RESEARCH ARTICLE
Stress Reactivity and Cognitive Performance
in a Simulated Firefighting Emergency
Sarita J. Robinson, John Leach, P. Jane Owen-Lynch,
and Sandra I. Sünram-Lea
ROBINSON SJ, LEACH J, OWEN-LYNCH PJ, SUNRAM-LEA SI. Stress reactivity and cognitive performance in a simulated firefighting emergency. Aviat Space Environ Med 2013; 84:592–9.
Background: During emergencies maladaptive behavior can reduce
survival. This study compared the effects of a basic firefighter training
course on 21 volunteers (with no firefighting experience) with age and
gender-matched controls. Methods: Stress reactivity (salivary cortisol
and anxiety) were monitored across the course: day 1 (classroom), day 2
(physical equipment training), and day 3 (simulated fire emergency).
Cognitive performance (visual attention, declarative and working memory) considered important in surviving a fire emergency were measured
immediately post-training or after a 20-min delay. Results: Prior to threat
subjects showed an anticipatory cortisol increase but no corresponding
increase in self-reported anxiety. On day 3 cortisol was higher in firefighters tested immediately after (10.37 nmol z L21) and 20 min after
training (7.20 nmol z L21) compared to controls (3.13 nmol z L21). Differences in cognitive performance were observed post-threat, with impairments in visual declarative memory in the firefighting subjects tested
immediately, and working memory impairments observed in those
tested after a 20-min delay. Conclusions: Cognitive impairments were
found following a simulated emergency and could explain maladaptive
responses observed during real fires. Moreover, the results suggest the
type of cognitive impairments observed may be time dependent, with
different cognitive difficulties becoming evident at different times following an emergency.
Keywords: fire emergency, anxiety, memory, cortisol.
dangerous situation tend to be based on information
which is ambiguous, incomplete, or unusual (13,14), thus
placing increased demands on WM just at the time when
its available resources are decreasing. Restrictions in
cognitive function during an emergency can, therefore,
result from physiological factors or psychological factors, although their relationship is little understood.
Under threat two key physiological systems are activated that enhance self-preservation (25): the sympathetic division of the autonomic nervous system (ANS)
and the more slowly responding hypothalamic-pituitaryadrenal (HPA) axis. The role of these two systems is to
prepare the body to cope with threat by increasing the
production of adrenaline and cortisol, which, in turn,
increases heart rate and the supply of glucose to the
muscles while inhibiting digestive system activity (20).
Activation of the ANS and HPA axis occurs in response
to psychological stressors (9), including firefighting activities (17,28). Although these physiological changes
may enhance the physical response to an emergency, the
associated neurochemical changes can increase the risk
of injury
Delivered by Ingenta
to: ?and death by impairing cognitive function.
HPA
axis
activation
increases the level of cortisol which
IP: 5.10.31.210 On: Sun, 18 Jun 2017
06:48:48
ORKING IN A dangerous environment
can produce
acts directly
on those areas of the brain with high densiCopyright:
Aerospace Medical
Association
considerable physiological and psychological disties of glucose receptors such as the frontal cortex and
tress. Exposure to danger can produce impairments in
hippocampus, producing a decrease in attention, workcognitive functioning and behaviors that seem countering memory capacity, inhibitory control (11), and deindicated for survival (12); for example, during a discoclarative memory function (29), which can compromise
theque fire in Gothenberg (Sweden), partygoers failed
chances of survival.
to evacuate even when the building filled with black
Survival requires goal-directed behavior (15) to support
smoke and there was a smell of burning (4). Firefighters
such actions as finding a fire-extinguisher or initiating
are prone to physical and mental strain (28) and, as well
escape procedures (19). However, physiological changes
as the risk of burns and smoke inhalation, dehydration
such as increases in cortisol (10) or psychological changes
from heat exposure can impair visuo-motor performance
(5), working memory, and decision making (6), while
From the School of Psychology, University of Central Lancashire,
physiological stress during smoke diving correlates inPreston, Lancashire, UK; the Centre for Intelligence Studies, FSES
versely with cognitive function (17).
(NorDISS), and the Center for the Study of Human Cognition, UniverWorking memory (WM) is a fundamental component
sity of Oslo, Oslo, Norway; the Department of Biological Sciences,
Lancaster University, Lancaster, Lancashire, UK; and the Department
of cognition and WM capacity is known to become imof Psychology, Lancaster University, Lancaster, Lancashire, UK.
paired during exposure to hazardous activities such as
This manuscript was received for review in March 2012. It was
helicopter underwater evacuation (24) and parachuting
accepted for publication in November 2012.
(16). WM may become more inefficient during a threat
Address correspondence and reprint requests to: Sarita Robinson,
School of Psychology, Darwin Building, Room 213, University of Central
due to an increase in worry and anxiety, which absorbs
Lancashire, Preston, Lancashire, UK PR1 2HE; [email protected].
the limited storage and processing resources, leaving
Reprint & Copyright © by the Aerospace Medical Association,
fewer available to process information from the threat
Alexandria, VA.
environment (8). Furthermore, decisions made in a
DOI: 10.3357/ASEM.3391.2013
W
592
Aviation, Space, and Environmental Medicine x Vol. 84, No. 6 x June 2013
COGNITION DURING EMERGENCIES—ROBINSON ET AL.
consists of morning classroom-based activities followed
by a 2-h practical session in the afternoon on each of the
3 d. The psychological and physiological intensity of the
afternoon sessions was increased over the 3-d course.
