International Journal of Psychophysiology 98 (2015) 470–476

Contents lists available at ScienceDirect

International Journal of Psychophysiology

journal homepage: www.elsevier.com/locate/ijpsycho

Does cigarette smoking relieve stress? Evidence from the event-related

potential (ERP)
Damee Choi a,⁎, Shotaro Ota b, Shigeki Watanuki a
a
Faculty of Design, Kyushu University, 4-9-1 Shiobaru, Minami-ku, Fukuoka 815-8540, Japan
b
Graduate School of Integrated Frontier Science, Kyushu University, 4-9-1 Shiobaru, Minami-ku, Fukuoka 815-8540, Japan

a r t i c l e i n f o a b s t r a c t

Article history: Previous studies have reported a paradox that cigarette smoking reduces stress psychologically; however, it in-
Received 29 June 2015 creases the arousal level physiologically. To examine this issue, our study aimed to investigate whether cigarette
Received in revised form 10 October 2015 smoking relieves stress by measuring the late positive potential (LPP), a component of the event-related potential
Accepted 19 October 2015
(ERP). In Experiment 1, participants first watched emotionally neutral images; second, they received a break; and
Available online 20 October 2015
finally, they watched emotionally neutral images again. In the break, they smoked a cigarette (smoking condi-
Keywords:
tion) or simply rested without smoking (non-smoking condition). The procedure of Experiment 2 was the
Cigarette smoking same as that of Experiment 1, except that the participants watched unpleasant images as stress stimuli before
Late positive potential the break. In Experiment 1, the LPP decreased from before to after the break in the smoking condition, but not
Event-related potentials in the non-smoking condition, suggesting that smoking cigarettes in the neutral state reduces the arousal level.
Arousal In Experiment 2, the LPP for 400–600 ms decreased from before to after the break, both in the smoking and
Stress non-smoking conditions; however, the LPP for 200–400 ms decreased from before to after the break only in
the smoking condition. This suggests the possibility that cigarette smoking in the unpleasant state may facilitate
a decrease in the arousal level faster than with non-smoking. In both Experiments 1 and 2, the subjective rating
results also suggested that cigarette smoking decreased anxiety. Taken together, both the physiological (LPP) and
the psychological responses from our study suggest that cigarette smoking perhaps relieves stress.
© 2015 The Authors. Published by Elsevier B.V. All rights reserved. This is an open access article under the
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction To address whether cigarette smoking actually has an effect of re-

ducing stress and negative emotions, previous studies have examined
Previous studies have suggested that people smoke cigarettes in the effect of smoking by using both psychological and physiological in-
order to regulate emotion and relieve negative emotions (for review, dices. In the results of subjective ratings, participants reported that
Kassel et al., 2003). For example, according to a survey study of adoles- acute smoking decreased their anxiety and stress (Nesbitt, 1973;
cent smokers (Dozois et al., 1995; Nichter et al., 1997), the most fre- Pomerleau and Pomerleau, 1987; Perkins et al., 1992; Parrott, 1995),
quently mentioned reasons for cigarette smoking were stress which is in line with the expectation of smokers that cigarette smoking
reduction and relaxation. Also, stress from life events is thought to be relieves stress. On the other hand, physiological results have indicated
one of the important factors of smoking motivation (Warburton et al, that smoking increases the heart rate (Pomerleau and Pomerleau,
1991), and one laboratory experiment reported that stress decreases 1987; Perkins et al., 1992; Woodson et al., 1986), which is a sign of in-
the ability to resist smoking (McKee et al., 2011). Moreover, smokers creased arousal. Arousal is an important component of the stress re-
have higher neuroticism and anxiety traits than non-smokers (McCrae sponse (Winsky-Sommerer et al., 2005) and an increased arousal level
et al., 1978). This relationship between smoking behavior and personal- is a typical physiological response elicited by a stressor (Chrousos,
ity traits (McCrae et al., 1978) suggests the possibility that cigarette 1998). In addition, previous research on therapy has related low arousal
smoking might be able to reduce anxiety, although there is also a possi- levels with low stress levels, or with the relaxation effects of therapy
bility that smoking behavior might increase neuroticism and anxiety. (for example, music therapy, reviewed in Pelletier, 2004; mindfulness
therapy, Mendelson et al., 2010). Therefore, there seems to be a close re-
lationship between arousal and stress, although this relationship is not
Abbreviations: LPP, late positive potential; ERP, event-related potential; EEG, electro- always supported (Bennett et al., 2003). Given this close relationship
encephalogram; POMS, Profile of Mood States; EOG, electrooculography; ANOVA, analysis
between arousal and stress, surprisingly, cigarette smoking and stress
of variance.
⁎ Corresponding author. share a common physiological response (Kassel et al., 2003), indicating
E-mail address: [email protected] (D. Choi). that the physiological responses to cigarette smoking are opposite to

http://dx.doi.org/10.1016/j.ijpsycho.2015.10.005
0167-8760/© 2015 The Authors. Published by Elsevier B.V. All rights reserved. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
 D. Choi et al. / International Journal of Psychophysiology 98 (2015) 470–476 471

