further
empirical study (Zautra et al., 2005).
The Journal of Nervous and Mental Disease
& Volume 199, Number 9, September 2011
Copyright © 2011 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
The Journal of Nervous and Mental Disease
& Volume 199, Number 9, September 2011
Depending on a range of background factors, the same stressful
experience can result in sensitization or habituation because the impact of many risk factors for psychiatric illness is context dependent
(Kendler, 2008). Until now, no studies have investigated how individual differences in personality characteristics can moderate the
impact of life events on the course of depression in euthymic previously depressed persons at risk of recurrence. Identification of individual differences characteristic of previously depressed persons at
risk of symptom deterioration and, consequently, recurrence could
yield important clinical information that will help to reduce the risk
of recurrence. This dearth of studies is remarkable given previous
studies showing that psychosocial adversity interacts with neuroticism
and that the impact of neuroticism on the incidence of a first depressive episode is greater at high than at low levels of adversity
(Kendler et al., 1993; Ormel et al., 1989). In addition, in predicting
remission rates, highly neurotic patients with depression also seem
to benefit more from positive life change than do others (Oldehinkel
et al., 2000). Finally, the question of whether the history of depression
or personality traits predict the course of depression mediated by life
events has hardly been investigated, although there is limited evidence
that neuroticism predicts a greater likelihood of experiencing a variety
of negative life stressors (Kendler et al, 2003), whereas extraversion
predicts a greater likelihood of experiencing (subjectively) positive
events (Headey and Wearing, 1989; Magnus et al., 1993).
The primary purpose of the present study in euthymic previously depressed persons and healthy controls was to investigate the
association between positive and negative life events and the course of
depressive symptoms and to evaluate history of depression and personality traits as putative moderators of these relationships. Moreover, whether life events mediate possible relationships of history
of depression or personality traits with symptom course will be tested.
It was hypothesized that positive and negative life events will independently predict the course of depressive symptoms. Moreover,
the highest impact of negative and positive life events on the severity
of depressive symptoms was expected in persons with a history of
depression and persons with higher levels of neuroticism and lower
levels of extraversion. Moreover, we expected that the predictive
value of neuroticism may be partly caused by differential exposure
to negative life events, whereas the prognostic value of extraversion may be partially attributed to differential exposure to positive
life events.
METHODS
Participants
The data for the present study were drawn from the Netherlands
Study of Depression and Anxiety (NESDA), an ongoing 8-year longitudinal cohort study aimed at examining the long-term course of
depressive and anxiety disorders in different healthcare settings and
phases of illness. A total of 2981 respondents were recruited from
primary care, specialized mental health care, and the community,
including healthy controls, respondents with subthreshold symptoms,
and those with an anxiety and/or depressive disorder. All 2981
respondents were administered a baseline assessment, which lasted,
on average, for 4 hours and included an assessment of psychopathology, demographic and personal characteristics, psychosocial
functioning, and biomarkers. Further details about NESDA are provided elsewhere (Penninx et al., 2008). A general inclusion criterion
was age 18 to 65 years. Excluded were patients with a primary diagnosis of psychotic disorder, obsessive-compulsive disorder, bipolar disorder, or severe addiction disorder (requiring care in specialized
addiction clinics). A second exclusion criterion was not being fluent
in Dutch. The study protocol was approved centrally by the Ethical
Review Board of the VU Medical Centre Amsterdam and subsequently
* 2011 Lippincott Williams & Wilkins
Personality, Life Events, Depression
by the local review boards of each participating institute. After full
verbal and written information about the study, written informed consent was obtained from all participants.
The 1-year follow-up assessment completed by 2445 participants (82.0% response) consisted of a written questionnaire containing the most important self-report instruments to determine
demographic changes, recent life events, and course of anxiety and
depressive symptoms and health consequences.