On day 1 (baseline), volunteers received a demonstration of different types of fire extinguisher. On day 2, subjects received training in the use of the self-contained
breathing apparatus (SCBA). This task was physically
demanding as subjects completed a number of activities
in full SCBA kit while wearing standard protective firefighting turn-out gear during a 2-h session. On day 2 all
training was completed without the presence of heat or
smoke and so was considered to be a physical but not a
psychological stressor. On day 3 subjects entered a mockup of a ship’s galley in full turn-out gear, including SCBA
kit. The galley was heated to temperatures between
60°C and 130°C and filled with thick, black smoke during a demonstration of rapid fire development. Subjects
completed a 60-min search and rescue exercise followed
by an exercise to extinguish the fire with a variety of
equipment. The training on day 3 was considered to be
both physically and psychologically demanding (Fig. 1).
State-Trait Anxiety Inventory: Anxiety levels were measured using Form Y of the State-Trait Anxiety Inventory
(27). The inventory comprises 40 statements, 20 of which
assess state anxiety and 20 assess trait anxiety. For example, ‘I feel at ease’ or ‘I am a steady person.’ Subjects
circle one of four options relating to how much they
agree with each statement (ranging from ‘Not at all’ to
METHODS
‘Almost Always’). The answers circled give a total score
for state and trait anxiety with a high final score indicatSubjects
ing a high level of anxiety.
Via opportunity sampling from a 3-d basic fire-training
The Stress Arousal Checklist: The Stress Arousal Checklist
course at Fleetwood Nautical College (UK), 21 volun(SACL) (18), comprising two subscales, was used. The
teers with no previous firefighting training were restress subscale uses 19 positive and negative adjective
cruited. The novice firefighters were divided into two
mood-related words, such as ‘Worried’ or ‘Peaceful.’
Delivered
to: ?
groups: those tested immediately post-training
(N 5 11,by Ingenta
TheJun
arousal
IP:
5.10.31.210
On:
Sun,
18
2017 subscale
06:48:48 contains 15 positive and negative
5 women, mean age 29.55 yr, age range 18-43 yr); and
items, such
as ‘Active’ or ‘Drowsy.’ Subjects are required
Copyright:
Aerospace
Medical
Association
those tested after a 20-min delay (N 5 10, 6 women,
to select the word which best describes their current
mean age 25 yr, age range 18-45 yr). Control subjects
state from the options: ‘Definitely Feel,’ Slightly Feel,’
(N 5 11) were recruited from the staff and students at
‘Cannot Decide,’ and ‘Definitely Do Not Feel.’ A value
Lancaster University (UK). Exclusion criteria, identified
of 1 is assigned if the positive adjectives ‘Definitely Feel’
from a modified version of the Blood Services screening
or ‘Slightly Feel’ or the negative adjective options
and medical questionnaires included: 1) active infections,
‘Cannot Decide’ or ‘Definitely Do Not Feel’ are selected.
jaundice within the last year, hepatitis, hemophilia, or
Otherwise a value of 0 is given. The maximum score on
HIV antibody positive; 2) any history of neurological or
the stress scale is 19 while the maximum score on the
psychiatric illness; 3) subjects who awoke earlier than
arousal scale is 15. A higher score represents higher sub06:30 or later than 08:00 to reduce the impact of cortisol
jective feelings for arousal and stress.
diurnal patterns (9); 4) subjects who consumed food or
Saliva collection and biochemical analysis: Saliva samples
drink (apart from water) within 1 h before testing; and
were taken using a salivette saliva sampling device
5) subjects taking medication known to affect cortisol
(Sarstedt Ltd, Leicester, UK). Subjects were instructed to
levels, such as antidepressants (9). All subjects gave
give unstimulated saliva samples by placing a salivette
written consent and were tested in accordance with the
under their tongue for a timed 2-min period. Samples
national and local ethics guidelines according to the
were stored at 240°C and were recovered by thawing
Declaration of Helsinki.
the salivette at room temperature for 15 mins, then centrifuging (1500 rpm) for 15 mins. Cortisol concentration
Equipment and Materials
(nmol z L21) in the saliva was then determined by a high
Hazardous environment: The Fleetwood Nautical Campus
sensitivity salivary cortisol enzyme immunoassay kit
3-d firefighting course trains crew aboard passenger
(Salimetrics, State College, PA) as per the manufacturferries in how to deal with an onboard fire. The course
er ’s instructions.
such as increases in anxiety and worry (26) may impair
a person’s ability to respond appropriately to the extant
dangers. That such impairment in goal-directed behavior occurs is evident from the official inquiry reports of
incidents such as the fire aboard the Boeing-737 at Manchester airport in which 55 people perished (1), and the
explosion and fire aboard the oil platform Piper Alpha in
which 167 men died (7). Therefore, this study examined
anticipatory changes in cortisol and state anxiety prior
to exposure to a firefighting exercise. Furthermore, as
previous studies have reported gender differences in cortisol reactivity, with women having a different cortisol profile compared to men (9), therefore gender differences in
cortisol response during the training were addressed.
Surviving the danger itself is not the end of the problem. Upon rescue and during the early stages of recovery, some people experience ‘disaster syndrome’ (30),
continuing to act in a stunned and bewildered manner
(3). The psychological and physiological factors which
underlie this cognitive collapse in survivors are not understood. As a result this study also addressed changes
in anxiety, cortisol, and cognitive performance after,
as well as before, exposure to a naturalistic firefighting
exercise. Moreover, to allow evaluation of the poststressor response profile (recovery versus deterioration)
during the earlier stages of recovery, stress reactivity
and cognitive performance were assessed immediately
and 20 min after exposure to the naturalistic stressor.
