Psychopharmacology (2005) 182: 508

–515

DOI 10.1007/s00213-005-0110-8

O R I G I N A L I N V E S T I G ATI O N

Yu Ohmura . Taiki Takahashi . Nozomi Kitamura

Discounting delayed and probabilistic monetary gains
and losses by smokers of cigarettes

Received: 6 February 2005 / Accepted: 22 June 2005 / Published online: 16 September 2005

# Springer-Verlag 2005

Abstract

Rationale: Nicotine dependence has been

associated with impulsivity and discounting delayed/un-
certain outcomes. Objectives: This study had two main
objectives: (1) to examine the relationship between the
number of cigarettes consumed per day and the degree to
which delayed and uncertain monetary gains and losses
are discounted by smokers, and (2) to determine the re-
lationship between the estimated dose of nicotine intake
per day and the degree to which four types of discounting
occur. Methods: Twenty seven habitual smokers and 23
never smokers participated in this experiment. They were
required to choose between immediate and delayed mone-
tary rewards (or losses), or between guaranteed and prob-
abilistic rewards (or losses). Results: The degree to which
delayed monetary gains were discounted was significant-
ly and positively correlated with both the number of cig-
arettes smoked and the estimated dose of nicotine intake
per day. Conversely, there was no relationship between
smoking and the remaining three types of discounting.
Also, mild smokers in our sample did not differ from
never smokers in discounting monetary gains or losses.
Conclusions: In general, our results suggest that both the
frequency of nicotine self-administration, as well as the
dosage, are positively associated with greater delay dis-
counting of gains. One neuropsychopharmacological ex-
planation for this effect is that chronic nicotine intake
may induce neuroadaptation of the neural circuitry in-
volved in reward processing.

Keywords Addiction . Delay discounting . Probability
discounting . Cigarette smoking . Nicotine . Impulsivity .
Neuroeconomics . Intertemporal choice

Introduction

Impulsive behavior, broadly defined as

“actions that are

poorly conceived, prematurely expressed, unduly risky, or
inappropriate to the situation and that often result in un-
desirable outcomes

” (Daruna and Barnes

1993

), are fre-

quently observed in drug-dependent subjects (see Bickel
and Marsch

2001

for a review). In most psychopharma-

cological studies of intertemporal choice, impulsivity is
often operationalized in terms of delay discounting

—the

tendency to choose smaller, relatively immediate rewards
over larger but more delayed rewards (e.g., Kirby et al.

1999

; Richards et al.

1999b

; Petry

2001

; Pietras et al.

2003

).

Dependence on drugs such as cocaine, heroin, nicotine,

or alcohol has been associated with greater discounting of
delayed rewards in a number of psychopharmacological
studies (e.g., Kirby and Petry

2004

; Kirby et al.

1999

;

Bickel et al.

1999

; Petry

2001

). However, to date, little is

known regarding the relationship between the frequency of
drug administration, or the drug dosage, and delay discount-
ing. One notable exception is a recent study by Reynolds
(

2004

) reporting that the number of cigarettes consumed

per day was positively correlated with impulsive choice in
delay discounting. Understanding the dose-dependent rela-
tionship between a drug and delay discounting is critical
to (a) better estimate a drug

’s effect on impulsive behavior

in general, and intertemporal decision-making specifical-
ly, which can often result in problematic outcomes for the
drug user (see Bickel and Marsch

2001

for a review) and

(b) predict vulnerability to drug dependence as a function
of discounting behavior, as suggested by a previous study
(Perry et al.

2005

).

Recently, in the emerging field of neuroeconomics (see

Glimcher and Rustichini

2004

; Schultz

2004

for a review),

several neuroscientists and economists have collaborated to
reveal some of the neural substrates involved in economic
decision making, including those governing delay dis-
counting. For instance, we have reported that low cortisol
levels were associated with impulsive choice in delay
discounting (Takahashi

2004

) partly via modulation of the

Y. Ohmura (

*) . T. Takahashi . N. Kitamura

Department of Behavioral Science, Hokkaido University,
N10 W7 Kita-ku,
Sapporo, 060-0810, Japan
e-mail: [e-mail]

reward-processing, dopaminergic circuitry in the brain. In
one neuroimaging study, Loewenstein and Cohen

’s group

demonstrated that choosing a smaller, immediate mone-
tary reward was associated with activation of the reward-
processing dopaminergic circuitry located in the midbrain
(McClure et al.