We selected a subsample of previously depressed persons with
low levels of depressive symptoms because the level of residual
symptoms is an established predictor of relapse (Burcusa, and Iacono,
2007) and because we wanted comparable levels of depression severity in remitted persons versus in healthy controls. In this way, we
also tried to obtain a more valid measurement of neuroticism not
confounded by current depressive symptoms. Consequently, we selected euthymic remitted participants who did not fulfill the criteria for
major depressive disorder during the last 6 months and had IDS scores
below the cutoff of 14 for mild depression (Rush et al., 1996). In the
total sample, 273 participants with a history of depression and no
current 6-month anxiety or depressive disorder at baseline had IDS
scores below the cutoff of 14 for mild depression (Rush et al., 1996)
and constituted our subgroup of euthymic previously depressed
individuals, hereafter called remitted patients. Among them, 254
(93.0%) had remitted more than 12 months ago, and 144 (60.2%) had
had a single previous episode. Of the remitted participants, 234 still
manifested mild to severe residual symptoms as evidenced by IDS
scores of 14 or more (mean [SD], 22.59 [7.13]) and were consequently
not included in the present study. Among them, 198 (84.6%) achieved
remission status more than 12 months ago, and 109 (46.6%) had had a
single previous depressive episode. The control group consisted of
498 healthy controls without current or lifetime anxiety or depressive
or alcohol diagnosis.
Materials
The diagnoses of current 6-month and lifetime depressive
(dysthymia, major depressive disorder) and anxiety disorders (generalized anxiety disorder, social phobia, panic disorder with or without agoraphobia, or agoraphobia) were established at baseline using
the Composite International Diagnostic Interview (CIDI; life time
version 2.1), which classifies diagnoses according to the DSM-IV
criteria (APA, 2001; Ter Smitten et al., 1998). The CIDI is used
worldwide and World Health Organization field research has found
high interrater reliability, high test-retest reliability, and high validity
for depressive and anxiety disorders (Wittchen et al., 1989, 1991).
Specially trained clinical research staff conducted the CIDI interview.
At baseline and after 1-year follow-up, the severity of depressive symptoms was measured using the 30-item Inventory of Depressive Symptoms self-report version (IDS-SR), which has shown
high correlations with observer-rated scales such as the Hamilton
Depression Rating Scale and with established responsiveness to
change (Rush et al., 1996). Internal consistency of the IDS-SR at
baseline in the present study was 0.94.
At baseline, neuroticism and extraversion were measured using
12-item subscales from the NEO Five-Factor Inventory (NEO-FFI;
Costa and McCrae, 1992; Dutch version: Hoekstra et al., 1996).
Cronbach alpha in the present study was 0.75 for the neuroticism
subscale and 0.78 for the extraversion subscale.
At 1-year follow-up, the incidence of twelve negative life
events (such as serious illness and injury or death of close friend
or relative) during the 1-year follow-up period was assessed using
the List of Threatening Events Questionnaire (LTE-Q; Brugha et al.,
1985). The LTE-Q has shown good test-retest reliability, high
agreement between participant and informant ratings, and good
agreement with interview-based ratings (Brugha and Cragg, 1990).
The LTE-Q was extended by seven items referring to positive life
www.jonmd.com
Copyright © 2011 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
685
The Journal of Nervous and Mental Disease
Spinhoven et al.
events: a) an immediate family member recovering from a serious
illness, b) meeting a new partner, c) becoming friends, d) having been
on holiday, e) having a new job or an important promotion, f ) an
education completed, and g) being better off financially. Negative
and positive life events during the follow-up period were summed
up to derive separate measures for the number of negative and positive life events. The subscales proved to be totally unrelated (r = 0.01,
not significant).