Aviation, Space, and Environmental Medicine x Vol. 84, No. 6 x June 2013
593
COGNITION DURING EMERGENCIES—ROBINSON ET AL.
Fig. 1. Diagrammatic representation of the experimental protocol. STAI 5 State-Trait Anxiety Inventory; SACL 5 Stress Arousal Checklist; MST 5
Map Search Test; GRT 5 Grammatical Reasoning Test; RVDLT 5 Rey Visual Design Learning Test.
The Map Search Test: Selective visual attention was asSubjects viewed the sheet for 3 min during the initial
sessed using the Map Search subtest (MST) from the Test
testing session with recall of the symbols after fire trainof Everyday Attention (23). The test consisted of an A3
ing. At recall subjects were presented with an A4 sheet
map of Philadelphia inserted in a clear A3 plastic wallet.
of paper printed with an empty 4 3 4 grid and were reDelivered by Ingenta to: ?
Subjects were given 2 min to search
the
map
and
circle
quired
draw
as many of the symbols as they could
IP: 5.10.31.210 On: Sun, 18 Jun to
2017
06:48:48
one of three symbols representing aCopyright:
restaurantAerospace
(knife
recall inAssociation
any order in a 2-min period.
Medical
and fork), a vehicle repair garage (screwdriver and a
Procedure
wrench), or a petrol station (petrol pump). The total
number of symbols located in the 2-min period was
A mixed-subjects design was used to explore state
recorded.
anxiety, HPA axis activation (cortisol), and changes in
Grammatical Reasoning Task: The 3-min Grammatical
cognitive performance (visual declarative memory, viReasoning Task (GRT) (2) was used to assess decisionsual selective attention, and decision making) over 3
making (working memory) ability. Subjects were pretraining days. Data were collected on each day between
sented with 64 statements, each being followed by a
13:30 and 16:00 in order to minimize the effects of diurletter pair. The task was to evaluate each sentence and
nal cycle on cortisol levels (9). On days 1, 2, and 3, subdecide if it correctly describes the letter pair that follows
jects in the experimental group completed a morning
of the form: True, False; S is not followed by E 2 ES X.
classroom activity. Before lunch subjects completed an
The task was scored for the number of correct and incoranxiety questionnaire and provided a saliva sample.
rect responses made after 3 min. Three versions of the
Subjects were then presented with the Rey Visual Design
Grammatical Reasoning Task were used and counterLearning Test stimuli for later recall. Following the afterbalanced across the 3 d.
noon activity, which lasted approximately 2 h, subjects
The modified Rey Visual Design Learning Test: Subjects
provided a second saliva sample and completed both
were presented with one of three versions of a modified
the anxiety questionnaire and the three cognitive perforform of the Rey Visual Design Learning test (RVDL) (22)
mance tasks. One group completed these measures imto assess visual declarative memory. Each task commediately on completion of the afternoon activity. The
prised 16 2-D symbols presented on a 4 3 4 grid on one
second group waited in the fire-ground seating area for
side of white A4 paper. Each symbol (4 cm by 4 cm) was
20 min before testing to allow investigation of cognitive
printed in black ink and displayed in one grid box.
collapse/recovery post-stressor. The control group was
594
Aviation, Space, and Environmental Medicine x Vol. 84, No. 6 x June 2013
COGNITION DURING EMERGENCIES—ROBINSON ET AL.
also tested over a 3-d period, but completed everyday
office-based activities instead of firefighter training.
Statistical Analysis
Anxiety and salivary cortisol measures were recorded
on each of the 3 test days. Comparisons between the experimental (firefighting) and the control groups were
analyzed using two-way ANOVAs with Greenhouse
Geisser correction as appropriate. Post-training the firefighting subjects were divided into two groups: those
who were tested immediately after threat exposure (immediate group) and those who were tested 20 min after
threat cessation (delayed group).
day 3). No main effects were found for group [F(2,29) 5
0.48, P 5 0.62, Eta2 5 0.03] or for day [F(2,58) 5 2.47, P 5
0.09, Eta2 5 0.08] and there was no significant interaction
between day and group [F(4,58) 5 2.46, P 5 0.06, Eta2 5
0.15].
Finally, state anxiety was compared for each group
before and after training on each of the 3 d. Analysis revealed that on day 3 the immediate group had significantly higher state anxiety after training compared to
before training [t(10) 5 3.96, P 5 0.003]. No other differences were found (see Fig. 2).
SACL Arousal Levels Post-Training
Post-training state arousal was analyzed using a twoway mixed ANOVA with one between factor (group) at
three levels (immediate, delayed, control) and one
RESULTS
within factor (day) at three levels (day 1, day 2, day 3).
State Anxiety
No main effects were found for group [F(2,29) 5 0.29,
Anticipatory anxiety: State anxiety was measured prior
P 5 0.75, Eta2 5 0.02] or for day [F(2,58) 5 1.58, P 5 0.21,
to training on each of the 3 d and compared using a twoEta2 5 0.05]. However, there was a significant interacway mixed ANOVA with one between factor (group) at
tion between day and group [F(4,58) 5 3.92, P 5 0.007,
two levels (firefighting, controls) and one within factor
Eta2 5 0.21].