smokers' expectations that cigarette smoking reduces stress. This incon- the quality of one participant's EEG was poor. They all reported smoking
gruity between the psychological and physiological responses to at least one or more cigarettes per day. We focused only on smokers,
smoking is a phenomenon known as “Nesbitt's paradox”, in which cig- since smokers and non-smokers are fundamentally different in terms
arette smoking relieves smokers psychologically; however, it increases of baseline levels of stress (Gilbert and Gilbert, 1995). Written informed
the arousal level physiologically (Nesbitt, 1973). Thus, it is difficult to consent was obtained from all participants prior to the start of the study.
conclude whether cigarette smoking can actually reduce stress and neg- All study protocols were approved by the ethics committee in the Facul-
ative emotions. To address this issue, it may be useful to measure other ty of Design at Kyushu University, Japan.
physiological indices, since Kassel et al. (2003) have pointed out that
previous studies investigating the effect of cigarette smoking on stress 2.2. Stimuli
reduction have mainly focused on peripheral processes such as the
heart rate, as mentioned above (Pomerleau and Pomerleau, 1987; A total of 60 images (40 neutral and 20 unpleasant images) were se-
Perkins et al., 1992; Woodson et al., 1986). lected from the International Affective Picture System (IAPS) (Lang
In this context, our study focused on the late positive potential (LPP) et al., 2008). The neutral images included images of shoes, a cup, and a
component of the event-related potential (ERP), in order to test the ef- tissue box, while the unpleasant pictures included images of a dirty toi-
fect of cigarette smoking on stress reduction. The ERP is one of the elec- let, a cockroach, and an injured animal. The mean IAPS normative va-
trophysiological indices of brain activity and is derived from an lence ratings (Lang et al., 2008) of the images selected in the present
electroencephalogram (EEG) time-locked to a specific event (for exam- study were 5.0 for neutral images and 2.9 for unpleasant images
ple, the presentation of a picture or a sound). The LPP is a component of (10 = very pleasant; 0 = very unpleasant).
the ERP and is a sustained positive polarity shown approximately The picture identification numbers from IAPS were as follows: Before
200 ms after stimuli (picture or sound) onset (reviewed in Hajcak the break (Experiment 1): 5471, 6150, 7001, 7003, 7010, 7017, 7020,
et al., 2010; Hajcak et al., 2012; Olofsson et al., 2008). Many previous 7035, 7037, 7041, 7050, 7059, 7090, 7150, 7161, 7175, 7185, 7211, 7255,
studies have reported that the LPP amplitude is greater in response to and 7705; Before the break (Experiment 2): 1205, 1220, 1270, 1271,
pleasant and unpleasant stimuli than to emotionally neutral stimuli 3010, 6020, 7078, 9090, 9110, 9183, 9290, 9295, 9300, 9301, 9320, 9325,
(for example, Cuthbert et al., 2000; Schupp et al., 2000; Weinberg 9395, 9405, 9571, and 9590; After the break (Experiment 1 and
et al., 2012). Moreover, there is a positive relationship between the sub- Experiment 2): 5510, 7000, 7002, 7004, 7009, 7012, 7019, 7032, 7036,
jective arousal of stimuli and the LPP amplitude (Cuthbert et al., 2000). 7038, 7045, 7055, 7077, 7130, 7160, 7170, 7207, 7233, 7491, and 7950.
Taken together, the LPP indexes the level of arousal induced by emo-
tional stimuli (reviewed in Hajcak et al., 2010; Hajcak et al., 2012;
Olofsson et al., 2008). As mentioned above, an increased arousal level 2.3. Procedures
is related to stress (Chrousos, 1998). Thus, the LPP amplitude might be
a useful index of the increase or decrease of stress, especially in cases The participants arrived 3 h before the experiment and rested (for
where the arousal level is increased by emotionally unpleasant stimuli. example, read a book) without cigarette smoking during this time.
Using this characteristic of the LPP, we conducted 2 experiments After that, they moved to the testing room and sat on chairs which
with smokers. In order to collect baseline data, in Experiment 1 we were placed approximately 80 cm from a monitor (17-inch monitor,
assessed how cigarette smoking in an emotionally neutral state changes 1024 × 768 resolution). EEG and electrooculography (EOG) sensors
the LPP amplitude. To do so, we analyzed the ERP elicited when partic- were attached to the participants' scalps.
ipants watched emotionally neutral images before and after cigarette Fig. 1 shows the procedures for Experiment 1 and Experiment 2. In
smoking, or while just resting. In Experiment 2, we assessed how Experiment 1 (Fig. 1A), 20 neutral images were presented for 3 min,
smoking a cigarette in a stressful state changes the LPP amplitude. To after which the participants assessed their subjective ratings. The partic-
do so, we recorded the ERP while the participants watched emotionally ipants then took a 3-minute break, in which they smoked a cigarette
unpleasant images. After that, participants smoked cigarettes (or just (smoking condition) or simply rested without smoking a cigarette
rested without cigarette smoking), and the ERP was then measured (non-smoking condition). The cigarettes were the individual
while participants watched emotionally neutral images. We hypothe- participant's personal usual brand. After the break, 20 neutral images
sized that if cigarette smoking actually reduces the arousal level, the (different from the images presented before the break) were presented
LPP amplitude elicited by the unpleasant images would be decreased
more by cigarette smoking than by resting.
In addition to the ERP, we measured the participants' subjective
state before and after cigarette smoking (or resting) by using the Profile
of Mood States (POMS) (McNair et al., 1971). The POMS assesses mood
states mainly related to anxiety along 6 dimensions (for details, please
refer to the Methods section) and has been used in previous studies
on the effect of cigarette smoking on stress (for example, Pomerleau
and Pomerleau, 1987). We hypothesized that the participants would re-
port less anxiety after cigarette smoking than after resting. By measur-
ing both physiological (LPP) and psychological (POMS) responses, we
expected that our study would offer helpful data to examine whether
cigarette smoking actually has the effect of reducing stress.

