Soc Psychiatry Psychiatr Epidemiol (2003) 38 : 262–268
1007/s00127-003-0633-4
ORIGINAL PAPER
I. Myin-Germeys · L. Krabbendam · P. Delespaul · J. van Os
Can cognitive deficits explain differential sensitivity
to life events in psychosis?
Accepted: 3 December 2002
■ Abstract Background Life events (LE) have been
found to influence the onset and course of psychotic dis-
orders. It has been suggested that LE have their effect by
increasing underlying sensitivity to daily life stress, a re-
ported vulnerability marker for psychosis. As increased
stress-sensitivity and cognitive impairments have been
shown to be negatively associated with each other in pa-
tient populations, it is attractive to hypothesise that the
impact of LE on sensitivity to daily stress is modified by
the degree of cognitive impairment, higher levels of cog-
nitive impairment giving rise to reduced impact of LE
on daily life stress-sensitivity. Methods Patients with
psychotic illness (n = 42) were studied with a) a stan-
dard battery of neuropsychological tests to assess cogni-
tive functioning, b) the Experience Sampling Method (a
structured diary technique assessing current context
and mood in daily life) to assess (i) appraised subjective
stress related to daily events and activities, and (ii) emo-
tional reactivity conceptualised as changes in both neg-
I. Myin-Germeys, MA, PhD · L. Krabbendam, MA, PhD ·
P. Delespaul, MA, PhD · J. van Os, MD, PhD, MRCPsych
Dep. of Psychiatry and Neuropsychology
azM/Mondriaan /RIBW/Riagg/Vijverdal Academic Centre
Maastricht University
Maastricht, The Netherlands
I. Myin-Germeys, MA, PhD
Mondriaan Zorggroep
Section Social Cognition
P.O. BOX 4436, 6401 CX Heerlen, The Netherlands
J. van Os, MD, PhD, MRCPsych
Division of Psychological Medicine
Institute of Psychiatry
De Crespigny Park
Denmark Hill, London SE5 8AF, UK
Dr. I. Myin-Germeys ()
Dep. of Psychiatry and Neuropsychology
Maastricht University
P.O. BOX 616 (PAR 45)
6200 MD Maastricht, The Netherlands
SPPE 633
Tel.: +31-43/3 29 97 73
Fax: +31-43/3 29 97 08
E-Mail: i.germeys@sp.unimaas.nl
DOI 10.
ative affect (NA) and positive affect (PA), and c) the
Brown and Harris Life Event and Difficulties Schedule to
assess LE over the last year. Results Multilevel regression
analyses showed that a prior history of LE increased the
sensitivity for daily life stress in subjects with the best
performance on the cognitive tests. Conclusions Subjects
with cognitive impairments, who already were shown to
have lower levels of sensitivity to daily life stress (Myin
Germeys et al. 2002), may additionally be less reactive to
prior exposure to LE. This result fits with the notion of
separate affective and cognitive pathways of symptom
formation in psychosis, the cognitive pathway being
characterised by severe cognitive deficits and the affec-
tive pathway by increased levels of stress-sensitivity as-
sociated with prior exposure to LE.
■ Key words psychosis – cognition – stress – life events
Introduction
Life events (LE) have been found to influence the onset
and course of psychotic disorders. Increased numbers of
LE have been associated with higher levels of sympto-
matology and increased relapse rates (Bebbington et al.
1993; Bebbington et al. 1996; Lukoff et al. 1984; Mik-
lowitz 1994; Norman and Malla 1993). However, the
mechanism by which LE influence the occurrence and
course of psychotic disorders is not known. Some au-
thors have suggested that LE do not trigger relapses di-
rectly, but rather act by increasing vulnerability cumu-
latively with successive exposures (Bebbington et al.
1996; Hirsch et al. 1996).
The underlying mechanism by which LE act cumula-
tively to the risk function may reside in their impact on
sensitivity for daily life stress. It has been shown in the
general population that life events affect emotional re-
activity to smaller daily events (Kanner et al. 1981; Eck-
enrode 1984). Furthermore, these smaller daily events
have been reported to be important predictors of psy-
chological symptoms in general (Kanner et al. 1981;

Monroe 1983), of subjective distress (Norman and Malla
1991), and of relapse rates in schizophrenia (Malla et al.
1990). Recent work reported that prior exposure to LE in
patients with schizophrenia did not affect the appraised
stressfulness of daily events. However, prior exposure to
LE did modify the emotional reaction to daily life stress,
both in terms of increases in negative affect and de-
creases in positive affect (Myin-Germeys et al. in press).