Statistical Analysis
First, the course of depressive symptoms as measured with
the IDS was analyzed using a 2 2 repeated-measures analysis of
variance (ANOVA) with Group (healthy controls versus remitted
persons) as the between-subjects variable and Time (baseline versus
1<year follow-up) as the within-subjects variable. To determine the
effect of positive and negative life events during follow-up on IDS
scores at 1-year follow-up, these variables were forced into a multiple
regression analysis controlling for age, sex, education, and IDS scores
at baseline (Model 1). Afterward, to investigate whether the history
of depression (Model 2) and the personality traits of neuroticism and
extraversion (Model 3) independently predict symptom course, these
variables were forced into the model in consecutive steps while controlling for history of anxiety disorder. Moreover, to investigate
whether the impact of life events on symptom course is moderated
by history of depressive disorders and personality traits, six dummy
variables representing the interaction of positive and negative life
events with history of depression, neuroticism and extraversion were
added to the prediction model. The resulting model (Model 4) was
subsequently reduced by removing nonsignificant interaction terms
(p G 0.10) one at a time, after which the model was rerun. To guard
against multicollinearity, the Variance Inflation Factor score for each
variable in each predictor model was examined. Besides standardized
betas and t- and p-values, zero-order and part (or semipartial) correlation coefficients are also provided as measures for the strength of the
association of predictor variables with outcome. A semipartial correlation coefficient represents the correlation between the criterion
and a predictor that has been residualized with respect to all other
predictors in the equation. The hypothesis that life events mediate
possible relationships of the history of depressive disorders or personality traits with symptom course was tested following the analytic
steps outlined by Baron and Kenny (1986). The macro for SPSS
developed by Preacher and Hayes (2008) was used to generate estimates for the indirect effects in the mediator models, drawing 5000
samples. Analyses were carried out using SPSS (version 17.0) and
alpha was set at G0.05.
RESULTS
Sociodemographic and Clinical Data
Of the 771 selected participants (273 persons with history of
depression and 498 healthy controls), 689 (89.4%) completed the
1<year follow-up questionnaire (239 persons with history of depression and 450 healthy controls). No significant differences regarding
sex, education, severity of depressive symptoms at baseline, neuroticism, and extraversion between completers and noncompleters were
observed. However, compared with completers, noncompleters were
younger (40.0 T 13.2 vs. 45.9 T 12.2 years; t[769] = 4.161, p G 0.001).
Compared with healthy controls, remitted persons were younger
(t[687] = 2.589, p G 0.01), were predominantly female (W2 = 4.23, p =
0.04), and reported higher levels of neuroticism (t[687] = 4.37, p G
0.001) and lower levels of extraversion (t[687] = 3.33, p G 0.001).
There were no significant differences with respect to the level of
education, depression severity at baseline, and number of negative and
positive life events during the 1-year follow-up (Table 1).
686
www.jonmd.com
& Volume 199, Number 9, September 2011
TABLE 1. Demographic and Clinical Characteristics (n = 689)
Remitted
Patients
(n = 239)
Variables
Female
Age, y
Education, y
IDS baseline
IDS 1-y follow-up
Previous anxiety
disorder
Negative life
events
Positive life events
Neuroticism score
Extraversion score
Healthy
Controls
(n = 450)
Test
Statistic
Mean
or %
SD or n
Mean
or %
SD or n
W2(1) or t(687)
71.1
44.2
13.1
7.73
9.19
39.3
(170)
T12.6
T3.1
T3.81
T6.24
(94)
63.3
46.7
13.1
8.29
7.42
Y
(285)
T11.9
T3.3
T7.39
T6.91
Y
4.23*a
2.59**
0.19
1.10
j3.32***
Y
1.0
T1.2
0.9
T1.1
j0.55
1.9
28.4
39.5
T1.2
T8.2
T8.8
1.7
26.4
41.7
T1.1
T7.5
T6.7
j1.33
j4.37***
3.33***
a
chi-square value.
* pG 0.05.
** p G 0.01.
*** p G 0.001.
IDS indicates Inventory of Depressive Symptoms.
Course of Symptoms During One-Year Follow-Up
As can be seen in Table 1, depression severity of healthy
controls diminished during the follow-up period but increased in remitted patients. The course of depressive symptoms (IDS) during the
1-year follow-up was analyzed using a 2 2 repeated-measures
ANOVA. This analysis only yielded a significant interaction effect for
Time Group (F[1687] = 25.338, p G 0.001), whereas the main
effects for Group and Time were nonsignificant. A subsequent analysis of covariance controlling for baseline IDS scores showed a significant effect for Group (F[1,686] = 25.055, p G 0.001, G2 = 0.035), with
remitted persons reporting higher IDS scores at follow-up than did
healthy controls with a small (defined as G2 = 0.01) to medium (defined
as G2 = 0.06) effect size.