2. Methods

2.1. Participants

Fourteen young, male, Japanese smokers (right-handed undergrad-

uate or graduate students; age range 21–25 years) participated in the
2 experiments. The participants were the same in both Experiments 1
and 2. Thirteen participants were included in the final analysis, since Fig. 1. Procedures of Experiment 1 (A) and Experiment 2 (B).
472 D. Choi et al. / International Journal of Psychophysiology 98 (2015) 470–476

Fig. 2. Grand-averaged event-related potential (ERP) waveforms (Pz site) in smoking condition (thick line) and non-smoking condition (thin line). The first and second rows show ERP
waveforms from Experiment 1 and Experiment 2, respectively. The left and right columns indicate before and after the break, respectively.

for 3 min, and the participants again assessed their subjective ratings. artifacts N 50 μV. The target stimulus presentation of − 200 to
The smoking and non-smoking conditions were conducted on different 2000 ms was averaged for each Session (before and after the break) in
days, and the order of the conditions was counterbalanced. each Condition. Prior to averaging, individual trial waveforms were
The procedure for Experiment 2 (Fig. 1B) was the same as that for baseline corrected (− 200 to 0 ms). The mean number of trials was
Experiment 1, except that the unpleasant images were presented before 27.4 (standard deviation (SD) = 8.1). Following the stimulus onset,
the break for 3 min as stress stimuli. The same neutral images were pre- the LPP was scored as the averaged activity in 4 time windows:
sented after the break in both experiments. Experiment 2 was conduct- 200–400 ms, 400–600 ms, 600–1000 ms, and 1000–2000 ms. We fo-
ed approximately 3 months after Experiment 1. The participants' brands cused on the LPP at the Pz site, since it has been reported to be maximal
of cigarettes were the same in both experiments. at this site (for example, Cuthbert et al., 2000).
The protocol for the image presentation was the same before and
after the break in both Experiments 1 and 2. The 20 IAPS images were 2.5. Statistical analysis
presented twice, and yellow circles were presented 10 times (a total
of 50 trials). The yellow circles were presented to sustain attention, For the LPP, we conducted a repeated measures analysis of variance
and the participants were asked to press a button with their right (ANOVA) with Time Window (200–400 ms, 400–600 ms, 600–1000 ms,
hand as quickly as possible in response to the circles. In each trial, a and 1000–2000 ms), Condition (smoking versus non-smoking), and Ses-
cross shape was presented (500 ms), followed by an IAPS image or a yel- sion (before versus after the break) as within-subject factors. The Green-
low circle (2000 ms). The interstimulus interval time was 1000 ms. On house–Geisser correction was applied where sphericity was violated.
our analysis, we focused only on responses to the IAPS images. For the POMS scores, we conducted a repeated measures ANOVA, with
For the subjective ratings, the participants completed the brief Condition and Session as within-subject factors. When an interaction
Japanese version (Yokoyama, 2005) of the POMS (McNair et al., 1971) was significant, pairwise comparisons were performed with the
with 6 subscales: tension–anxiety, depression, anger–hostility, vigor, fa- Bonferroni correction. Statistical significance was accepted at the 5%
tigue, and confusion. The ratings included 30 adjectives (for example, level (p b 0.05) (SPSS, Chicago, IL, USA).
anxious, angry, active, and sad), and the participants were asked to eval-
uate their moods on a 5-point scale, ranging from 0 (“not at all”) to 4 3. Results
(“quite frequently”).
3.1. LPP
2.4. ERP measurements and analysis
Fig. 2 illustrates the grand-averaged ERP waveforms elicited at the Pz
While the participants watched the images, EEG signals were re- site. The differences in the waveforms between the smoking and non-
corded at the Fz (medial frontal), Cz (medial central), and Pz (medial smoking conditions are relatively clearer after rather than before the
parietal) sites, based on the International 10–20 System (Towle et al., break. Table 1 summarizes the statistical results of the LPP responses.
1993), using a Polymate AP1532 system (TEAC, Tokyo, Japan). The elec- In Experiment 1, there was reliable interaction of Time Window and
trodes were referenced to averaged ears, and a ground electrode was at- Session (Table 1). Pairwise comparisons between Sessions within each
tached to the middle of the forehead. EOG was recorded to detect Time Window (critical p value = 0.0125 for 4 comparisons) did not re-
blinking, with electrodes placed above and below the right eye. All elec- veal significant difference (200–400 ms, p = 0.046; 400–600 ms, p =
trode impedances were below 10 k Ω. 0.028; 600–1000 ms, p = 0.687; 1000–2000 ms, p = 0.704). The inter-
The EEG signals were digitized at a sampling rate of 500 Hz and were action of Condition and Session was also reliable (Table 1). As shown in
amplified (band pass, 0.1–40 Hz) using the EMSE Suite (Source Signal Fig. 3A, pairwise comparisons between Sessions within each Condition
Imaging, San Diego, CA). We excluded trials containing (critical p value = 0.025 for 2 comparisons) revealed that the LPP
 D. Choi et al. / International Journal of Psychophysiology 98 (2015) 470–476 473