The apparent discordance between effects of LE on sub-
jective appraisal of stress and emotional reactivity can
be understood in terms of underlying vulnerability. It
has been reported that relatives of patients with psy-
chosis are not different from controls in the average level
of appraised stress, nor in average mood level (Myin-
Germeys et al. 2001b). However, they did report a signif-
icantly larger emotional reactivity compared to con-
trols. Therefore, it was argued that emotional reactivity
constitutes an underlying vulnerability marker for psy-
chosis. It appears that LE may impact specifically on this
underlying vulnerability.
Previous work has shown that the two domains of
vulnerability of cognitive impairment and sensitivity to
daily life stress have non-overlapping distributions in
patient populations: individuals with greater levels of
cognitive impairment have lower levels of sensitivity to
daily life stress (Myin-Germeys et al. 2002). Therefore, if
LE impact on sensitivity to daily life stress, it is attrac-
tive to hypothesise that this effect of LE is modified by
the degree of cognitive impairment, higher levels of cog-
nitive impairment giving rise to a reduced impact of LE
on daily life stress-sensitivity. The present study will in-
vestigate the effect of LE on emotional reactivity to daily
life stress in relation to cognitive performance in 42 pa-
tients diagnosed with psychotic disorder.
Subjects and methods
■ Subjects
The sample consisted of 50 psychotic subjects. All patients were un-
der current treatment. Selection criteria, assessed by a research physi-
cian or research psychologist, were a lifetime occurrence of psychotic
symptoms according to the RDC (Research Diagnostic Criteria) for at
least 2 weeks in clear consciousness. Inclusion criteria were: 1) age
18–55 years, 2) sufficient command of the Dutch language, and 3) nor-
mal physical examination. Exclusion criteria were: 1) endocrine, car-
diovascular, or brain disease, 2) use of alcohol in excess of five stan-
dard units per day, 3) weekly use of illicit drugs, 4) history of
head-injury with loss of consciousness, and 5) being in need of in-pa-
tient care, intensive case management home care or crisis interven-
tion. Written informed consent, conforming to the local ethics com-
mittee guidelines, was obtained from all subjects. Patients were
recruited through the ambulatory mental health facilities in Maas-
tricht, The Netherlands, and through patient associations in the
southern part of the Netherlands.
The diagnostic procedure included extensive screening with di-
agnostic interviews that included the Life Chart (Susser et al. 2000),
the Brief Psychiatric Rating Scale (Ventura et al. 1993), and the Posi-
tive and Negative Syndrome Scale (Kay et al. 1987) to map psychiatric
symptomatology. Interview data and clinical record data were used to
complete the Operational Criteria Checklist for Psychotic Illness
(OCCPI) yielding DSM-III-R diagnoses through the OPCRIT com-
puter program (McGuffin et al. 1991).
263
■ Experience Sampling Method (ESM)
The Experience Sampling Method is a within-day self-assessment
technique. Previous applications of ESM in schizophrenia (Delespaul
et al. 2002; Myin-Germeys et al. 2000; Myin-Germeys et al. 2001a;
Myin-Germeys et al. 2001b) have demonstrated the feasibility, valid-
ity, and reliability of the method in this population. Subjects were
studied in their normal daily living environment. They received a dig-
ital wristwatch and a set of ESM self-assessment forms collated in a
booklet for each day. Ten times a day on six consecutive days, the
watch emitted a signal (beep) at unpredictable moments between 7.30
a. m. and 10.30 p. m. After every “beep”, subjects were asked to stop
their activity and fill out the ESM self-assessment forms previously
handed to them, collecting reports of thoughts, current context (ac-
tivity, persons present, location), appraisals of the current situation,
and mood. All self-assessments were rated on 7-point Likert scales.
The ESM procedure was explained to the subjects during an ini-
tial briefing session and a practice form was completed to confirm
that subjects were able to understand the 7-point Likert scale format.
Subjects were instructed to complete their reports immediately after
the beep, thus minimising memory distortions, and to record the time
at which they completed the form. During the actual sampling period,
research staff repeatedly called the subjects to assess whether they
were complying with the instructions. In order to know whether the
subjects had completed the form within 15 minutes of the beep, the
time at which subjects indicated they completed the report was com-
pared to the actual time of the beep. All reports completed more than
15 minutes after the signal were excluded from the analysis. Previous
work has shown that reports completed after this interval are less re-
liable and consequently less valid (Delespaul 1995). Subjects with less
than 20 valid reports were excluded from the analysis.
■ Emotional stress reactivity assessment
Previously, emotional stress reactivity was conceptualised as mood in
reaction to daily events and minor disturbances in daily life (Myin-
Germeys et al. 2001b). Both the mood measures and the stress mea-
sures were derived from the experience sampling reports as described
below.