(day) at three levels (day 1, day 2, day 3). No main effect
Post hoc testing using paired samples t-tests revealed
was found for day on anxiety levels [F(2,60) 5 0.30, P 5
no significant differences in arousal in the immediate
0.74, Eta2 5 0.01] nor for group [F(1,30) 5 0.19, P 5 0.66,
group between day 1 and day 2 [t(10) 5 0.000, P 5 1.00],
Eta2 5 0.01], and there was no significant interaction
day 1 and day 3 [t(10) 5 20.31, P 5 0.76] or day 2 and
[F(2,60) 5 0.51, Eta2 5 0.60]. These results indicate that
day 3 [t(10) 5 20.41, P 5 0.69]. No significant differthere was no difference in anxiety between controls and
ences were found in the delayed group between day 1
the experimental group, suggesting no anticipatory inand day 2 [t(9) 5 20.52, P 5 0.62], day 1 and day 3 [t(9) 5
creases in anxiety levels.
0.99, P 5 0.35], or day 2 and day 3 [t(9) 5 0.56, P 5 0.59];
Reactive anxiety: State anxiety levels for the three
however, in the control group a significant reduction in
groups (immediate, delayed, and control) were recorded
arousal was observed between day 1 and day 3 [t(10) 5
both prior to and following the training session on each
2.65, P 5 0.02]. No differences were found between day
of the 3 training days. Descriptive statistics for all three
1 and day 2 [t(10) 5 2.02, P 5 0.07] or between day 2 and
groups across the 3 d for anxiety (both anticipatory and
day 3 [t(10) 5 1.97, P 5 0.078]. Further analyses revealed
reactive), arousal, and stress levels post-stressor
are givenby Ingenta
no signifi
Delivered
to: ?cant differences between the three groups’
in Table I. Post-training state anxiety
was analyzed
arousal
levels
on day 1 or day 2; however, on day 3 a
IP: 5.10.31.210
On: Sun, 18
Jun 2017
06:48:48
Copyright:
Aerospace
Association
using a two-way mixed ANOVA with
one between
fac- Medical
significant
difference in arousal was observed between
tor (group) at three levels (immediate, delayed, control)
the control and firefighting delayed group post-training
and one within factor (day) at three levels (day 1, day 2,
[t(19) 5 2.10, P 5 0.049].
TABLE I. MEAN STATE ANXIETY, AROUSAL, AND STRESS LEVELS (WITH SDs) FOR THE FIREFIGHTING GROUPS (IMMEDIATE AND DELAYED)
AND CONTROL GROUP ON EACH OF THE 3 TEST DAYS.
Group/Day
Immediate
Day One
Day Two
Day Three
Delayed
Day One
Day Two
Day Three
Control
Day One
Day Two
Day Three
Anticipatory Anxiety
Reactive Anxiety
Stress
Arousal
34.91 (10.63)
34.45 (12.67)
33.91 (10.06)*
30.09 (8.47)
35.91 (14.76)
40.00 (10.95)*
4.09 (2.98)
5.73 (4.47)
6.73 (4.54)
10.36 (3.50)
10.36 (2.69)
10.64 (3.23)
31.60 (5.72)
36.10 (6.81)
38.10 (7.56)
34.30 (6.98)
41.20 (9.95)
32.70 (10.19)
6.10 (4.75)
7.00 (4.99)
5.00 (4.83)
9.80 (3.85)
10.50 (3.63)
10.80 (2.94)1
34.18 (8.22)
32.45 (10.99)
34.18 (9.77)
32.55 (8.36)
32.91 (9.69)
33.55 (8.51)
4.73 (4.65)
3.91 (3.81)
4.64 (5.28)
11.64 (3.23)**
10.00 (4.22)
7.09 (4.83)**,1
*, **, 1Denote a significant difference between group or time points (P , 0.05).
* Denotes a significant differences between the anticipatory and the reactive anxiety time points.
** Denotes a significant difference between the control groups level of arousal on day one compared to day three.
1
Denotes a significant difference in arousal between the delayed group and the control group on day three.
Aviation, Space, and Environmental Medicine x Vol. 84, No. 6 x June 2013
595
COGNITION DURING EMERGENCIES—ROBINSON ET AL.
Fig. 2. Mean state anxiety scores (with SDs) for the firefighting
group (immediate and delayed) and control group day 3 prior to and
post-training. *Denotes a significant difference before and after training
(P , 0.01).
SACL Stress Levels Post-Training
Stress levels post-training were analyzed using a twoway mixed ANOVA with one between factor (group) at
three levels (immediate, delayed, control) and one within
factor (day) at three levels (day 1, day 2, day 3). No main
effects were found for group [F(2,29) 5 0.59, P 5 0.56,
Eta2 5 0.04] or for day [F(2,58) 5 0.24, P 5 0.78, Eta2 5
0.01]. Further, no significant interaction between day and
group was found [F(4,58) 5 1.15, P 5 0.34, Eta2 5 0.07].