significantly decreased from before to after the break (p = 0.015) in the physiological responses. In order to compensate for the weak points in
smoking condition, but not in the non-smoking condition (p = 0.815). these studies, our study adapted the LPP as the physiological index
In Experiment 2, there were main effects of Time Window and Ses- and examined how cigarette smoking changes the LPP amplitude. The
sion, and reliable interaction of Time Window and Session (Table 1). subjective ratings (POMS) were also measured before and after ciga-
Pairwise comparisons between Sessions within each Time Window (crit- rette smoking.
ical p value = 0.0125 for 4 comparisons) revealed that the LPP signifi-
cantly decreased from before to after the break in time windows 4.1. LPP
200–400 ms (p = 0.003) and 400–600 ms (p = 0.000), but not in
time windows 600–1000 ms (p = 0.135) and 1000–2000 ms (p = Before discussing the effect of cigarette smoking in the unpleasant
0.445). Unlike Experiment 1, there was no reliable interaction of Condi- state on the LPP (Experiment 2), the effect of cigarette smoking in the
tion and Session in Experiment 2 (Table 1). However, for comparison be- emotionally neutral state in Experiment 1, in order to establish baseline
tween Experiments 1 and 2, we conducted pairwise comparisons data, should be discussed. In Experiment 1, we found different patterns
between Sessions within each Condition (critical p value = 0.025 for 2 of change of the LPP between after smoking and non-smoking; the LPP
comparisons). As shown in Fig. 3B, these revealed that the LPP signifi- was decreased by smoking, but not by non-smoking. Given that the LPP
cantly decreased from before to after the break in both the smoking con- amplitude reflects the arousal level (Cuthbert et al., 2000; Schupp et al.,
dition (p = 0.001) and the non-smoking condition (p = 0.012). In 2000; Weinberg et al., 2012), the results suggest that cigarette smoking
Experiment 2, there was also reliable interaction of Time Window, Con- in the emotionally neutral state decreases the arousal level, while non-
dition, and Session (Table 1). As shown in Fig. 4A, pairwise comparisons smoking does not. Since emotionally neutral images were presented be-
among Sessions within each Condition in each Time Window (critical p fore and after the break in Experiment 1, it would be difficult to inter-
value = 0.00625 for 8 comparisons) revealed that the LPP significantly pret this result as an effect of smoking on stress reduction. However, it
decreased from before to after the break for time window 200–400 ms might be suggested that cigarette smoking in the neutral state reduces
(p = 0.002) in the smoking condition, but not in the non-smoking con- the arousal level and thus results in relaxation.
dition (p = 0.090). However, for the time window 400–600 ms (Fig. In Experiment 2, unpleasant images were presented before the
4B), the LPP significantly decreased from before to after the break in break, and emotionally neutral images were presented after the break.
both the smoking and the non-smoking conditions (all p b 0.001). For Since many previous ERP studies have reported that the LPP is greater
time windows 600–1000 ms and 1000–2000 ms, the LPP did not signif- in response to emotional (pleasant and unpleasant) stimuli than to
icantly change from before to after the break in the smoking or the non- emotionally neutral stimuli (for example, Cuthbert et al., 2000;
smoking conditions (all p ≥ 0.05). Schupp et al., 2000; Weinberg et al., 2012), it was expected that the
LPP amplitude would decrease after the break. The results showed
3.2. POMS that the LPP amplitude for the time window 400–600 ms was decreased
both by smoking and non-smoking; however, the LPP amplitude for the
Table 2 shows the ratings for each subscale of the POMS in Experi- time window 200–400 ms was decreased only by cigarette smoking. A
ment 1 and Experiment 2. In Experiment 1, the tension–anxiety sub- possible interpretation for this result is that the high arousal level
scale showed a significant main effect of Session (F(1,12) = 10.400, evoked by watching unpleasant images decreased at a faster rate in
p = 0.007) and a reliable interaction (F(1,12) = 6.353, p = 0.027). the smoking condition than in the non-smoking condition. Thus, ciga-
Pairwise comparisons between Sessions within each Condition (critical rette smoking in the unpleasant state might facilitate decreased stress
p value = 0.025 for 2 comparisons) indicated that subjective tension– levels more than in the resting (non-smoking) state, given that there
anxiety decreased from before to after the break in the smoking condi- is a close relationship between arousal level and stress (Chrousos,
tion (p = 0. 003), but not in the non-smoking condition (p = 0.544). 1998; Winsky-Sommerer et al., 2005). Thus, the Experiment 2 results
In Experiment 2, the tension–anxiety, depression, and anger– suggest that cigarette smoking is more effective in relieving stress
hostility POMS subscales showed a significant main effect of Session than resting without smoking.
(tension–anxiety: F(1,12) = 5.149, p = 0.043; depression: F(1,12) = In Experiment 2, the LPP amplitude for the time windows
9.432, p = 0.010; anger–hostility: F(1,12) = 5.571, p = 0.036), indicat- 600–1000 ms and 1000–2000 ms was not affected by Condition or Ses-
ing that the subjective ratings had decreased from before to after the sion. According to previous studies (Weinberg and Hajcak, 2011;
break. However, no subscale showed any significant interaction or Weinberg et al., 2012), the earlier part of the LPP reflects capture of at-
main effect of Condition (p N 0.05). tention in a relatively obligatory manner, whereas the later part reflects
relatively sustained attention and elaborate processing of stimuli. This
4. Discussion suggests that cigarette smoking affects the earlier attention process,
but not the later sustained process. Future ERP studies are needed to
Previous studies on the effect of cigarette smoking on stress reduc- confirm this issue.
tion have reported a disparity between the subjective ratings and the There are some differences between the present study and the pre-
vious studies that assessed changes in the ERP responses to cigarette
smoking (for example, Domino, 2003; Houlihan et al., 1996; Ilan and
Table 1
Analysis of variance (ANOVA) effects of ERP responses.
Polich, 2001). For example, in an ERP study by Houlihan et al. (1996),
Significant p-values are in bold. the P300 was measured while participants completed a visual oddball
task, both before and after cigarette smoking. The oddball task indicates
Factor df(h,e) Experiment 1 Experiment 2
a paradigm in which 2 or 3 types are presented in different probability,
F P F P and the individual has to press a button as soon as possible in response
Time window 3, 36 2.041 0.167 13.598 0.001 to a stimulus with a low presentation probability (i.e., the target stimu-
Condition 1, 12 0.001 0.973 0.846 0.376 lus). In the study by Houlihan et al. (1996), the target stimulus was the
Session 1, 12 3.437 0.088 18.643 0.001
letter “X” and the non-target stimulus was the letter “O”; the results in-
Time window × Condition 3, 36 0.876 0.463 0.597 0.621
Time window × Session 3, 36 3.878 0.017 16.693 0.000 dicated that the P300 latency to the target stimulus decreased after cig-
Condition × Session 1, 12 8.329 0.014 2.592 0.133 arette smoking. This suggests that cigarette smoking improved
Time window × Condition × Session 3, 36 0.711 0.552 5.590 0.019 cognitive performance, in accordance with other ERP studies on ciga-
Note. df(h,e): degrees of freedom. The Greenhouse–Geisser correction was applied where rette smoking (Pritchard et al., 2004), given that the P300 latency
sphericity was violated; the uncorrected degrees of freedom are reported in this table. evoked by the oddball task usually reflects the degree of difficulty of
474 D. Choi et al. / International Journal of Psychophysiology 98 (2015) 470–476