Prediction of Symptom Course by Life Events
Positive and negative life events during follow-up were forced
into a multiple regression analysis controlling for age, sex, education,
and IDS scores at baseline, with follow-up IDS scores as the dependent variable (F change[2682] = 29.88, p G 0.001; Model 1). The
intercorrelations among predictors indicated no serious problems of
multicollinearity in these and subsequent analyses (all values G2.5).
Of note is the relatively high association of neuroticism with IDS
depression scores (r = 0.60, p G 0.001), whereas the association of
extraversion with IDS scores was much smaller (r = j0.21, p G
0.001). The results indicated that positive and negative life events
were independent and statistically significant predictors of the course
of depression (positive life events: A = j0.105, t = j3.24, p G 0.001,
part correlation = j0.09; negative life events: A = 0.223, t = 7.30,
p G 0.001, part correlation = 0.21).
Prediction of Symptom Course by Life
Events and History of Depression
Next, history of depression and anxiety were added to the
base model (Model 2) and significantly improved upon Model
1 (F change[2680] = 14.86, p G 0.001). Both history of depressive
(A = 0.092, t = 2.67, p = 0.008, part correlation = 0.14) and anxiety
* 2011 Lippincott Williams & Wilkins
Copyright © 2011 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
The Journal of Nervous and Mental Disease
& Volume 199, Number 9, September 2011
Personality, Life Events, Depression
TABLE 2. Standardized Betas of Predictor Variables of IDS Scores in the Total Sample at 1-Year Follow-Up in the Final Multiple
Regression Model (n = 689)
Total
Variables
Step 1 (Covariates)
Baseline IDS
Age
Sex
Education
Step 2 (Model 1)
PLE
NLE
Step 3 (Model 2)
Previous depression
Previous anxiety
Step 4 (Model 3)
Neuroticism
Extraversion
A
t
0.515
j0.030
0.813
0.056
12.779***
j1.732
1.961*
0.918
j0.542
1.302
j2.797**
7.433***
F Change
F
95.43***
95.43***
29.88***
14.86***
1.112
1.267
0.871
j4.096***
Part r
0.59***
j0.03
0.11**
j0.06*
0.36***
j0.05
0.06*
0.03
j0.11**
0.34***
j0.08**
0.21***
78.95***
65.34***
2.261*
1.861
0.13***
0.15***
8.77***
0.028
j0.112
r
0.06*
0.05
55.22***
0.41***
j0.27***
0.02
j0.12***
*p G 0.05.
**p G 0.01.
*** p G 0.001.
IDS indicates Inventory of Depressive Symptoms; PLE, Number of positive life events; NLE, Number of negative life events; r, zero-order correlation; part r, semipartial correlation
coefficient.
(A = 0.088, t = 2.547, p = 0.011, part correlation = 0.07) disorders were
statistically significant predictors.
Prediction of Symptom Course by Life Events,
History of Depression, and Personality Traits
In the next step (Model 3), neuroticism and extraversion were
added to Model 2. The inclusion of neuroticism and extraversion
significantly improved upon Model 2 (F change[2678] = 8.77, p G
0.001), but only extraversion (A = j0.125, t = j4.096, p G 0.001, part
correlation = j0.12), not neuroticism (A = 0.032, t = 0.871, p = 0.38,
part correlation = 0.02), independently predicted the course of depressive symptoms (Table 2).
History of Depression and Personality Traits as
Moderators of Life EventsYSymptom Course
Relationship
To investigate whether the relationship between life events and
symptom course was moderated by the history of depressive disorders
or personality traits, six dummy variables representing the interaction
of history of depression, neuroticism, and extraversion with positive
and negative life events were added to Model 3. The resulting model
(Model 4) was subsequently reduced by removing nonsignificant
interaction terms (p G 0.10) one at a time, after which the model was
rerun. None of the two-way interaction terms made any further significant contribution to the prediction model.