2004

). Montague and Berns (

2002

) have

proposed that estimating future reward values is mediated
by dopaminergic circuitry (e.g., the ventral tegmental
area). Taken together, dopaminergic systems may play a
pivotal role in impulsive choice in delay discounting.

Concerning the relationship between chronic self-ad-

ministration of nicotine and dopaminergic response to
monetary gains, Schultz and his colleagues have reported a
lower dopaminergic response to monetary rewards in ha-
bitual cigarette smokers when compared to nonsmokers,
which may indicate an exaggerated devaluation of delayed
monetary rewards (i.e., greater delay discounting) in smok-
ers (Martin-Solch et al.

2001

,

2003

). Nevertheless, the re-

lationship between self-administration of a dopaminergic
drug such as nicotine and discounting behavior can be
better defined. Therefore, it is of neuropsychopharmaco-
logical interest to investigate delay discounting as a func-
tion of both the frequency of nicotine self-administration
and the strength of the dosage.

Moreover, although most studies have focused on

discounting delayed monetary gain, we feel it is important
to expand the current research design to include additional
forms of discounting. Neuroeconomic research has re-
vealed that distinct brain regions are activated in response
to monetary gains and losses (Knutson et al.

2000

,

2003

;

Breiter et al.

2001

). Therefore, in examining the relation-

ship between cigarette smoking and impulsive discounting
delayed outcomes, it is important to include the tendency
to discount delayed monetary losses in addition to the
discounting of delayed monetary gains. Furthermore, the
relationship between drug intake and the discounting of
uncertain rewards (probability discounting) has been at-
tracting more attention recently, but with mixed results.
For instance, Mitchell, a psychopharmacologist, failed to
observe a difference between smokers and never smokers
in discounting of uncertain rewards (Mitchell

1999

). On

the other hand, although there was no significant correla-
tion between probability discounting and breath CO levels
taken at the time of participation, smokers in the Reynolds
study discounted an uncertain monetary reward more dra-
matically when compared to never smokers (Reynolds
et al.

2004

). One possible reason for the discrepant find-

ings may have to do with the degree to which the samples
engaged in smoking: smokers in the Mitchell study con-
sumed as few as 15 cigarettes per day, whereas those in the
Reynolds study smoked more than 20 per day. Thus, an
elevation in probability discounting may only be observed
in relatively heavy smokers. Nevertheless, the inconsis-
tency of these findings suggests that the link between
smoking and probability discounting requires further in-
vestigation. It should also be noted that whether impulsive
behavior can be defined as strong probability discounting

is still controversial (cf. Myerson et al.

2003

). Again, con-

sidering that the neural responses involved in economic
gains and losses are distinct, it is important to examine
discounting of both uncertain monetary gains and losses
with regard to smoking frequency and nicotine dosage.

As far as we know, this study is the first to investigate the

relationship between the frequency and dosage of nicotine
self-administration and the tendency toward four types of
decision making (i.e., discounting of delayed and uncer-
tain monetary gains and losses) within the same subjects.
It should also be noted that, as far as we know, this study
is the first to examine discounting of uncertain monetary
losses in smokers of cigarettes. Furthermore, we examined
differences in four types of discounting between never
smokers and smokers. Additionally, to further elucidate
their distinct psychological processes (and possibly the
distinct neural mechanisms underlying these processes as
well), we compared discounting of gains and losses. Final-
ly, we examined whether a positive correlation is observed
between delay and probability discounting as expected
from the hypothesis that an increase in delay is equivalent
to a decrease in probability (Rachlin et al.

1991

).

Materials and methods

Participants

A total of 50 subjects participated in the present study,
including 27 young-adult habitual smokers (20 males and
7 females) between 21 and 33 years of age (M=24.15;
SD=3.68), and 23 never smokers (16 males and 7 females)
between 21 and 28 years of age (M=23.26; SD=1.96). It
should be noted that the present smoker population was
relatively mild in comparison to other studies (e.g., Bickel
et al.

1999

; Reynolds et al.

2004

; in their studies, only

heavy smokers who consumed a minimum of 20 ciga-
rettes per day were employed); only eight smokers con-
sumed a minimum of 20 cigarettes per day, and the mean
number of cigarettes consumed per day was 14.38 (Table

1

).