Fig. 3. The late positive potential (LPP) in Experiment 1 (A) and Experiment 2 (B). Amplitude of the LPP was averaged across all time windows. Mean and standard errors: a, p = 0.015; b,
p = 0.012; c = p = 0.001 (pairwise comparisons between before and after the break in each Condition; Bonferroni-corrected critical p-value: 0.025).

the task. In the present study, we also conducted the oddball task; how- previous studies have reported on Nesbitt's paradox issue; although par-
ever, we analyzed only responses to non-target stimuli (i.e., neutral or ticipants reported that cigarette smoking made them feel less anxious and
unpleasant images), not responses to target stimuli (i.e., a yellow circle). stressed (Nesbitt, 1973; Pomerleau and Pomerleau, 1987; Perkins et al.,
This was because we focused on changes in the emotional state between 1992), the results of peripheral processes, such as the heart rate, indicated
before and after smoking, rather than changes in cognitive performance. that cigarette smoking and stress have common physiological responses
The images that were presented as non-target stimuli in the present (Pomerleau and Pomerleau, 1987; Perkins et al., 1992; Woodson et al.,
study were more related to emotion (pleasant versus unpleasant; re- 1986). The subjective ratings (POMS) and physiological responses (LPP)
laxed versus aroused) than the letters presented in the previous study from our study indicate the same results, which again sheds light on the
(i.e., “O” and “X”) (Houlihan et al., 1996). To the best of our knowledge, Nesbitt's paradox issue.
there has been no ERP study that has tested the effect of cigarette However, in Experiment 2, we could not find a different pattern in
smoking on emotion evoked by watching images. Thus, we expect changes of the subjective ratings between the smoking and non-
that the present study expands the availability of ERP in studies on ef- smoking conditions; tension–anxiety, depression, and anger–hostility
fects of cigarette smoking not only as an index of cognitive performance, were all decreased by both smoking and non-smoking. This is a some-
but also as an index of emotion and stress. what different result from the LPP results, which suggested that ciga-
rette smoking in the unpleasant state might facilitate decreased stress
4.2. Subjective ratings levels more than non-smoking. One possible reason for this incongruity
between the subjective ratings and the LPP responses might be that the
In addition to the LPP, we measured the participants' mood state by ERP is able to reflect very early and subtle changes in arousal between
using the POMS. In Experiment 1, the results revealed that the smoking and resting, while the self-report is not. Indeed, the difference
subjective tension and anxiety levels decreased by smoking, but not by in the LPP between smoking and non-smoking was shown in very short
non-smoking. This supports the results of the previous studies, which re- time windows (i.e., 200–400 ms). Another possible reason for the dis-
ported that smoking decreases subjective anxiety and stress (Nesbitt, parity between the subjective ratings and the LPP response in Experi-
1973; Pomerleau and Pomerleau, 1987; Perkins et al., 1992). More impor- ment 2 might be that the images presented as stress stimuli were too
tantly, this is compatible with the results of the LPP shown in Experiment unpleasant, and thus almost the same level of decrease of subjective
1, which suggested that the physiological arousal level decreased only by anxiety between smoking and non-smoking might have resulted. Fu-
smoking, not by non-smoking. As mentioned in the Introduction, the ture studies that apply other stress stimuli (i.e., less unpleasant images