Cortisol
Reactive cortisol: Cortisol levels were measured on
each day after training and were analyzed using a twoway mixed ANOVA with one between factor (group)
at three levels (immediate, delayed, control) and one within factor (day) at three levels (day 1, day 2, day 3). Post
training cortisol levels for the subjects who undertook
the firefighting training and controls can be viewed in
Fig. 3. No main effect of day was observed [F(2,52) 5
0.22, P 5 0.80, Eta2 5 0.01] and there was no significant
group 3 day interaction [F(4,52) 5 2.21, P 5 0.08 Eta2 5
0.15]. Strictly there was no main effect of group [F(2,26) 5
3.33, P 5 0.052, Eta2 5 0.20]; however, given the borderline P-value coupled with a literature that suggests
a robust effect of stressor exposure on cortisol reactivity,
post hoc analysis using planned comparisons was undertaken. A one-way ANOVA suggested that groups
differed only on day 3 post-task in levels of cortisol
[F(2,28) 5 3.98, P 5 0.03] with no differences on day 1
[F(2,28) 5 3.21, P 5 0.056] or day 2 [F(2,26) 5 1.19, P 5
0.32]. Further analysis using independent t-test on the
groups on day 3 revealed no significant difference between the immediate and the delayed groups [t(13.86) 5
0.97, P 5 0.35]. However, both the immediate [t(9.41) 5
2.50, P 5 0.03] and the delayed group [t(10.40) 5 2.53,
P 5 0.03] had significantly higher cortisol than the controls on day 3 post-firefighting.
Cognitive Performance
Following training on each of the 3 d subjects underGender differences: Previous research has suggested
took tests of grammatical reasoning, visual declarative
that cortisol reactivity can vary between men and
memory, and visual search. The descriptive statistics are
women post-stressor. As a result cortisol levels for men
given in Table II.
and women were compared across the 3 d using a twoGrammatical Reasoning Task: Scores for the GRT were
way mixed ANOVA with one between factor (gender) at
analyzed using a two-way mixed ANOVA with one betwo levels (men, women) and one within factor (day) at
tween factor (group) at three levels (immediate, delayed,
three levels (day 1, day 2, day 3). No main effects of gencontrol)to:and
Delivered
by
Ingenta
? one within factor (day) at three levels (day
der [F(1,27) 5 0.198, P 5 0.66, Eta2IP:
5 0.007] or day were
1, day 2, day 3). The results revealed a main effect of day
2 5.10.31.210 On: Sun, 18 Jun 2017 06:48:48
found [F(2,54) 5 0.251, P 5 0.78, Eta 5Copyright:
0.009]. Further,
no
Aerospace
Medical
Association
[F(1.36, 39.31)
5 6.56, P 5 0.003, Eta2 5 0.19] and a main
significant interaction between day and group [F(2,54) 5
effect of group which approached significance [F(2,29) 5
1.658, P 5 0.97, Eta2 5 0.001] was observed between
3.22, P 5 0.055, Eta2 5 0.18]. However, the group 3 day
men (day 1, 8.56 nmol z L21; day 2, 6.70 nmol z L21; day
interaction was not significant [F(2.71, 39.31) 5 0.89, P 5
3, 7.70 nmol z L21) and women (day 1, 7.30 nmol z L21;
0.48, Eta2 5 0.06].
day 2, 6:33; nmol z L21, and day 3, 6.51 nmol z L21).
Anticipatory cortisol: Prior to training on each of the
3 d subjects in the firefighting group had their cortisol
measured and these scores were compared to those in
the control group. Anticipatory cortisol was analyzed
using a two-way mixed ANOVA with one between factor (group) at two levels (firefighting, controls) and one
within factor (day) at three levels (day 1, day 2, day 3).
The results revealed no main effect of day on cortisol
levels [F(2,54) 5 1.17, P 5 0.32, Eta2 5 0.04], but there
was a significant main effect of group [F(1,27) 5 12.34,
P 5 0.002, Eta2 5 0.314] with anticipatory cortisol elevated in the firefighters (day 1, 17.29 nmol z L21; day 2,
19.60 nmol z L21; day 3, 11.80 nmol z L21) compared to
controls (day 1, 4.89 nmol z L21; day 2, 6.17 nmol z L21;
day 3, 4.81 nmol z L21) across all 3 d. No significant interFig. 3. Cortisol levels (means and SDs) for the firefighting groups and
action was found between group 3 day [F(2,54) 5 0.66,
control group on each of the 3 d post-training. *Denotes a significant
P 5 0.52, Eta2 5 0.024].
difference between two groups (P , 0.05).
596
Aviation, Space, and Environmental Medicine x Vol. 84, No. 6 x June 2013
COGNITION DURING EMERGENCIES—ROBINSON ET AL.
TABLE II. COGNITIVE PERFORMANCE TASK MEANS (WITH STANDARD DEVIATIONS) FOR SUBJECTS IN FIREFIGHTING GROUPS
(IMMEDIATE AND DELAYED) AND CONTROL GROUP ON EACH OF THE THREE TEST DAYS.
Day 1
Declarative Memory Rey Visual Design Learning
Immediate
Delayed
Control
Total
Working Memory Grammatical Reasoning
Immediate
Delayed
Control
Total
Visual Search Map Search Task
Immediate
Delayed
Control
Total
Day 2
Day 3
7.09 (2.21)
9.00 (2.67)
7.27 (2.15)
7.75 (2.42)†
3.82 (2.09)
5.70 (1.95)
5.27 (1.74)
4.91 (2.04)†
5.73 (3.04)*
8.50 (2.55)*
7.45 (2.91)
7.19 (2.99)†
17.82 (7.45)
16.40 (6.45)
21.82 (10.65)
18.75 (8.49)‡
22.18 (9.15)
18.10 (8.69)
27.91 (12.96)
22.88 (10.96)‡
24.27 (11.58)**
18.60 (7.79)**
32.27 (15.90)
25.25 (13.21)‡
74.64 (6.10)
70.80 (7.955)
74.55 (4.458)
73.41 (6.32)
72.27 (15.13)
72.90 (7.39)
74.27 (5.00)
73.16 (9.93)
72.82 (7.77)
69.70 (6.90)
75.64 (2.98)
72.81 (6.49)
*, **Denote a significant difference between groups (P , 0.05). †, ‡Denote a significant difference over time (P , 0.05).