Graduate and undergraduate students were recruited to par-
ticipate through advertisements posted on bulletin boards
at Hokkaido University in Sapporo, Japan. The partici-
pants were informed that the experiment involved a de-
cision-making task involving monetary gains and losses.
They signed an informed consent form before participat-
ing and received 1,000 yen (about US $10) following
completion of the experiment.

Materials and procedure

Because Johnson and Bickel (

2002

) showed a strong cor-

relation between discounting rates for hypothetical and real
monetary gains, and Baker et al. (

2003

) demonstrated that

discounting rates for hypothetical and real money were not
significantly different, a computerized procedure consist-

509

ing of hypothetical monetary outcomes was used to assess
discounting in the laboratory. The procedure was com-
posed of four different types of discounting (i.e., delay of
gain, delay of loss, uncertain gain, and uncertain loss).

Participants were seated individually in a semisound-

proof room and received the following instruction on the
computer screen in Japanese:

[From now, you are required to perform tasks of
decision-making on monetary reward/loss. The task is
to choose between two options. The monetary reward/
loss in this experiment is hypothetical, but we want
you to think as though it is real money].

Next, they received instructions describing the four

discounting tasks with corresponding examples. At the
beginning of each trial, the participant was asked to select
one of two cards displayed on their computer monitor. The
left card indicated the sum of money that could be received
(or lost) immediately (or certainly, in the probability-dis-
counting tasks), whereas the right card always indicated
100,000 yen (about US $1,000) that could be received (or
lost) after a certain delay (or with a certain probability).

The sum of money indicated on the left card ranged from

100,000 to 5,000 yen (or from

−100,000 to −5,000 yen, in

loss-frame tasks) in 5,000-yen intervals. In the delay-
discounting task, the delay indicated on the right card
changed between five time frames (1 week, 1 month, 6
months, 1 year, and 5 years). For the probability-discount-
ing task, the right card indicated one of five probability
values (90, 70, 50, 30, and 10%). These changes were
computerized according to the algorithm used by Richards
et al. (

1999b

). This algorithm is designed to determine

the point at which the participant switches his or her pref-
erence from the left card (immediate/guaranteed reward or

loss) to the right card (delayed/probabilistic reward or loss)
by changing the type of task and the sum of money in
accordance with previous decisions. The switching point
is regarded as the indifference point and was used to cal-
culate the dependant variable. In the present study, 20
indifference points were determined (five for each type of
discounting). This algorithm masks the true nature of the
procedure, and in the present study, distractor trials were
inserted after ten indifference points were established (for
more details, see Richards et al.

1999b

).

Following the computer task, all participants were re-

quired to answer four questions regarding their smoking
behavior, including the number of months they had
smoked, the average number of cigarettes they smoked
per day, the variation (i.e., the minimum and maximum
number of cigarettes they smoked in a day), and their usual
brand of cigarettes. The entire experimental procedure took
between 30 and 60 min to complete. For no subject did the
variation in the number of cigarettes they smoked per day
exceed five cigarettes. It should also be noted that the
results remained essentially unchanged when the number
of cigarettes smoked per day was recoded into categories
of low (1

–10 cigarettes), medium (11–20 cigarettes), and

high (21 cigarettes or more) frequencies.

Data analysis

To parametrize the degree to which each subject discounted
delayed and uncertain monetary gains and losses, we com-
puted an area under the curve (AUC) for each of the four
discounting tasks (cf. Fig.

1

). The procedure for calculat-

ing an AUC was as follows (for more details, see Myerson
et al.

2001

). First, indifference points were plotted in two

dimensions, with either delay or odds-against [=(1/prob-
ability)

−1] (cf. Rachlin et al.

1991

) plotted along the

horizontal axis and gain (or the absolute value of loss)
plotted along the vertical axis. Connecting the individual
indifference points defined the indifference curve. Note
that the steepness of the indifference curve indicates the
degree to which the monetary outcomes were discounted
by the subject. Second, both the horizontal and vertical
scales were divided by the largest value on their respective
axes to produce a range of values between 0 and 1. Third,
the AUC was defined as the total area under this nor-
malized indifference curve. The smaller this area, the more
dramatic the subject

’s discounting tendency. We adopted

the use of AUCs primarily to avoid equation-dependent
systematic errors that can result from specific fitting func-
tions, something the AUC parameter does not depend on.