Fig. 4. The late positive potential (LPP) for time windows 200–400 ms (A) and 400–600 ms (B) in Experiment 2. Mean and standard errors: a, p = 0.002; b, p b 0.001 (pairwise comparisons
between before and after the break in each Condition; Bonferroni-corrected critical p-value: 0.006).
 D. Choi et al. / International Journal of Psychophysiology 98 (2015) 470–476 475

Table 2 5. Conclusion
Ratings of the Profile of Mood States (POMS).
Mean and standard errors.
The aim of our study was to investigate whether cigarette smoking
Smoking condition Non-smoking relieves stress and reduces unpleasant emotions by using the LPP, a
condition component of the ERP, which reflects the arousal level. In Experiment
Before After Before After 1, the LPP in the emotionally neutral state decreased only by smoking,
break break break break suggesting that cigarette smoking in the neutral state reduced the
Experiment 1 Tension–anxiety 3.8 (0.8) 1.3 (0.5) 2.5 (0.7) 2.8 (0.9) arousal level. In Experiment 2, smoking in the unpleasant state de-
Depression 0.5 (0.2) 0.2 (0.2) 1.2 (0.5) 1.2 (0.6) creased the LPP at a faster rate (time window 200–400 ms) than non-
Anger–hostility 0.3 (0.2) 0.2 (0.1) 0.7 (0.4) 1.2 (0.7) smoking, suggesting that cigarette smoking in the unpleasant state facil-
Vigor 1.5 (0.5) 1.4 (0.5) 0.9 (0.4) 0.5 (0.3)
Fatigue 3.3 (1.1) 2.3 (0.7) 3.6 (1.0) 3.8 (1.1)
itates a decrease in the arousal level. Taken together, the results of the
Confusion 1.8 (0.5) 1.6 (0.4) 1.5 (0.4) 1.7 (0.6) LPP shown in the present study indicate that cigarette smoking might
Experiment 2 Tension–anxiety 2.6 (0.7) 1.7 (0.7) 2.8 (0.9) 1.2 (0.7) decrease the arousal level and perhaps relieve stress for smokers.
Depression 1.8 (0.6) 0.5 (0.4) 1.9 (0.7) 0.9 (0.5)
Anger–hostility 1.2 (0.6) 0.6 (0.4) 0.8 (0.4) 0.3 (0.3)
Vigor 1.5 (0.5) 1.1 (0.5) 0.9 (0.3) 1.5 (0.6) Acknowledgments
Fatigue 3.4 (0.8) 3.0 (0.7) 4.0 (1.1) 4.4 (1.6)
Confusion 1.4 (0.3) 1.2 (0.5) 1.1 (0.3) 1.1 (0.4)
The authors sincerely thank Mutsuhiro Fujiwara for his technical as-
sistance with this study. We also thank the participants of the study.
This research was supported by grants from Kyushu University (grant
than those presented in Experiment 2) are needed to address this number: LAKF 620212).
question.
References