* Denotes a significant difference between the immediate and delayed time points on day 3.
** Denotes a significant in working memory performance between immediate and delayed time points on day 3.
†
Denotes a significant difference between the total scores of the Rey Visual design learning task over the three days (i.e between d 1 and d 2; d 2 and
d 3 and d 1 and d 3).
‡
Denotes a significant difference in WM grammatical reasoning (total) over the three days.
Although the main effect of group fell just outside sig0.001]. However, no difference was found between day 1
and day 3 [t(31) 5 1.20, P 5 0.24]. Planned comparinificance level, the results were in the expected direction
sons for group revealed better performance in the deand so a series of planned comparisons was undertaken
to further explore this finding. The results of the indelayed group than the immediate group [t(19) 5 2.55,
pendent t-tests revealed that those in the immediate
P 5 0.019]; however, no differences were found in
group had better performance than the delayed group
performance between the control group and either the
[t(19) 5 2.55, P 5 0.019]. However, no differences were
immediate [t(20) 5 1.38, P 5 0.18] or the delayed groups
found in performance between the immediate and con[t(19) 5 1.35, P 5 0.20].
trols [t(20) 5 1.38, P 5 0.18] and the delayed and conVisual Map Search Task: Descriptive statistics for the
MST are given in Table II. The data were analyzed using
trols [t(19) 5 1.35, P 5 0.20].
a two-way mixed ANOVA with one between factor
Planned comparisons were undertaken for post hoc
analysis. Paired t-test analysis compared performance
(group) at three levels (immediate, delayed, control) and
on the GRT across the 3 d for all groups. Across the three
one within factor (day) at three levels (day 1, day 2, day 3).
Delivered by Ingenta to: ?
NoJun
signifi
cant
difference in the total number of symbols
groups there was a difference in performance
on
day
1
IP: 5.10.31.210 On: Sun, 18
2017
06:48:48
compared to day 2 [t(31) 5 22.20, P 5
0.04], withAerospace
signifi- Medical
located Association
was observed between the three groups [F(2,29)
Copyright:
cantly more GRT questions answered correctly on day 2
5 1.07, P 5 0.36, Eta2 5 0.07], or across the 3 d [F(1.50,
than on day 1. Furthermore, a significant difference was
43.40) 5 0.08, P 5 0.88, Eta2 5 0.003]. No significant
found between day 2 and day 3 [t(31) 5 22.27, P 5 0.03],
group 3 day interaction was found [F(3.00, 43.40) 5 0.56,
with significantly more correct answers on day 3 than
P 5 0.64, Eta2 5 0.04].
day 2. Finally, significantly better performance was
found on day 3 compared to day 1 [t(31) 5 22.96, P 5
DISCUSSION
0.01], indicating higher reasoning ability on day 3 than
day 1. This suggests that all groups improved in their
The chances of surviving in a hazardous environment,
GRT performance over the 3 d.
such as a fire emergency, are enhanced if people do
The modified Rey Visual Design Learning Test: Descriptive
not suffer impaired cognition during the event (13,14).
statistics for the performance on the RVDL test are given
Previous research has suggested that both psychologin Table II. The data were analyzed using a two-way
ical factors, e.g., increased anxiety (8), and physiological
mixed ANOVA with one between factor (group) at three
changes, e.g., increased cortisol, may lead to cognitive
levels (immediate, delayed, control) and one within facimpairments (11). Consequently, this study addressed
tor (day) at three levels (day 1, day 2, day 3). The results
possible changes in state anxiety, cortisol, and cognition
revealed main effects for both group [F(2,29) 5 3.49, P 5
during a simulated fire emergency.
0.044, Eta2 5 0.19] and day [F(2,58) 5 22.47, P , 0.001,
The results revealed that people exposed to a naturalEta2 5 0.44]. There was no group 3 day interaction
istic threat environment did not show any anticipatory
[F(4,58) 5 0.75, P 5 0.56, Eta2 5 0.05].
anxiety. However, those tested immediately after expoPlanned comparisons for day revealed that subjects in
sure to the simulated fire emergency on day 3 showed
all groups performed worse on day 2 compared to either
increased self-reported anxiety compared to pre-session
levels. Although subjective anxiety was not elevated in
day 1 [t(31) 5 6.36, P , 0.001] or day 3 [t(31) 5 5.35, P ,
Aviation, Space, and Environmental Medicine x Vol. 84, No. 6 x June 2013
597
COGNITION DURING EMERGENCIES—ROBINSON ET AL.
was elevated (in this case on day 3 post-training), cognianticipation of the training, HPA axis activation did
tive impairments were expected in the three cognitive
occur with high cortisol levels being observed in firetasks (visual search, visual declarative memory, and gramfighters prior to each training session. This finding is
matical reasoning). However, only subjects tested 20 min
consistent with previous work on anticipatory rises in
after exposure to the threat showed impairment in the
cortisol (10). However, cortisol only remained elevated
grammatical reasoning task compared to controls. The
on day 3 when both physiological and psychological
lack of impact of elevated cortisol on declarative memory
strain were present. No differences in cortisol reactivity
was unexpected as exposure to the fire emergency induced
were observed between men and women. With respect
cortisol increases in the subjects that were similar to those
to cognitive function, visual search ability was prefound in previous studies that have demonstrated postserved immediately after and 20 min following the
stressor memory impairment (11).