Additionally, although it is well established that delayed

and probabilistic gains are discounted hyperbolically rather
than exponentially (e.g., Mazur

1987

; Rachlin et al.

1991

;

Richards et al.

1999b

; Simpson and Vuchinich

2000

), we

compared the four types of discounting (delayed and un-
certain, gains and losses), to determine whether losses are
also discounted hyperbolically. Specifically, we performed
a nonlinear regression (SAS, PROC NLIN) to fit hyper-
bolic and exponential discounting functions to the indif-

Table 1 Mean and standard deviations for demographic variables,
smoking behavior, and AUCs for current and never smokers

Current smokers

Never smokers

Mean

SD

Mean

SD

Sex (% men)

69.57

74.07

Age (years)

24.15

3.68

23.26

1.96

Education (% graduate)

32.00

47.83

Cigarette number (per day)

14.38

6.66

Smoking history (months)

69.04

48.89

DD of gain

0.54

0.27

0.58

0.33

PD of gain

0.23

0.15

0.18

0.08

DD of loss

0.69

0.25

0.74

0.29

PD of loss

0.36

0.16

0.39

0.17

In calculating the AUC, the horizontal axis is delay (in delay

discounting) or odds-against (=1/probability

− 1, in probability

discounting), and the vertical axis is the indifference point. Note
that smaller AUC values correspond to greater discounting.
Regarding education and cigarette number, two data points
are missing due to the participants

’ omission in answering

these questions

DD Delay discounting, PD probability discounting

510

ference points at each level of delay and probability. The
exponential function was defined as

V ¼ Ae

kD

;

where V is the subjective value of a reward, A is the
(objective) amount of the reward (the monetary gain or
loss), k is a free parameter and an index of the steepness
of the discounting function (i.e., larger k values corre-
spond to steeper delay discounting), and D is the length of
the delay (delay discounting) or the odds-against (proba-
bility discounting). The hyperbolic function was defined as

V ¼ A= 1 þ kD

ð

Þ;

with the same notations. To determine which equation fits
the data better, we compared the respective R

2

values of

the hyperbolic and exponential equations.

The estimated amount of daily nicotine intake was cal-

culated by multiplying the average number of cigarettes
smoked per day by the amount of nicotine per cigarette
that was printed on the cigarette pack of the brand con-
sumed by the subject. It is noteworthy that the number of
cigarettes smoked per day indicates the frequency of nic-
otine self-administration, whereas the estimated amount
of nicotine intake per day indicates the level of chronic
nicotine exposure. To test statistical significance of corre-
lations and mean differences, Pearson

’s product–moment

correlation coefficients and t tests were utilized, respec-
tively. Alpha level was set at 5% throughout.

Results

Relationships involving demographics
and discounting behavior

Means and standard deviations for demographic variables,
smoking behavior, and AUCs for discounting are summa-
rized in Table

1

. Smoking and nonsmoking samples did

not differ in age, sex, or level of education, but there was a
significant difference in the average number of cigarettes
smoked per day between men and women (16.36 vs 9.29,
t (23)=2.67, P<0.05). The correlation between the average
number of cigarettes smoked per day and age was not
significant.

The present AUC values were similar to values reported

in previous studies (Myerson et al.

2001

). A participant

’s

age did not correlate with any of the AUCs for discount-
ing, and there were no significant differences between the
AUCs of men and women or between the AUCs of grad-
uate and undergraduate students. Thus, neither age, sex,
nor education seemed to have any effect on discounting
behavior in our sample.

Fitness of discounting equations

Hyperbolic and exponential R

2

values were calculated

using both group medians and individual scores. For the
group data, each R

2

value associated with the hyperbolic

function (delay discounting of gain, 0.99; probability
discounting of gain, 0.98; delay discounting of loss, 0.97;
and probability discounting of loss, 0.98) was larger than
its corresponding exponential function (delay discounting
of gain, 0.91; probability discounting of gain, 0.81; delay
discounting of loss, 0.96; and probability discounting of
loss, 0.84). Moreover, except when discounting delayed
gains, each individual R

2

value associated with the hyper-

bolic function was significantly larger than its correspond-
ing exponential function [dependent sample t tests: delay
discounting of gain t (39)=1.77, P=0.08; probability dis-
counting of gain t (39)=2.80, P<0.01; delay discounting of
loss t (33)=4.93, P<0.01; and probability discounting of
loss t (38)=4.82, P<0.01]. These results suggest that the
subjects discounted most types of monetary outcomes hy-
perbolically, rather than exponentially, as a number of
previous studies have reported (e.g., Rachlin et al.