Assessment of mood
Mood states reported after each beep were assessed with ten mood
adjectives rated on 7-point Likert scales (1 not at all to 7 very). Factor
analyses (principal component analysis with Harris-Kaiser rotation)
on the raw within-subject scores identified two factors with eigenval-
ues greater than 1 explaining 41 % of the total variance. Two factor-
based scales with equal weights for each item were created. The items
down, guilty, insecure, lonely, and anxious formed the NA scale (Neg-
ative Affect), (Cronbach’s α = 0.79). The items happy, cheerful, relaxed,
and satisfied formed the PA scale (Positive Affect) (Cronbach’s
α = 0.89). The item angry had low loadings on both factors and was
excluded to enhance differentiation between the two factors.
Assessment of stress
Stress was conceptualised as subjective appraised stressfulness of dis-
tinctive events as well as of minor disturbances that continually hap-
pen in the natural flow of daily life. These were:
1. Event-related stress:
after each beep, subjects were asked to report the most important
event that happened between the current and the previous report.
This event was subsequently rated on a 7-point bipolar scale
(–3 = very unpleasant, 0 = neutral, 3 = very pleasant). Responses
were recoded to allow high scores to reflect stress (–3 = very pleas-
ant, 0 = neutral, 3 = very unpleasant). Response on this item is
called event-related stress.
2. Activity-related stress:
after each beep, subjects judged their current activity on three
self-report items (scored on 7-point Likert scales, 1 = not at all and
7 = very). The mean of the scales ‘I am not skilled to do this activ-
264
ity‘, ‘I would rather do something else’, and ‘This activity requires
effort’ forms the activity-related stress scale (α = 0.54).
■ Assessment of life events
All subjects were interviewed with the structured Brown and Harris
Life Events and Difficulties Schedule (LEDS) (Brown and Harris
1978). The LEDS is a structured interview, assessing serious life events
in several domains of life such as health, marriage, work, etc. Inter-
views were conducted by a research psychologist (I. M.-G.) or a
trained research nurse. Each interview was audiotaped and subse-
quently scored by the two researchers based on consensus. Each in-
terview investigated the presence of life events over the past year (1
year back from the date of the interview), irrespective of the occur-
rence of symptoms.
In the present study, life events (LE) with a moderate amount of
threat or unpleasantness were included (scores of 1 = marked
threat/unpleasantness, 2 = moderate threat/unpleasantness; accord-
ing to the Brown and Harris scale). The total number of such LE was
used as a continuous variable in the analyses, as a measure of the
amount of social adversity befalling the person (Van Os and Jones
1999).
■ Neuropsychological assessments
The neuropsychological assessment was directed at the following
cognitive domains: episodic memory, semantic fluency, attentional
span and speed of complex information processing. The Auditory
Verbal Learning Task (AVLT) (Brand and Jolles 1985; Lezak 1995) was
used to evaluate learning and retrieval of information in episodic
memory. In five consecutive trials, a list of 15 words has to be memo-
rised and reproduced. The measure used was the total number of
words recalled over the five trials. The backward digit span from the
Wechsler Adult Intelligence Scale-Revised (Wechsler 1981) was used
as a measure of attentional span (Lezak 1995). The Stroop Colour-
Word Test (SCWT) (Stroop 1935) and the Concept Shifting Test (CST)
(Houx et al. 1991), which is a modified version of the Trailmaking Test
(Reitan 1958), were used as measures of speed of information pro-
cessing. For both tests, the measure used was the interference score
which is the difference between performance on the complex task and
performance on the simple task. For the Stroop Colour-Word Test,
this means the time needed for the subject to name the print colour
while ignoring the word minus the time needed for reading the colour
names.For the Concept Shifting Test,the interference score is the time
needed for the subject to switch between numbers and letters minus
the number-only condition. Word fluency was used to evaluate strat-
egy-driven retrieval from semantic memory. The subjects had to gen-
erate as many animal names as possible in one minute (Lezak 1995).
For all neuropsychological measures, a z-score was calculated
(subtracting the mean and dividing by the standard deviation) and
the sum of these five z-scores was used in the analyses.
■ Statistical analyses
A multilevel linear random regression model (Goldstein 1987) was
used. Multilevel or hierarchical linear modelling techniques are a
variant of the more often used unilevel linear regression analyses and
are ideally suited for the analysis of ESM data, consisting of multiple
observations within one person, i. e. at two levels (ESM-beep level and
subject level) (Schwartz and Stone 1998). Since, in ESM, observations
from the same subject are more similar than observations from dif-
ferent subjects, the residuals are not independent. Conventional re-
gression techniques do not adequately take into account the variance
components at two different levels.