Life Events as Mediators of the
ExtraversionYSymptom Course Relationship
The hypothesis that the effect of extraversion on symptom
course would be mediated by life events was tested following the
analytic steps outlined by Baron and Kenny (1986). First, it was investigated whether extraversion was predictive of the number of
positive or negative life events during the 1-year follow-up, controlling for demographic variables, history of depression and anxiety,
baseline depression severity, neuroticism, and the number of negative
or positive life events. Extraversion significantly predicted the number
* 2011 Lippincott Williams & Wilkins
of positive (but not negative) life events during the 1-year follow-up
(A = 0.144, t = 3.927, p G 0.001, part correlation = 0.14), with persons
high in extraversion reporting more positive life events.
Using a bootstrapping sampling procedure while controlling
for demographic variables, history of depression, history of anxiety,
baseline IDS scores, negative life events, and neuroticism, positive
life events proved to be a significant mediator between extraversion
and depressive symptoms (point estimate = j0.0114; bias-corrected
and accelerated 95% CI, j0.0246 to j0.0031). Extraversion has a
significant indirect effect on the severity of depressive symptoms at
1<year follow-up because highly extravert persons are more likely to
experience positive life events.
The hypothesis that the effect of history of depression on
symptom course is also mediated by life events was not further tested
because the number of positive and negative life events did not differ
between previously depressed subjects and controls. Moreover, although neuroticism was associated with the number of negative life
events (r = 0.13, p G 0.001), neuroticism was not predictive of symptom course and, consequently, was not further analyzed.
DISCUSSION
The primary purpose of the present study conducted in
euthymic previously depressed persons and healthy controls was
a) to examine the link between positive and negative life events and
the course of depressive symptoms, b) to evaluate the history of depression, neuroticism, and extraversion as moderators of this relationship, and c) to investigate whether life events mediate possible
relationships of history of depression and personality traits with
symptom course.
As hypothesized, negative life events during the 1-year
follow-up predicted the exacerbation of depressive symptoms over
and above the effects of age, sex, education, and baseline IDS scores
(Paykel, 2003), whereas positive life events had the opposite effect
(Neeleman et al., 2003). However, the size of the effects was small,
with part correlations varying between 0.10 and 0.30 (Cohen, 1988).
By measuring both negative and positive events in the same study, we
www.jonmd.com
Copyright © 2011 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
687
The Journal of Nervous and Mental Disease
Spinhoven et al.
were able to show that negative and positive life events independently predict the course of depressive symptoms in healthy controls
and in remitted persons. These results parallel those found in other
participants of the NESDA study with a current DSM-IV diagnosis
of depression or anxiety, in whom positive and negative life events
also both independently predicted symptom course (Spinhoven et al.,
in press).
Our second aim was to evaluate the history of depressive
disorders and personality traits as putative moderators of these relationships. Compared with healthy controls, euthymic remitted individuals were more likely to be younger and female and also manifested
risk factors for recurrence, such as higher levels of neuroticism and
lower levels of extraversion at baseline. As a consequence of selecting
euthymic remitted persons, both groups manifested comparable low
levels of depression severity making it unlikely that personality
measurements were confounded by current depressive symptoms.
In line with the notion that depression may be highly recurrent in
individuals with a history of depression, remitted persons reported
higher IDS scores at 1-year follow-up than did the controls. However,
no evidence was found that previously depressed persons in general
are more sensitive to negative or positive life events. With respect to
personality traits, only extraversion, not neuroticism, predicted prospective increases in depressive symptoms. Contrary to expectations,
no evidence that one of these traits strengthens the impact of positive
and negative life events on depressive symptoms was found. In particular, the strong correlation of neuroticism with severity of depressive symptoms (Jylhä et al., 2009), as was also found in the present
study (r = 0.60), makes it harder to establish an independent effect
of neuroticism on prognosis. Consequently, the size of the associations with symptom severity at 1-year follow-up was greatly reduced
after controlling for baseline severity and became statistically nonsignificant. We may have ‘‘overcontrolled’’ for the effects of neuroticism in our multivariate analyses, and the possibility that, in our study
group, higher depression scores are a direct consequence of neuroticism awaits further longitudinal studies. In contrast, extraversion
was only moderately related to the severity of depressive symptoms
and also predicted prospective decreases in depressive symptoms
after correcting for the effects of demographic and clinical variables.