4.3. Limitations and future directions Bauer, T., Göhlmann, S., Sinning, M., 2007. Gender differences in smoking behavior. Health
Economics 16 (9), 895–909. http://dx.doi.org/10.1002/hec.1259.
Bennett, M.P., Zeller, J.M., Rosenberg, L., McCann, J., 2003. The effect of mirthful laughter
Several limitations must be considered when interpreting the results on stress and natural killer cell activity. Nursing Faculty Publications 9, 38–45.
of the current study. First, it is not possible to conclude whether Chrousos, G.P., 1998. Stressors, stress, and neuroendocrine integration of the adaptive re-
sponse: the 1997 Hans Selye Memorial Lecture. Ann N Y Acad Sci 851 (1), 311–335.
smoking generally has an effect to reduce stress levels, since our partic-
http://dx.doi.org/10.1111/j.1749-6632.1998.tb09006.x.
ipants included only smokers, not non-smokers. Previous findings have Cook, M.R., Gerkovich, M.M., Hoffman, S.J., McClernon, F.J., O'Connell, K.A., 1996. Effects of
suggested that smokers have higher baseline stress levels than non- smoking and telic/paratelic dominance on the contingent negative variation (CNV).
smokers. For example, in the study by Parrott (1995), smokers reported Int J Psychophysiol. 23 (1–2), 101–110. http://dx.doi.org/10.1016/0167-
8760(96)00043-8.
lower stress levels after smoking than before smoking (baseline), while Cuthbert, B.N., Schupp, H.T., Bradley, M.M., Birbaumer, N., Lang, P.J., 2000. Brain potentials
nonsmokers reported similar baseline stress levels to the stress levels in affective picture processing: covariation with autonomic arousal and affective re-
reported by smokers after smoking. Additionally, as mentioned in the port. Biol Psychol 52 (2), 95–111. http://dx.doi.org/10.1016/S0301-0511(99)00044-
7.
Introduction, smokers have higher neuroticism and anxiety traits than Domino, E.F., 2003. Effects of tobacco smoking on electroencephalographic, auditory
non-smokers (McCrae et al., 1978), suggesting that smokers also have evoked and event related potentials. Brain Cogn 53 (1), 66–74. http://dx.doi.org/10.
higher baseline arousal levels than nonsmokers. Thus, the reason why 1016/S0278-2626(03)00204-5.
Dozois, D.N., Farrow, J.A., Miser, A., 1995. Smoking patterns and cessation motivations
smoking reduces the arousal level and perhaps the stress level in during adolescence. Int J Addict 30 (11), 1485–1498. http://dx.doi.org/10.3109/
smokers might be simply because smoking reduces the stress levels of 10826089509055844.
smokers to the same level as that of non-smokers, and not because File, S.E., Fluck, E., Leahy, A., 2001. Nicotine has calming effects on stress-induced
mood changes in females, but enhances aggressive mood in males. Int J
smoking generally has an effect on stress levels. Future studies are re-
Neuropsychopharmacol 4 (4), 371–376. http://dx.doi.org/10.1017/S1461145701002577.
quired using the same protocol as the present study for both smokers Gilbert, D.G., Gilbert, B.O., 1995. Personality, psychopathology, and nicotine response as
and non-smokers, in order to examine whether smoking generally has mediators of the genetics of smoking. Behav Genet 25 (2), 133–147. http://dx.doi.
org/10.1007/BF02196923.
an effect to reduce stress levels.
Hajcak, G., MacNamara, A., Olvet, D.M., 2010. Event-related potentials, emotion, and emo-
Second, all of the participants were males. Since there are reported tion regulation: an integrative review. Dev Neuropsychol 35 (2), 129–155. http://dx.
gender differences in smoking behavior (Bauer et al., 2007; File et al., doi.org/10.1080/87565640903526504.
2001; Perkins et al., 1992), it is difficult to predict whether female Hajcak, G., Weinberg, A., MacNamara, A., Foti, D., 2012. ERPs and the study of emotion. In
In: Luck, S.J., Kappenman, E.S. (Eds.), Oxford Handbook of Event-related Potential
smokers would show the same results. Thus, future studies need to be Components. New York: Oxford University Press, pp.441–474.
done with female smokers, using the same experiment protocol as the Hasenfratz, M., Michel, C., Nil, R., Bättig, K., 1989. Can smoking increase attention in rapid
present study. information processing during noise? Electrocortical, physiological and behavioral ef-
fects. Psychopharmacology (Berl) 98 (1), 75–80. http://dx.doi.org/10.1007/
Third, in our study, the participants rested without doing anything as BF00442009.
a control condition. Previous studies investigating the effect of smoking Houlihan, M.E., Pritchard, W.S., Robinson, J.H., 1996. Faster P300 latency after smoking in
have adapted both a pre/post smoking versus pre/post resting design visual but not auditory oddball tasks. Psychopharmacology 123 (3), 231–238. http://
dx.doi.org/10.1007/BF02246577.
(Hasenfratz et al., 1989; Michel et al., 1987) and a pre/post smoking ver- Ilan, A.B., Polich, J., 2001. Tobacco smoking and event-related brain potentials in a Stroop
sus pre/post sham smoking design (Cook et al., 1996). However, some task. Int J Psychophysiol 40 (2), 109–118. http://dx.doi.org/10.1016/S0167-
authors (for example, Pritchard et al., 2004) have argued that merely 8760(00)00156-2.
Kassel, J.D., Stroud, L.R., Paronis, C.A., 2003. Smoking, stress, and negative affect: correla-
resting is not thought to have the same sensory or motor components
tion, causation, and context across stages of smoking. Psychol Bull 129 (2),
as smoking. Thus, future studies would need to use identical- 270–304. http://dx.doi.org/10.1037/0033-2909.129.2.270.
appearing denicotinized cigarettes or gum in the non-smoking Lang, P.J., Bradley, M.M., Cuthbert, B.N., 2008. International affective picture system
(IAPS): affective ratings of pictures and instruction manual. Technical Report A-8.
condition.
University of Florida, Gainesville, FL.
Finally, we did not measure heart rates; it was therefore impossible McCrae, R.R., Costa, P.T., Bossé, R., 1978. Anxiety, extraversion and smoking. Br J Soc Clin
to compare changes in peripheral responses between the current study Psychol 17 (3), 269–273. http://dx.doi.org/10.1111/j.2044-8260.1978.tb00277.x.
and the previous studies. Future studies would need to measure brain McKee, S.A., Sinha, R., Weinberger, A.H., Sofuoglu, M., Harrison, E.L., Lavery, M., Wanzer, J.,
2011. Stress decreases the ability to resist smoking and potentiates smoking intensity
activity and heart rate simultaneously, in order to obtain rich physiolog- and reward. J Psychopharmacol 25 (4), 490–502. http://dx.doi.org/10.1177/
ical data to resolve Nesbitt's paradox. 0269881110376694.
476 D. Choi et al. / International Journal of Psychophysiology 98 (2015) 470–476