Data were analysed with the XTREG module in STATA (StataCorp
2001). The β is the fixed regression coefficient of the predictor in the
multilevel model and can be interpreted identically to the estimate in
a unilevel linear regression analysis.
To test the hypothesis that the occurrence of LE modified the
emotional reaction to daily life stress more in patients with better
cognitive performance, multilevel linear regression analyses were
conducted with NA and PA as the dependent variables. LE, the two
stress measures, the cognitive score, as well as their interactions were
the independent variables: mood = B0 + B1 daily stress + B2 LE + B3
cognition + B4 daily stress*LE + B5 daily stress*cognition + B6
LE*cognition + B7 daily stress*LE*cognition + residuals. As is re-
quired in a regression analysis testing the effect of higher-order in-
teraction terms, all lower-order interaction terms were included in
the model as well. Analyses were conducted separately for each mea-
sure of daily life stress.
The interaction term (daily stress*LE*cognition) was of most in-
terest in the present study as the main hypothesis revolved around the
question whether cognitive functioning modifies the interaction ef-
fect of LE and daily life stress on mood. Therefore, stratified analyses
were conducted whenever there was evidence of significant interac-
tion effects. To this end, the subjects were divided into three groups
according to their tertile group level of functioning on the neuropsy-
chological tests: the worst-performance group (the approximately
33 % of the subjects with the lowest overall score on the tests), the in-
termediate performance group (the 33 % of the subjects whose scores
on the tests fell into the middle tertile), and the best performance
group (the 33 % of the subjects with the highest scores on the tests).
In each of the three groups, it was investigated how LE modified the
emotional reactivity to daily life stress according to the following
model: mood = B0 + B1 daily stress + B2 LE + B3 daily stress*LE.
Results
Of the 50 subjects who entered the study, 8 subjects were
excluded from the analysis due to insufficient ESM data.
Two patients did not return the diary booklets and 6 pa-
tients were unable to comply with the research protocol
(they had fewer than 20 valid reports and were, there-
fore, excluded from the analyses, see ESM). The final
study sample thus consisted of 42 subjects (Table 1).
Sociodemographic and clinical characteristics of the
patients are summarised in Table 1, and the mean scores
on the independent and dependent variables are shown
in Table 2.
The multilevel random regression analyses showed a
significant main effect of the two stress measures on NA
[activity-related stress: B = 0.16 (SE = 0.02), p = 0.0001;
event-related stress: B = 0.08 (SE = 0.01), p = 0.0001] and
on PA [activity-related stress: B = –0.28 (SE = 0.02),
p = 0.0001; event-related stress: B = –0.14 (SE = 0.01),
p = 0.0001]. No significant main effect of LE and cogni-
tive functioning on mood was found (results not
shown). Significant three-way interactions were found
in three of the four models (Table 3), indicating that cog-
nitive performance modified the daily stress by LE in-
teraction effect on mood.
Stratified analyses were conducted to further clarify
the differential interaction effect between LE and daily
stress on mood in the group with worst, with intermedi-
ate, and with best cognitive performance (Table 4).
Overall, LE did modify the effect of daily stress on mood
in the group with the best performance on the cognitive
tests. In the groups with intermediate and worst perfor-
mance on the cognitive tests, no significant interaction
effect was found between daily stress and LE on mood.
 265

Table 1 Sociodemographic and clinical characteristics of the research sample Table 3 The interaction effect of daily stress, life events and cognitive perfor-
mance on mood
Psychotic subjects
(n = 42) B (SE) z P

Sociodemographic variables NA
Activity-related stress * LE * cognition –0.05 (0.01) –6.44 0.000
Age 31.9 (7.7); (range 20–48) Event-related stress * LE * cognition –0.02 (0.01) –3.10 0.002
Sex (M–F) 22–20 PA
Education Activity-related stress * LE * cognition 0.03 (0.01) 3.16 0.002
Elementary school 24% Event-related stress * LE * cognition 0.00 (0.01) 0.04 0.970
Secondary school 67%
Higher education 9%
Marital status
Married or living together 21% Discussion
Divorced 5%
Never married 74%
In terms of impacting on the underlying stress-sensitiv-
Work situation ity, life events may particularly influence patients with
Working 24%
Unemployed 0% preserved cognitive functioning. The altered stress-sen-
Incapable of work 66% sitivity that characterises patients with higher levels of
Protected work 10% cognitive functioning is also more reactive to a history
of prior LE, while patients with cognitive impairments,
Clinical variables
who already have lower levels of sensitivity to daily life
OPCRIT DSM-III-R diagnosis (lifetime) stress (Myin-Germeys et al. 2002), may additionally be
Schizophrenia 39 less reactive to prior exposure to LE.