These results suggest that higher levels of extraversion may exert
some protective effects against worsening of depressive symptoms
(Farmer et al., 2002).
Our hypothesis that personality moderated the impact of life
events on depressive symptoms received no support. We found that
neither neuroticism nor extraversion was a ‘‘positive life event amplifier’’ or ‘‘negative life event amplifier’’ regarding the course of
depressive symptoms. These findings are inconsistent with the results
of a limited number of previous studies reporting that highly neurotic patients with depression benefit more from positive life change
(Oldehinkel et al., 2000) or react more strongly to exposure to negative life events (Ormel et al., 1989) compared with patients with low
neuroticism. A possible reason for these discrepant results could be
that, in previous studies, finding support for neuroticism as a moderating variable, the incidence of life events was measured by a
semistructured interview scored by independent raters (Brown and
Harris, 1978). Replication studies are warranted using both life event
checklists and contextual measures of life event to assess whether
discrepant findings may depend on the method of assessing life
events.
Our third and last aim was to investigate to what extent history
of depression or personality factors influence whether someone is
more likely to experience (positive or negative) life events. Compared
with controls, remitted persons did not report a higher number of
negative and positive life events during the 1-year follow-up. These
results are inconsistent with the stress generation model, according
to which previously depressed persons are expected to produce more
688
www.jonmd.com
& Volume 199, Number 9, September 2011
stressful conditions for themselves than are controls (Hammen, 1991).
Therefore, we did not find evidence that the prospective value of
history of depression for the course of depression is mediated by
positive or negative life events. Personality traits, however, were
significantly related to the self-reported occurrence of life events.
Neurotic persons reported more negative life events (Magnus et al.,
1993), but neuroticism proved to be unrelated to the course of depressive symptoms after controlling for baseline severity. Moreover,
persons with high extraversion scores were more likely to experience
positive life events, and most importantly, the predictive value of
extraversion for the course of depressive symptoms seemed to be
partly caused by differential exposure to or, perhaps more appropriately labeled, engagement in positive life events (Zautra et al., 2005).
However, the size of the associations was relatively small. Overall, our
results are partly supportive of the life event mediation model, which
assumes that life events mediate the relationship of personality
dimensions with depressive symptoms.
Strengths of the present study include a) a prospective and
longitudinal design in a representative sample of participants with
a previous depressive disorder from different recruitment settings,
b) use of a structured diagnostic interview to assess diagnostic status
at baseline, and c) examination of healthy controls and euthymic
previously depressed individuals, making it unlikely that the differential effects of neuroticism and life events are confounded by depression severity.
Finally, three limitations of this study merit consideration:
a) the prevalence of life events was self-reported retrospectively, and
we did not verify whether persons did, in fact, experience these positive or negative life events. Therefore, the reports may have been
biased by mood state at 1-year follow-up, and future studies would
benefit from interview-based measures of life events. b) Part of the
association between life events and course of symptoms might have
arisen from life events being a consequence, rather than a cause, of
the course of symptoms. Because symptom levels, as measured
during the last week of the 1-year follow-up, may have already been
present at an earlier moment during the 1-year follow-up period,
reverse causality cannot be ruled out before the occurrence of positive and negative life events. c) There was no structured diagnostic
interview to assess the recurrence of depression during the 1-year
follow-up period.
CONCLUSIONS
Remitted persons reported higher IDS scores at 1-year followup than did controls. Extraversion and positive and negative life
events independently predicted the course of depressive symptoms.
The impact of life events on symptom course was not moderated by
the history of depression or personality traits. The effect of extraversion on symptom course was partly caused by differential engagement in positive life events. Acquiring strategies to actively
engage in positive life events and to cope more adequately with
negative life events may be indicated to prevent relapse of depression.
REFERENCES
American Psychiatric Association (2001) Diagnostic and Statistical Manual of
Mental Disorders (4th ed). Washington, DC: American Psychiatric Publishing.
Baron RM, Kenny DA (1986) The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations.
J Pers Soc Psychol. 51:1173Y1182.