McNair, D.M., Lorr, M., Droppleman, L.F., 1971. Manual for the Profile of Mood States. Ed- Pritchard, W., Sokhadze, E., Houlihan, M., 2004. Effects of nicotine and smoking on event-
ucational and Industrial Testing Services, San Diego, CA. related potentials: a review. Nicotine Tob Res 6 (6), 961–984. http://dx.doi.org/10.
Mendelson, T., Greenberg, M.T., Dariotis, J.K., Gould, L.F., Rhoades, B.L., Leaf, P.J., 2010. Fea- 1080/14622200412331324848.
sibility and preliminary outcomes of a school-based mindfulness intervention for Schupp, H.T., Cuthbert, B.N., Bradeley, M.M., Cacioppo, J.T., Ito, T., Lang, P.J., 2000. Affective
urban youth. J. Abnorm. Child Psychol. 38 (7), 985–994. picture processing: the late positive potential is modulated by motivational rele-
Michel, C., Nil, R., Buzzi, R., Woodson, P.P., Bättig, K., 1987. Rapid information processing vance. Psychophysiology 37 (2), 257–261. http://dx.doi.org/10.1111/1469-8986.
and concomitant event-related brain potentials in smokers differing in CO absorp- 3720257.
tion. Neuropsychobiology 17 (3), 161–168. http://dx.doi.org/10.1159/000118356. Towle, V.L., Bolaños, J., Suarez, D., Tan, K., Grzeszczuk, R., Levin, D.N., Cakmur, R., Frank,
Nesbitt, P.D., 1973. Smoking, physiological arousal, and emotional response. J Pers Soc S.A., Spire, J.P., 1993. The spatial location of EEG electrodes: locating the best-fitting
Psychol 25 (1), 137–144. http://dx.doi.org/10.1037/h0034256. sphere relative to cortical anatomy. Electroencephalogr Clin Neurophysiol 86 (1),
Nichter, M., Vuckovic, N., Quintero, G., Ritenbaugh, C., 1997. Smoking experimentation 1–6. http://dx.doi.org/10.1016/0013-4694(93)90061-Y.
and initiation among adolescent girls: qualitative and quantitative findings. Tob Con- Warburton, D.M., Revell, A.D., Thompson, D.H., 1991. Smokers of the future. Br J Addict 86
trol 6 (4), 285–295. http://dx.doi.org/10.1136/tc.6.4.285. (5), 621–625. http://dx.doi.org/10.1016/S0306-4603(02)00249-6.S0306460302002496.
Olofsson, J.K., Nordin, S., Sequeira, H., Polich, J., 2008. Affective picture processing: an in- Weinberg, A., Hilgard, J., Bartholow, B.D., Hajcak, G., 2012. Emotional targets: evaluative
tegrative review of ERP findings. Biol Psycho 77 (3), 247–265. http://dx.doi.org/10. categorization as a function of context and content. Int J Psychophysiol 84 (2),
1016/j.biopsycho.2007.11.006. 149–154. http://dx.doi.org/10.1016/j.ijpsycho.2012.01.023.
Parrott, A.C., 1995. Stress modulation over the day in cigarette smokers. Addiction 90, Weinberg, A., Hajcak, G., 2011. The late positive potential predicts subsequent interfer-
233–244. http://dx.doi.org/10.1046/j.1360-0443.1995.9022339.x. ence with target processing. J CognNeurosci 23 (10), 2994–3007. http://dx.doi.org/
Pelletier, C.L., 2004. The effect of music on decreasing arousal due to stress: A meta- 10.1162/jocn.2011.21630.
analysis. Journal of Music Therapy 41 (3), 192–214. http://dx.doi.org/10.1093/jmt/ Winsky-Sommerer, R., Boutrel, B., de Lecea, L., 2005. Stress and arousal. Molecular neuro-
41.3.192. biology 32 (3), 285–294. http://dx.doi.org/10.1385/MN:32:3:285.
Perkins, K.A., Grobe, J.E., Fonte, C., Breus, M., 1992. “Paradoxical” effects of smoking on Woodson, P.P., Buzzi, R., Nil, R., Báuttig, K., 1986. Effects of smoking on vegetative reactiv-
subjective stress versus cardiovascular arousal in males and females. Pharmacol ity to noise in women. Psychophysiology 23 (3), 272–282. http://dx.doi.org/10.1111/
Biochem Behav 42 (2), 301–311. http://dx.doi.org/10.1016/0091-3057(92)90531-J. j.1469-8986.1986.tb00632.x.
Pomerleau, C.S., Pomerleau, O.F., 1987. The effects of a psychological stressor on cigarette Yokoyama, K., 2005. Manual of the Japanese Version of Profile of Mood States — Brief
smoking and subsequent behavioral and physiological responses. Psychophysiology Form. Kaneko Shobo, Tokyo (in Japanese).
24 (3), 278–285. http://dx.doi.org/10.1111/j.1469-8986.1987.tb00295.x.