Schizo-affective disorder 2 These results fit with the notion of separate affective
Atypical psychosis 1
and cognitive pathways of symptom formation in psy-
Total BPRS score 38 (9.8); range (24–73) chosis (Myin-Germeys et al. 2002), the affective pathway
Age of first psychotic episode 22.5 (5.8); (range 14–41) characterised by increased levels of stress-sensitivity
Medication status (psychotropics) and the cognitive pathway characterised by severe cog-
Typical antipsychotics 50%
Atypical antipsychotics 45%
nitive deficits. These pathways may contribute to the
Antidepressants 19% mechanisms underlying the extensive clinical hetero-
Benzodiazepines 24% geneity in schizophrenia which several authors (Car-
Lithium 5% penter et al. 1988; Crow 1980; Murray et al. 1992; Robins
Anticholinergics 14% and Guze 1970; Van Os et al. 1998) have suggested can be
No medication 5%
reduced to two main forms: a more chronic form of ill-

Table 2 Ratings and correlations of the number of

valid reports and the independent and dependent Score Correlation (r)
variables Mean SD Range 1 2 3 4

Valid reports 45 10 21–60

Life event (LE) variables
1. # LE 1.1 0.2 0–4
Neuropsychological measures
1. Word learning 48.6 8.2 29–64
2. Word fluency 21.7 7.0 8–37 0.31***
3. Stroop Colour-Word 51.1 19.1 23–96 0.18*** 0.32***
Test
4. Concept Shifting 12.7 10.4 0.4–56.2 0.26*** 0.06* 0.41***
Test
5. WAIS-R backward 5.8 1.6 3–11 0.29*** 0.09*** 0.07** 0.31***
digit span
6. Sum of five z-scores 0 3.03 –4.9–9.6
Stress-related variablesa
1. Event –1.2 0.9 –3–3 1
2. Activity 2.5 0.7 1 to 7 0.17*** 1
Mood statesa, b
1. NA 1.7 0.7 1–7 1
2. PA 4.4 1.0 1–7 –0.43*** 1
a For each subject, a mean was calculated over all reports, and these means were aggregated over the group to

obtain the group mean (SD); b NA = Negative Affect; PA = Positive Affect

* p < 0.05; ** p < 0.01; *** p < 0.001
266
Table 4 The interaction effect of daily stress and LE
on mood, stratified by performance on the cognitive
tasks
NA
Activity-related stress * LE
Event-related stress * LE
PA
Activity-related stress* LE
* p < 0.05; *** p < 0.001
ness characterised by high levels of negative symptoms
and neurocognitive impairment which may be associ-
ated with the cognitive pathway, and an episodic, reac-
tive, good-outcome form with higher levels of positive
symptoms associated with the affective pathway. The
latter has been postulated to be most reactive to envi-
ronmental stress, which may be mediated by underlying
stress-sensitivity. Previous LE research has already sug-
gested that LE are related in particular to positive symp-
toms, better outcomes (e. g. less need of anti-psychotic
maintenance medication and more time in complete re-
mission) and a more episodic type of illness (Bebbing-
ton et al. 1996; Hirsch et al. 1996; Miller et al. 2001; van
Os et al. 1994; Van Os et al. 1998).
The notion of separate affective and cognitive path-
ways has previously been mentioned in the literature by
Garety and colleagues (Garety et al. 2001), although
some differences become apparent. Garety et al. (2001)
hypothesise a cognitive-affective and an affective path-
way to the formation of positive psychotic symptoms,
while the current model suggests that the cognitive
pathway is related to the formation of negative symp-
toms. However, both views might be reconcilable. Firstly,
both the current model and Garety et al.’s model stress
the importance of affective disturbances in the forma-
tion of positive symptoms. Secondly, although the inclu-
sion of cognitive disturbances in the formation of posi-
tive symptoms seems contradictory to the affective and
cognitive pathways as suggested in the current paper,
this may not be the case. In the current model, the cog-
nitive pathway is characterised by a general impairment
in cognitive functioning, which is clearly trait-related.
Garety et al., on the other hand, focus on specific cogni-
tive disturbances such as reasoning biases and problems
in self-monitoring of actions and intentions. These cog-
nitive problems can be conceived as specific and state-
related as they become apparent due to triggering
events. Therefore, the affective pathway, as described in
the current model, could actually contain both the affec-
tive and the cognitive-affective pathway as described by
Garety et al. (2001).