threat. Visual declarative memory was impaired immeSubjects tested immediately after exposure to the psydiately after the simulated emergency, but not later;
chological and physiological stressor showed impairconversely, grammatical reasoning was most impaired
ment in visual declarative memory which was not present
20 min after the simulated fire but preserved when
in subjects tested a short time later. This suggests that, in
tested immediately post-threat.
the immediate aftermath of an emergency, people may
Several studies have linked high anxiety to impairhave difficulty recalling potentially vital visual informaments in cognition (8,26) and poor cognitive function in
tion such as plans of the building. However, it would apturn has been associated with reduced chances of surpear that threat-induced impairments in visual declarative
vival (13,14). As a result this study examined self-reported
memory recover quickly after an emergency.
anxiety both before and following fire training. No selfConversely, working memory appeared to be prereported increases in anticipatory anxiety were observed
served in subjects immediately after exposure to the
prior to threat exposure. This finding was unexpected as
simulated emergency. As working memory is generally
physiological activation of the HPA axis was recorded in
accepted to be important for dealing successfully with
subjects while anticipating the simulated fire emergency.
novel and threatening situations (16), this finding offers
This mismatch between anxiety and cortisol levels, howsome reassurance. The lack of marked impairments in
ever, has been observed in previous research, which found
working memory immediately post-threat suggests the
that subjective reports of stress reactivity do not always
majority of people have a good chance of responding
correspond to physiological stress markers (24).
appropriately to a threat environment. However, our reIn this current study, elevations in cortisol were obsults also show that working memory may become imserved prior to training on each of the 3 d. Anticipatory
paired shortly after facing a physiological and physical
increases in cortisol have not previously been reported
stressor. Previous studies suggest that stressor-induced
during firefighter training (21); however, the volunteers
cortisol does not always have an immediate impact on
in the current study were nonprofessionals and so had
cognition (11) and anecdotal reports also suggest that,
no prior experience of fighting a fire. Therefore, our
while people may appear to be coping with an emerfindings may be more reflective of members of the pubgency situation initially, they show some cognitive dyslic or workers not trained in firefighting techniques who
Delivered by Ingenta to: ?
upon06:48:48
rescue (3,30).
unexpectedly find themselves in a fi
re
emergency.
Given
IP: 5.10.31.210 On: Sun, function
18 Jun 2017
Failures
in working memory function in the aftermath
that cortisol levels have been associated
with reduced
Copyright:
Aerospace Medical
Association
of an emergency can have implications for the emercognitive performance (11,29), future research should
gency services in that they should be aware that people
consider whether an anticipatory increase in cortisol, as
who appear to be functioning normally at the point of
a consequence of working in at-risk environments, could
rescue may show a cognitive decline shortly after. Furesult in cognitive impairments that may influence surture studies should consider the importance of timing
vival should an emergency occur.
when assessing the effect of increased cortisol on cogniIncreases in cortisol levels have been previously found
tive functions as some deficits may only become apparwith physical exertion (17), but such increases were not
ent later.
observed in the current study. Cortisol was only elevated
The current study demonstrated that when people are
following training on day 3, when the trainees were explaced in threatening environments, such as a simulated
posed to a hazardous environment which included both
fire emergency, marked physiological stress reactivity
a physiological and psychological element. Changes in
(cortisol) both prior to and immediately post-exposure
cortisol were not observed when the environment concan occur. Increased self-reported anxiety levels were
tained only physiological strain (day 2). Therefore, in order
also evident, but only immediately after exposure to the
for cortisol levels to remain high throughout the training
simulated fire emergency. No clear-cut impairments in
session, psychological pressure in addition to physical
cognition were found after threat exposure, but some
strain was required. These findings support Dickerson
cognitive performance decrements were found between
and Kemeny (9), who found that for an increase in cortisol
those subjects tested immediately after their training peto occur, exposure to an uncontrollable and self-evaluative
riod and those tested after a short delay. These findings
psychological threat is needed.
begin to clarify our understanding of some of the malElevated cortisol has been shown to impair those cogadaptive behaviors observed in response to exposure to
nitive functions associated with both the frontal cortex
hazardous situations.
(11) and the hippocampus (29). Therefore, when cortisol
598
Aviation, Space, and Environmental Medicine x Vol. 84, No. 6 x June 2013
COGNITION DURING EMERGENCIES—ROBINSON ET AL.
ACKNOWLEDGMENTS
We thank the staff and students at Fleetwood Nautical Campus
(UK) for their assistance with this study.
Authors and affiliations: Sarita J. Robinson, Ph.D., School of Psychology,
University of Central Lancashire, Preston, Lancashire, UK; John
Leach, Ph.D., Centre for Intelligence Studies, FSES (NorDISS) and the
Center for the Study of Human Cognition, University of Oslo, Oslo,
Norway; and P. Jane Owen-Lynch, Ph.D., Department of Biological
Sciences, and Sandra I. Sünram-Lea, Ph.D., Department of Psychology,
Lancaster University, Lancaster, Lancashire, UK.
14. Leach J. Cognitive paralysis in an emergency: the role of the supervisory attentional system. Aviat Space Environ Med 2005; 76:134–6.