1991

;

Vuchinich and Simpson

1998

; Richards et al.

1999b

;

Bickel et al.

1999

; Simpson and Vuchinich

2000

).

Relationships involving smoking status
and discounting behavior

First, we investigated the relationship between the fre-
quency of nicotine intake and discounting behavior in
smokers by calculating the Pearson product

–moment cor-

relation between the number of cigarettes smoked per day
and each of the four types of discounting. The correlational
analysis revealed that the AUC for discounting delayed

0

0.25

0.5

0.75

1

0

0.25

0.5

0.75

1

Delay (proportion of maximum)

Subjective Value

Area Under the Curve

Fig. 1 Calculation of the area under the indifference curve in delay
discounting of monetary gain, a linkage of subjective values of
delayed monetary gain, for representative example (Subject 12).
Note that in probability discounting, the horizontal axis is odds-

against (=1/probability

−1)

511

gains in smokers decreased significantly with the number
of cigarettes they smoked per day [r (25)=−0.66, P<0.01,
see Fig.

2

], suggesting that more frequent smokers are

more impulsive when discounting delayed monetary gains.
It should be noted that this finding is consistent with the
recent study by Reynolds (

2004

), which reported a posi-

tive correlation between the number of cigarettes smoked
per day and a delay discounting rate (logged hyperbolic k)
of monetary rewards. Regarding the number of cigarettes
smoked per day and the remaining three types of dis-
counting, no other significant correlations were observed
in the present study [probability discounting of gain r (25)=

−0.00059, P=0.9978; delay discounting of loss r (25)=−0.12,
P=0.58; and probability discounting of loss r (25)= 0.10,
P=0.63]. In addition, because (a) the number of male and
female smokers (20 males and 7 females) was different and,
(b) as stated earlier, there was a significant difference in
the average number of cigarettes smoked per day between the
sexes, we omitted the female participants and repeated the
correlational analysis. Again, the AUC for delay discounting
of gains in male smokers was significantly correlated with the
number of cigarettes smoked per day [r (18)=−0.69, P<0.01],
and no other significant correlations were observed.

Second, correlations between the level of chronic nic-

otine exposure and the degree of discounting similarly re-
vealed a significant relationship between the AUC for
discounting delayed gains by smokers and the estimated
amount of nicotine intake per day [r (25)=−0.57, P<0.01,
see Fig.

3

]. This indicates that subjects with higher levels

of chronic nicotine exposure were more impulsive in dis-
counting delayed monetary gains. Regarding the relation-
ships between the estimated amount of nicotine intake per
day and the remaining three types of discounting, no other
significant relationships were observed [probability dis-

counting of gain r (25)=0.22, P=0.29; delay discounting
of loss r (25)=−0.05, P=0.82; and probability discounting
of loss r (25)=−0.12, P=0.56]. When the analysis was
restricted to male participants, the results were the same;
the only significant correlation between nicotine intake per
day and discounting was that involving delay discounting
of gains by male smokers [r (18)=−0.56, P<0.05].

Additionally, the number of months the subject had

smoked was unrelated to his or her discounting behavior
[delay discounting of gain r (25)=−0.18, P=0.38; proba-
bility discounting of gain r (25)=−0.10, P=0.63; delay
discounting of loss r (25)=0.34, P=0.08; and probability
discounting of loss r (25)=0.21, P=0.30] but was signif-
icantly correlated with both the estimated nicotine intake
per day and the number of cigarettes smoked per day
(r (25)=0.46, P<0.05 and r (25)=0.42, P<0.05, respec-
tively). These findings are consistent with those reported
in the report of Reynolds (

2004

). Also, the number of cig-

arettes smoked per day was correlated with the amount of
nicotine per cigarette [r (25)=0.55, P<0.01]. This indicates
that habitual smokers who consume more cigarettes reg-
ularly tend to smoke stronger cigarettes. Consequently, the
number of cigarettes smoked per day was correlated with
the estimated amount of nicotine intake per day [r (25)=
0.87, P<0.01].