The data suggest that LE engender their effect by im-
pacting on the underlying stress-sensitivity, especially
in the group with the best cognitive performance. Al-
though LE might impact through other mechanisms as
well, additional evidence points in the direction of al-
tered stress-sensitivity. Several authors have suggested
a form of continuity between affective and psychotic
Best performing Middle performing Worst performing
group group group
0.21 (0.03)*** –0.03 (0.04) –0.02 (0.02)
0.09 (0.02)*** 0.04 (0.04) 0.00 (0.02)
–0.09 (0.04)* –0.02 (0.05) 0.02 (0.04)
disorders, partly resulting from areas of shared vulner-
ability (Kendler et al. 1998; Maier et al. 1993; van Os et
al. 1999). Social adversities such as life events may con-
stitute a shared risk factor for both the affective and
psychotic disorders, with the largest effect on the affec-
tive end of the continuum (van Os et al. 1999), their ef-
fects being most pronounced in the affective disorders
and the affective psychoses (Bebbington et al. 1993;
Dohrenwend et al. 1995; Paykel 1978; Van Os et al. 1998).
If LE indeed impact in part through their effect on
raising underlying stress-sensitivity, one would expect
increased levels of stress-sensitivity in the affective
disorders as well. Recently, Myin-Germeys et al. (Myin-
Germeys et al. in press b) demonstrated that altered
stress-sensitivity is a general underlying vulnerability
shared by patients with psychotic and affective disor-
ders. This adds to the suggestion that altered stress-sen-
sitivity is an area of shared liability between affective
and psychotic disorders. Therefore, it is plausible that
one of the mechanisms through which LE influence the
onset and course of disorders is by impacting on the un-
derlying stress-sensitivity, irrespective of diagnostic
context.
One could argue that the results are an artefact of the
fact that patients with better cognitive performance are
more likely to experience LE, as they would be more ca-
pable of fulfilling some role in society such as having a
job. However, a post-hoc analysis showed that cognitive
performance was not associated with the number of LE
(B = –0.04, SE = 0.03, p = 0.29). Alternatively, it could be
asserted that LE impact on mood levels rather than on
stress-sensitivity and that mood levels may be expected
to be different in the three groups (for example, because
subjects with better cognitive functioning are more
aware of the devastating effect of their illness). However,
no main effect of cognition was found on mood, indi-
cating that mood is not mediated by level of cognitive
performance and no main effect of LE on mood was
found, indicating that LE impact on stress-reactivity
rather than on mood levels per se.
The differential effect in models predicting NA and
PA is in line with reports in the literature showing that
NA and PA are two independent factors, with specific
characteristics and daily patterns (Goldstein and Strube
1994; Watson et al. 1999). NA has been shown to be more
influenced by the occurrence of life events (Watson et al.
1999). Thus, negative mood levels are generally quite
low, sharply increasing when a serious LE befalls the in-

dividual. Positive affect, on the other hand, is a much
more stable state, with higher levels of intensity and less
reactive to LE.
■ Methodological issues
The present results should be viewed in the light of sev-
eral methodological issues. Firstly, most LE research dis-
tinguished independent LE and LE that are possibly re-
lated to the illness, the former being the most predictive
for psychotic relapse. In the present research, no such
distinction was made. The rationale for this was that life
events, even if they are “dependent” in some way on ill-
ness, would still have a psychological impact on the in-
dividual, so that excluding them may underrate life
stress experienced by patients. Secondly, we focused on
LE that possibly had the capacity to impact on the un-
derlying vulnerability. Therefore, LE with short-term
threat and with only mild long-term threat were not in-
cluded in the analyses. In a post-hoc analysis, we added
the LE with mild long-term threat. This resulted in
smaller but still significant effect sizes, except for the
event-related stress*LE*cognition interaction effect on
PA which became marginally significant in the opposite
direction. In addition, no long-term difficulties were in-
cluded in the analyses, although they also contribute to
the amount of social adversity befalling the person. The
reason for leaving these out,however,was that long-term
difficulties might impact on daily stress-sensitivity in
different ways to LE, e. g. when the difficulty is still going
on, it will directly be reflected in the ESM measures,
rather than impacting on the underlying vulnerability.
Thirdly, the measurements of daily stress and mood
are based on subjective reports. Although subjective re-
ports are considered less reliable (e. g. do all subjects in-
terpret or answer the questions identically?), they can be
valid, whereas the validity of objective approaches
should not be taken for granted (Strauss 1994).
Fourthly, emotional stress reactivity has been de-
fined in terms of emotional reactivity towards subjec-
tive stress. The cross-sectional analyses of the data, how-
ever, make it impossible to establish a causal
relationship. Therefore, the reverse might be true, in that
a worse mood might influence the subjective appraisal
of the environment. The overall effect, however, would
still be for the individual to experience distress associ-
ated with an environmental event.