Berlanga C, Heinze G, Torres M, Apiquian R, Caballero A (1999) Personality and
clinical predictors of recurrence of depression. Psychiatr Serv. 50:376Y380.
Brown GW, Harris TO (1978) Social Origins of Depression. A Study of Psychiatric
Disorder in Women. New York: Free Press.
Brugha T, Cragg D (1990) The list of threatening experiences: The reliability and
validity of a brief life events questionnaire. Acta Psychiatr Scand. 82:77Y81.
* 2011 Lippincott Williams & Wilkins
Copyright © 2011 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
The Journal of Nervous and Mental Disease
& Volume 199, Number 9, September 2011
Brugha T, Bebbington P, Tennant C, Hurry J (1985) The list of threatening
experiences: A subset of 12 life event categories with considerable long-term
contextual threat. Psychol Med. 15:189Y194.
Burcusa SL, Iacono WG (2007) Risk for recurrence in depression. Clin Psychol Rev.
27:959Y985.
Clark LA, Watson D, Mineka S (1994) Temperament, personality, and the mood
and anxiety disorders. J Abnorm Psychol. 103:103Y116.
Cohen J (1988) Statistical Power Analysis for the Behavioral Sciences (2nd ed).
New York: Lawrence Erlbaum Associates.
Costa PT, McCrae RR (1992) Revised NEO Personality Inventory (NEO-PR-I) and
the Five Factor Inventory (NEO-FFI): Professional manual. Odessa, FL:
Psychological Assessment Resources.
Enns MW, Cox BJ (1997) Personality dimensions and depression: Review and
commentary. Can J Psychiatry. 42:274Y284.
Fanous A, Gardner CO, Prescott CA, Cancro R, Kendler KS (2002) Neuroticism,
major depression, and gender: A population-based twin study. Psychol Med.
32:719Y728.
Farmer A, Redman K, Harris T, Mahmood A, Sadler S, Pickering A, McGuffin P
(2002) Neuroticism, extraversion, life events and depression: The Cardiff
Depression Study. Br J Psychiatry 181:118Y122.
Hammen C (1991) Generation of stress in the course of unipolar depression. J
Abnorm Psychol. 100:551Y561.
Headey B, Wearing A (1989) Personality, life events, and subjective well-being:
A dynamic equilibrium model. J Pers Soc Psychol. 57:731Y739.
Hettema JM, Prescott CA, Kendler KS (2004) Genetic and environmental sources
of covariation between generalized anxiety disorder and neuroticism. Am J
Psychiatry. 161:1581Y1587.
Hoekstra HA, Ormel J, De Fruyt F (1996) De NEO-PI-R/NEO-FFI; Big Five
Persoonlijkheidsvragenlijsten; Handleiding (Manual of the Dutch version of the
NEO- PI-R/NEO-FFI). Lisse, the Netherlands: Swets and Zeitlinger.
Jylhä P, Melartin T, Rytsälä H, Isometsä E (2009) Neuroticims, introversion, and
major depressive disorder-traits, states or scars. Depress Anxiety. 26:325Y334.
Kendler KS (2008) Explanatory models of psychiatric illness. Am J Psychiatry.
165:695Y702.
Kendler KS, Gardner CO, Prescott CA (2003) Personality and the experience of
environmental adversity. Psychol Med. 33:1193Y1202.
Kendler KS, Neale MC, Kessler RC, Heath AC, Eaves LJ (1993) A longitudinal
twin study of personality and major depression in women. Arch Gen Psychiatry.
93:853Y862.
Kupfer DJ, Frank E, Wamhoff J (1996) Mood disorders: Update on prevention of
recurrence. In Mundt C, MJ Goldstein (Eds), Interpersonal Factors in the
Origin and Course of Affective Disorders (pp 289Y302). London: Gaskell/Royal
College of Psychiatrists.
Lewinsohn PM, Allen NB, Seeley JR, Gotlib IH (1999) First onset versus
recurrence of depression: Differential processes of psychosocial risk. J Abnorm
Psychol. 108:483Y489.