Group differences in the degree of discounting
between never smokers and smokers

To examine whether a previously reported group difference
between never smokers and heavy smokers was replicated
in the present study, we compared delay discounting of
monetary gains in smokers and never smokers and found

0

0.2

0.4

0.6

0.8

1

0

10 20

30

40

The number of cigarettes smoked per day

AUC of delay discounting of monetary gains

r = -0.66

Fig. 2 Scatterplot of the number of cigarettes smoked per day and
the AUC for delay discounting of monetary gains in smokers. A
significant negative correlation was observed (n=25, P<0.01). Note

that a smaller AUC indicates a higher degree of discounting. Two
data points are missing due to the participants

’ omission in an-

swering questions about their smoking behavior

0

0.2

0.4

0.6

0.8

1

0

10

20

30

40

Estimated nicotine intake per day

AUC of delay discounting of monetary gains

r = -0.57

Fig. 3 Scatterplot of the estimated amount of nicotine intake per
day and the AUC for delay discounting of monetary gains in
smokers. A significant negative correlation was observed (n=25,

P<0.01). Note that a smaller AUC indicates a higher degree of

discounting. Two data points are missing due to the participants

’

omission in answering questions about their smoking behavior

512

no significant difference [t (48)=0.48, P=0.63]. Moreover,
there was no difference between smokers and never
smokers in any of the other three types of discounting
[probability discounting of gain t (40.5)=−1.78, P=0.08;
delay discounting of loss t (48)=0.57, P=0.57; and
probability discounting of loss t (48)=0.78, P=0.44]. It
has already been noted that the smokers in our sample were
relatively mild nicotine users compared to those who have
participated in previous studies (e.g., Bickel et al.

1999

;

Mitchell

1999

; Reynolds et al.

2004

), which may have

resulted in a nonsignificant group difference.

Relationship between discounting gains
and discounting losses

In addition, we examined the relationship between dis-
counting a gain and discounting a loss across all 50 partic-
ipants for delay and probability discounting and confirmed
positive and negative relationships, respectively [delay
discounting r (50)=0.60, P<0.01; probability discounting
r (50)=−0.49, P<0.01]. Moreover, mean AUCs for dis-
counting gains were significantly smaller than those for
losses [dependent sample t tests: delay discounting t (49)=
4.34, P<0.01; probability discounting t (49)=4.66, P<
0.01]. This implies that people more steeply discount fu-
ture/uncertain gains than losses.

Relationship between delay discounting
and probability discounting

Finally, we analyzed the relationship between delay dis-
counting and probability discounting independently for
gains and losses across all 50 participants. For gains,
Pearson product

–moment correlations revealed a posi-

tive, but nonsignificant, relationship between the AUCs
of delay and probability discounting [r (50)=0.17, P=
0.23]. This observation is consistent with the study of
Myerson et al. (

2003

) reporting a relatively weak to modest

positive relationship between discounting of delayed and
uncertain gains. The correlation between discounting de-
layed and uncertain losses was likewise nonsignificant
[r (50)=0.16, P=0.28]. Together, these results suggest that
the cognitive processes involved in evaluating delayed
rewards (or losses) may differ from those involved with
uncertain rewards (or losses).

Discussion

Our study is the first to examine the relationships between
the number of cigarettes smoked per day, the estimated
amount of nicotine intake per day, and four types of dis-
counting (i.e., delay and probability discounting of mon-
etary gains and losses) within the same subjects. Our data
suggest five general conclusions: (1) the frequency of
nicotine self-administration is positively associated with
impulsive behavior in discounting delayed monetary re-

wards, (2) the level of chronic nicotine exposure is sim-
ilarly associated with impulsive behavior in discounting
delayed monetary gains, (3) relatively mild smokers do not
discount delayed or uncertain gains or losses more than
never smokers, (4) discounting monetary losses is not
strongly associated with smoking, and (5) the relationship
between smoking and probability discounting is not as
strong as that observed between smoking and delay dis-
counting of gains.

The correlations observed between smoking behavior

and delay discounting of monetary rewards are consistent
with previous studies. For example, Reynolds (

2004

) re-

cently reported a significant positive correlation between
the number of cigarettes smoked per day and a delay dis-
counting rate (logged hyperbolic k). In addition to the
positive relationship observed between discounting of de-
layed rewards and smoking frequency (Fig.

1

), we also

reported a positive association involving nicotine dosage
(Fig.