References
1. Bebbington P, Wilkins S, Jones P, Foerster A, Murray R, Toone B,
Lewis S (1993) Life events and psychosis. Initial results from the
Camberwell collaborative psychosis study. Br J Psychiatry
162:72–79
2. Bebbington P, Wilkins S, Sham P, Jones P, van Os J, Murray R,
Toone B, Lewis S (1996) Life events before psychotic episodes: do
clinical and social variables affect the relationship? Soc Psychia-
try Psychiatr Epidemiol 31:122–128
267
3. Brand N, Jolles J (1985) Learning and retrieval rate of words pre-
sented auditorily and visually. J Gen Psychology 112:201–210
4. Brown GW, Harris TO (1978) Social origins of depression: a
study of psychiatric disorder in women. Tavistock, London
5. Carpenter WT,Heinrichs DW,Wagman A (1988) Deficit and non-
deficit forms of schizophrenia: the concept. Am J Psychiatry
145:578–583
6. Crow TJ (1980) Molecular pathology of schizophrenia: more
than one disease process? BMJ 280:66–68
7. Delespaul P, deVries M, van Os J (2002) Determinants of occur-
rence and recovery from hallucinations in daily life. Soc Psychi-
atry Psychiatr Epidemiol 37:97–104
8. Delespaul PAEG (1995) Assessing schizophrenia in daily life.
Universitaire Pers Maastricht, Maastricht
9. Dohrenwend BP, Shrout PE, Link BG, Skodol AE, Stueve A (1995)
Life events and other possible psychosocial risk factors for
episodes of schizophrenia and major depression: a case-control
study. In: Mazure CM (ed) Does stress cause psychiatric illness?
Progress in psychiatry, no. 46.American Psychiatric Press,Wash-
ington, pp. 43–65
10. Eckenrode J (1984) Impact of chronic and acute stressors on
daily reports of mood. J Pers Soc Psychol 46:907–918
11. Garety PA, Kuipers E, Fowler D, Freeman D, Bebbington PE
(2001) A cognitive model of the positive symptoms of psychosis.
Psychol Med 31:189–195
12. Goldstein H (1987) Multilevel models in educational and social
research. Charles Griffin & Co, London
13. Goldstein MD, Strube MJ (1994) Independence revisited: the re-
lation between positive and negative affect in a naturalistic set-
ting. Pers Soc Psychol Bull 20:57–64
14. Hirsch S, Bowen J, Emami J, Cramer P, Jolley A, Haw C, Dickinson
M (1996) A one-year prospective study of the effect of life events
and medication in the aetiology of schizophrenic relapse. Br J
Psychiatry 168:49–56
15. Houx P, Vreeling F, Jolles J (1991) Age-associated cognitive de-
cline is related to biological life-events. In: Iqbal K, McLachlan D,
Winblad B, Wisniewski H (eds) Alzheimer’s disease: basic mech-
anisms, diagnosis and therapeutic strategies. John Wiley & Sons
Ltd, Chichester, pp. 353–359
16. Kanner AD, Coyne JC, Schaefer C, Lazarus RS (1981) Compari-
son of two models of stress measurement: daily hassles and up-
lifts versus major life events. J Behav Med 4:1–39
17. Kay SR, Fiszbein A, Opler LA (1987) The positive and negative
syndrome scale (PANNS) for schizophrenia. Schizophr Bull
13:261–276
18. Kendler KS, Karkowski LM,Walsh D (1998) The structure of psy-
chosis: Latent class analysis of probands from the Roscommon
family study. Arch Gen Psychiatry 55:492–499
19. Lezak MD (1995) Neuropsychological assessment. Oxford Uni-
versity Press, New York, pp. 357–368
20. Lukoff D, Snyder K, Ventura J, Nuechterlein KH (1984) Life
events, familial stress, and coping in the developmental course of
schizophrenia. Schizophr Bull 10:258–292
21. Maier W, Lichtermann D, Minges J, Hallmayer J, Heun R, Benkert
O, Levinson DF (1993) Continuity and discontinuity of affective
disorders and schizophrenia. Results of a controlled family
study. Arch Gen Psychiatry 50:871–883
22. Malla AK, Cortese L, Shaw TS, Ginsberg B (1990) Life events and
relapse in schizophrenia. Soc Psychiatry Psychiatr Epidemiol
25:221–224
23. McGuffin P, Farmer A, Harvey I (1991) A polydiagnostic applica-
tion of operational criteria in studies of psychotic illness: devel-
opment and reliability of the opcrit system. Arch Gen Psychiatry
48:764–770
24. Miller P, Lawrie SM, Hodges A, Clafferty R, Cosway R, Johnstone
EC (2001) Genetic liability, illicit drug use, life stress and psy-
chotic symptoms: preliminary findings from the Edinburgh
study of people at high risk for schizophrenia. Soc Psychiatry
Psychiatr Epidemiol 36:338–342
25. Monroe SM (1983) Major and minor life events as predictors of
psychological distress: further issues and findings. J Behav Med
6:189–205
268
26. Murray RM, O’Callaghan E, Castle DJ, Lewis SW (1992) A neu-
rodevelopmental approach to the classification of schizophre-
nia. Schizophr Bull 18:319–332
27. Myin-Germeys I, Delespaul PA, deVries MW (2000) Schizophre-
nia patients are more emotionally active than is assumed based
on their behavior. Schizophr Bull 26:847–854
28. Myin-Germeys I,Nicolson NA,Delespaul PAEG (2001a) The con-
text of delusional experiences in the daily life of schizophrenia
patients. Psychol Med 31:489–498
29. Myin-Germeys I, van Os J, Schwartz JE, Stone AA, Delespaul PA
(2001b) Emotional reactivity to daily life stress in psychosis.