Magnus K, Diener E, Fujita F, Pavot W (1993) Extroversion and neuroticism as
predictors of objective life eventsVa longitudinal study. J Pers Soc Psychol.
65:1046Y1053.
Monroe SM, Roberts JE, Kupfer JD, Frank E (1996) Life stress and treatment of
* 2011 Lippincott Williams & Wilkins
Personality, Life Events, Depression
recurrent depression: II. Postrecovery associations with attrition, symptom
course, and recurrence over 3 years. J Abnorm Psychol. 105:313Y328.
Needles DJ, Abramson LY (1990) Positive life events, attributional style, and hopefulness: Testing a model of recovery from depression. J Abnorm Psycho1. 99:156Y165.
Neeleman J, Oldehinkel AJ, Ormel J (2003) Positive life change and remission
of non-psychotic mental illness. A competing outcomes approach. J Affect
Disord. 76:69Y78.
Oldehinkel AJ, Ormel J, Neeleman J (2000) Predictors of time to remission from
depression in primary care patients: Do some people benefit more from positive
life change than others. J Abnorm Psychol. 109:299Y307.
Ormel J, Stewart R, Sanderman R (1989) Personality as modifier of the life changedistress relationship: A longitudinal modeling approach. Soc Psychiatry
Psychiatr Epidemiol. 24:187Y195.
Paykel ES (2003) Life events and affective disorders. Acta Psychiatr Scand.
108:61Y66.
Penninx BW, Beekman ATF, Smit JH, Zitman FG, Nolen WA, Spinhoven P,
Cuijpers P, de Jong PJ, van Marwijk HWJ, Assendelft WJJ, van der Meer K,
Verhaak P, Wensing M, de Graaf R, Hoogendijk WJ, Ormel J, van Dyck R
(2008) The Netherlands Study of Depression and Anxiety (NESDA): Rationale,
objectives, methods. Int J Methods Psychiatr Res. 17:121Y140.
Post RM (1992) Transduction of psychosocial stress into the neurobiology of
recurrent affective disorder. Am J Psychiatry. 149:999Y1010.
Preacher KJ, Hayes AF (2008) Asymptotic and resampling strategies for assessing
and comparing indirect effects in multiple mediator models. Behav Res
Methods. 40:879Y891.
Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH (1996) The Inventory of
Depressive Symptomatology (IDS): Psychometric properties. Psychol Med.
26:477Y486.
Spinhoven P, Roelofs K, Hovens JGFM, Elzinga BM, van Oppen P, Zitman FG,
Penninx BJWH (in press) Personality, life events and the course of anxiety and
depression. Eur J Pers.
Ter Smitten MH, Smeets RMW, Van den Brink W (1998) Composite International
Diagnostic Interview (CIDI), version 2.1, 12 months [in Dutch]. Amsterdam,
the Netherlands: World Health Organization.
Watson D, Gamez W, Simms LJ (2005) Basic dimensions of temperament and their
relation to anxiety and depression: A symptom-based perspective. J Res Pers.
39:46Y66.
Viken RJ, Rose RJ, Kaprio J, Koskenvuo M (1994) A developmental genetic
analysis of adult personality: Extraversion and neuroticism from 18 to 59 years
of age. J Pers Soc Psychol. 66:722Y730.
Wittchen HU, Burke JD, Semler G, Pfister H, Von CM, Zaudig M (1989) Recall and
dating of psychiatric symptoms. Test-retest reliability of time-related symptom
questions in a standardized psychiatric interview. Arch Gen Psychiatry.
46:437Y443.
Wittchen HU, Robins LN, Cottler LB, Sartorius N, Burke JD, Regier D (1991)
Cross-cultural feasibility, reliability and sources of variance of the Composite
International Diagnostic Interview (CIDI). The Multicentre WHO/ADAMHA
Field Trials. Br J Psychiatry. 159:645Y653.
Zautra AJ, Affleck GG, Tennen H, Reich JW, Davis MC (2005) Dynamic
approaches to emotions and stress in everyday life: Bolger and Zuckerman
reloaded with positive as well as negative affects. J Pers. 73:1Y28.
www.jonmd.com
Copyright © 2011 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
689