15. Leach J, Ansell L. Impairment in attentional processing in a field
survival environment. Appl Cogn Psychol 2008; 22:643–52.
16. Leach J, Griffith R. Restrictions in working memory capacity during
parachuting: a possible cause of ‘no-pull’ fatalities. Appl Cogn
Psychol 2008; 22:147–57.
17. Lusa S, Louhevaara V, Smolander J, Kivimaki M, Korhonen O.
Physiological responses of firefighting students during simulated
smoke-diving in the heat. Am Ind Hyg Assoc J 1993; 54:
228–31.
18. Mackay C, Cox T, Burrows G, Lazzerini T. An inventory for the
measurement of self-reported stress and arousal. Br J Soc Clin
REFERENCES
Psychol 1978; 17:283–4.
1. AA Air Accident Investigations Board. Report on the accident
19. Mileti DS, Peek L. The social psychology of public response to
to Boeing 737-236, G-BGJL, at Manchester Airport on 22nd
warnings of a nuclear power plant accident. J Hazard Mater
August 1985. London: AAIB; 1988.
2000; 75:181–94.
2. Baddeley AD. A three-minute reasoning test based on grammatical
20. Miller DB, O’Callaghan JP. Neuroendocrine aspects of the response
transformation. Psychon Sci 1968; 10:341–2.
to stress. Metabolism 2002; 51:5–10.
3. Baker GW, Chapman DW. Man and society in disaster. New York:
21. Perroni F, Tessitore A, Cibelli G, Lupo C, D’Artibale E, et al. Effects of
Basic Books, Inc; 1962.
simulated firefighting on the responses of salivary cortisol, alpha4. Cassuto J, Tarnow P. The discothque fire in Gothenburg 1998: a
amylase and psychological variables. Ergonomics 2009; 52:484–91.
tragedy among teenagers. Burns 2003; 29:405–16.
22. Rey A. Memory. In: Spreen O, Strauss E. A compendium of neuro5. Cian C, Barraud PA, Melin B, Raphel C. Effects of fluid ingestion
psychological tests: administration, norms, and commentary.
on cognitive function after heat stress or exercise-induced
New York: Oxford University Press; 1991.
dehydration. Int J Psychophysiol 2001; 42:243–51.
23. Robertson IH, Ward T, Ridgeway V, Nimmo-Smith I. The structure
6. Cian C, Koulmann N, Barraud PA, Raphel C, Jimenez C, Melin B.
of normal human attention: the test of everyday attention. J Int
Influences of variations in body hydration on cognitive function:
Neuropsychol Soc 1996; 2:525–34.
effect of hyperhydration, heat stress, and exercise-induced de24. Robinson SJ, Sunram-Lea SI, Leach J, Owen-Lynch PJ. The effects of
hydration. J Psychophysiol 2000; 14:29–36.
exposure to an acute naturalistic stressor on working memory,
7. Cullen L. The public inquiry into the Piper Alpha disaster. London:
state anxiety and salivary cortisol concentrations. Stress 2008;
HMSO, Department of Energy; 1990.
11:115–24.
8. Darke S. Anxiety and working memory capacity. Cogn Emotion
25. Selye H. Stress without distress. New York: New American Library;
1988; 2:145–54.
1974.
9. Dickerson SS, Kemeny ME. Acute stressors and cortisol responses:
26. Somer E, Tamir E, Maguen S, Litz BT. Brief cognitive-behavioural
a theoretical integration and synthesis of laboratory research.
phone-based intervention targeting anxiety about the threat of
Psychol Bull 2004; 130:355–91.
attack: a pilot study. Behav Res Ther 2005; 43:669–79.
10. Garcia-Leal C, Alexandre CBV, Parente C, Del-Ben CM, Guimaraes
27. Spielberger CD. Manual for the State-Trait Anxiety Inventory.
FS, et al. Anxiety and salivary cortisol on symptomatic and
Palo Alto, CA: Consulting Psychologists; 1983.
nonsymptomatic panic patients and healthy volunteers per28. Sünram-Lea SI, Owen-Lynch J, Robinson SJ, Jones E, Hu H. The
forming simulated public speaking. Psychiatry Res 2005; 133:
effect of energy drinks on cortisol levels, cognition and mood
239–52.
during a fire-fighting exercise. Psychopharmacology (Berl). 2012;
11. Hoffman R, al’Absi M. The affect of acute stress on subsequent
219:83–97.
neuropsychological test performance. Arch Clin Neuropsychol
29. Wolf OT, Schommer NC, Hellhammer DH, Reischies FM,
2004; 19:497–506.
Kirschbaum C. Moderate psychosocial stress appears not to
12. Leach J. Maladaptive behaviour in survivors: dysexecutive survivor
impair recall of words learned 4 weeks prior to stress exposure.
syndrome. Aviat Space Environ Med 2012; 83:1152Delivered
–61.
by Ingenta Stress
to: ? 2002; 5:59–64.
13. Leach J. Why people ‘freeze’ in an emergency: temporal and cogIP:
5.10.31.210
On:
Sun,
18
Jun
2017 06:48:48
30.
Wolfenstein
M. Disaster: a psychological essay. Glencoe, IL: The
nitive constraints on survival responses. Aviat Space Environ
Copyright: Aerospace MedicalFree
Association
Press; 1957.
Med 2004; 75:539–42.
Aviation, Space, and Environmental Medicine x Vol. 84, No. 6 x June 2013
599