Arch Gen Psychiatry 58:1137–1144
30. Myin-Germeys I, Krabbendam L, Jolles J, Delespaul PA, van Os J
(2002) Are cognitive impairments associated with sensitivity to
stress in schizophrenia? An experience sampling study.Am J Psy-
chiatry 159:443–449
31. Myin-Germeys I, Krabbendam L, Delespaul P, van Os J (in press)
Do life events have their effect on psychosis by influencing the
emotional reactivity to daily life stress? Psychol Med
32. Myin-Germeys I, Peeters FPML, Havermans R, Nicolson NA, De-
lespaul P, deVries M, van Os J (2003) Emotional reactivity to daily
life stress in psychosis and affective disorder: an experience
sampling study. Acta Psychiatr Scand 107:124–131
33. Norman RMG, Malla AK (1991) Subjective stress in schizo-
phrenic patients. Soc Psychiatry Psychiatr Epidemiol 26:212–216
34. Norman RMG, Malla AK (1993) Stressful life events and schizo-
phrenia: a review of the research. Br J Psychiatry 162:161–166
35. Paykel ES (1978) Contribution of life events to causation of psy-
chiatric illness. Psychol Med 8:245–253
36. Reitan RM (1958) Validity of the trail making test as an indica-
tor of organic brain damage. Percept Mot Skills 8:271–276
37. Robins E, Guze SB (1970) Establishment of diagnostic validity in
psychiatric illness: its application to schizophrenia. Am J Psychi-
atry 126:983–987
38. Schwartz JE, Stone AA (1998) Strategies for analyzing ecological
momentary assessment data. Health Psychol 17:6–16
39. StataCorp (2001) Stata statistical software: Release 7.0. Texas,
College Station
40. Strauss JS (1994) The person with schizophrenia as a person. I.
Approaches to the subjective and complex. Br J Psychiatry Suppl:
103–107
41. Stroop JR (1935) Studies of interference in serial verbal reac-
tions. J Exp Psychol 18:643–662
42. Susser E, Finnerty M, Mojtabai R, Yale S, Conover S, Goetz R,
Amador X (2000) Reliability of the life chart schedule for assess-
ment of the long-term course of schizophrenia. Schizophr Res
42:67–77
43. van Os J, Fahy TA, Bebbington P, Jones P, Wilkins S, Sham P, Rus-
sell A, Gilvarry K, Lewis S, Toone B, et al. (1994) The influence of
life events on the subsequent course of psychotic illness. A
prospective follow-up of the Camberwell collaborative psychosis
study. Psychol Med 24:503–513
44. van Os J, Jones P, Sham P, Bebbington P, Murray RM (1998) Risk
factors for onset and persistence of psychosis. Soc Psychiatry
Psychiatr Epidemiol 33:596–605
45. van Os J, Jones PB (1999) Early risk factors and adult person-en-
vironment relationships in affective disorder. Psychol Med
29:1055–1067
46. van Os J, Verdoux H, Bijl RV, Ravelli A (1999) Psychosis as an ex-
treme of continuous variation in dimensions of psychopathol-
ogy. In: Gattaz WF, Hafner H (eds) Search for the causes of schiz-
ophrenia. Steinkopff, Darmstadt, vol. 4:59–79
47. Ventura J, Green MF, Shaner A, Liberman RP (1993) Training and
quality assurance with the brief psychiatric rating scale:‘the drift
busters’. Int J Methods Psychiatr Res 3:221–244
48. Watson D, Wiese D, Vaidya J, Tellegen A (1999) The two general
activation systems of affect: structural findings, evolutionary
considerations, and psychobiological evidence. J Pers Soc Psy-
chol 76:820–838
49. Wechsler D (1981) Wechsler adult intelligence scale-revised
manual. Psychological Corporation, New York