2

), whereby higher doses of nicotine were associated

with a greater tendency to discount delayed rewards. Neu-
ropsychopharmacologically, because chronic nicotine ex-
posure is known to associate with strong neuroadaptation,
predominantly in reward-processing brain regions (Liu
and Jin

2004

; Rahman et al.

2004

), it is possible that

chronic exposure to nicotine may reduce dopaminergic
activity in the neural circuitry, resulting in augmented de-
valuation of delayed monetary gains. However, whether
drug-intake-induced neuroadaptation actually causes strong
delay discounting of gain should be more extensively stud-
ied. Moreover, it was observed that habitual smokers who
consume more cigarettes regularly tend to smoke stronger
cigarettes. Therefore, smoking frequency (i.e., the num-
ber of cigarettes consumed per day) can alternatively be
adopted to estimate nicotine exposure when more biologi-
cally significant measures are unavailable (e.g., the amount
of daily nicotine intake or cotinine level).

Although smokers and never smokers did not differ in

their discounting behavior overall, this likely resulted from
the number of relatively mild nicotine users in our sample.
Most studies have focused on heavy smokers who consume
no less than 20 cigarettes per day (e.g., Bickel et al.

1999

;

Reynolds et al.

2004

); in contrast, only eight of the nicotine

users in our study met this criterion. As such, our inves-
tigation is the first to demonstrate that relatively mild
smokers do not more rapidly discount delayed monetary
rewards than never smokers. One possible interpretation
of this result is that the level of chronic nicotine exposure
in mild smokers might not be strong enough to affect
impulsivity.

Although there was a relationship with delay discount-

ing of gains, we did not detect a relationship between
smoking behavior and delay discounting of monetary
losses. This observation may reflect the unique neural ac-
tivation patterns observed in response to gains and losses
during decision making; dopaminergic neural responses
are evoked in response to monetary gains, whereas other
brain regions (e.g., the right anterior cingulate, thalamus,
and left amygdala) are more strongly activated in response
to monetary losses (Knutson et al.

2000

,

2003

; Breiter et al.

513

2001

). It should be noted that several studies have shown

that heavy smokers tend to discount delayed losses more
steeply than never smokers (Odum et al.

2002

; Baker et al.

2003

). Again, this discrepancy might be explained by the

relatively mild nicotine use exhibited by our sample of
smokers. This point should be further investigated in future
studies to draw more definitive conclusions.

Whereas delay discounting of gain was associated with

smoking behavior, we did not observe a significant correla-
tion between probability discounting of gain and smoking
behavior. Likewise, we did not find a significant differ-
ence in probability discounting of gain between smokers
and never smokers. The smokers employed in both our
present study and Mitchell

’s (

1999

) study were relatively

light smokers, which may have resulted in a nonsignif-
icant difference in probability discounting of gain between
smokers and never smokers. On the other hand, the study
of Reynolds and his colleagues reported that heavy smok-
ers discounted an uncertain monetary reward more drama-
tically when compared to never smokers, possibly because
they employed heavier smokers (Reynolds et al.

2004

).

Considering these results, it is possible that probability
discounting is related to smoking only in heavy smokers.

We further examined the relationships between the four

types of discounting regardless of smoking status. It was
revealed that the participants discounted both delayed and
uncertain gains more steeply than delayed and uncertain
losses, respectively, which is in line with previous reports
of an asymmetry in the decisions made regarding gains
and losses in discounting tasks (Tversky and Kahneman

1981

; Thaler

1981

; Baker et al.

2003

). It was also dem-

onstrated that the association between the tendency to
discount rewards and the tendency to discount losses was
positive in direction for delay discounting, but negative
for probability discounting. The latter finding can explain
paradoxical behavior observed in antisocial psychiatric
patients with comorbid drug dependence (Kausch

2003

)

who exhibit both low discounting of probabilistic rewards
(e.g., a preference for gambling) and high discounting of
probabilistic losses (e.g., low aversion to possible HIV
infection caused by needle sharing or high-risk sexual
behavior). Finally, although the direction of the correla-
tion observed between delay and probability discounting
of gains was positive, the coefficient did not reach sta-
tistical significance. This is consistent with the conclusion
of Myerson et al. (

2003

) that the tendencies to discount

delayed and uncertain gains is only weakly to modestly
related at best. However, because several studies have
shown a strong positive correlation between them (e.g.,
Richards et al.

1999b

; Reynolds et al.

2003

), further studies

are required to elucidate this relationship.

Although promising, there are limitations to our present

study. First, we did not restrict the participant

’s access to

nicotine prior to the experiment. In our opinion, it is im-
probable that the time of the last cigarette prior to partic-
ipating in the experiment significantly affected our results,
since (1) the time to complete our study was typically less
than 1 h, and (2) one previous study (Mitchell

2004

) has

shown that even a 24-h nicotine deprivation did not change

the discounting behavior of monetary outcomes (although
discounting cigarettes was significantly increased). Never-
theless, future studies should examine how the time of last
cigarette affects the subject

’s discounting behavior. Sec-

ond, we did not assess breath CO levels or urine cotinine
levels. It is, however, noteworthy that a number of studies
(Ueda et al.

2002

; Benowitz et al.

2003

; Binnie et al.

2004

)

have shown that there is a significant correlation between
self-reported smoking status and urinary cotinine levels,
especially in mild smokers, suggesting that self-reported
smoking status is a good estimate of actual nicotine in-
take. It should also be noted that our results are consistent
with the study of Reynolds et al. (

2004

) reporting that CO

levels were positively associated with delay discounting
of gains, but not with probability discounting of gains.
Nevertheless, it would be preferable for future studies to
assess biological markers of nicotine exposure such as
plasma cotinine levels and CSF (cerebrospinal fluid) nico-
tine levels.

Finally, we suggest future directions: (1) the effects of

acute and chronic nicotine administration on discounting
should be compared since psychopharmacological studies
have revealed that an acute administration of a dopami-
nergic drug may actually reduce impulsive behavior in
delay discounting of gains, whereas chronic exposure may
induce neuroadaptation and thereby increase impulsive
behavior in delay discounting of gains (Richards et al.

1999a

; Cardinal et al.

2000

; Wade et al.

2000

; de Wit et al.

2002

; Pietras et al.

2003

), and (2) future investigations

should combine a genetic analysis with a psychophar-
macological methodology to further elucidate the neuro-
psychopharmacological correlates of delay and probability
discounting of gains and losses because considerable evi-
dence indicates that nicotine use is influenced by our ge-
notype, such as DRD2 polymorphism (see Munafo et al.

2004

for a review).

Acknowledgements

The research reported in this paper was

supported by grants from the Grant-in-Aid for Scientific Research
(

“21st century center of excellence” grant and grant#17650074) from

the Ministry of Education, Culture, Sports, Science and Technology
of Japan, and a Yamaguchi endocrinological disorder grant. We are
grateful to Dr. Paul Wehr and anonymous reviewers for critical
reading of our manuscript.

References

Baker F, Johnson MW, Bickel WK (2003) Delay discounting in

current and never-before cigarette smokers: similarities and
differences across commodity sign and magnitude. J Abnorm
Psychol 112:382

–392

Benowitz NL, Pomerleau OF, Pomerleau CS, Jacob P III (2003)

Nicotine metabolite ratio as a predictor of cigarette consump-
tion. Nicotine Tob Res 5:621

–624

Bickel WK, Marsch LA (2001) Toward a behavioral economic

understanding of drug dependence: delay discounting pro-
cesses. Addiction 96:73

–86

Bickel WK, Odum AL, Madden GJ (1999) Impulsivity and cigarette

smoking: delay discounting in current never and ex-smokers.
Psychopharmacology 146:447

–454

514

Binnie V, McHugh S, Macpherson L, Borland B, Moir K, Malik K

(2004) The validation of self-reported smoking status by an-
alysing cotinine levels in stimulated and unstimulated saliva,
serum and urine. Oral Dis 10:287

–293

Breiter HC, Aharon I, Kahneman D, Dale A, Shizgal P (2001)

Functional imaging of neural responses to expectancy and
experience of monetary gains and losses. Neuron 30:619

–639

Cardinal RN, Robbins TW, Everitt BJ (2000) The effects of

D

-amphetamine chlordiazepoxide alpha-flupenthixol and be-

havioural manipulations on choice of signalled and unsignalled
delayed reinforcement in rats. Psychopharmacology 152:362

–

375

Daruna JH, Barnes PA (1993) A neurodevelopmental view of

impulsivity. In: McCown WG, Johnson JL, Shure MB (eds)
The impulsive client: theory research and treatment. American
Psychological Association, Washington, DC, pp 23

–37

de Wit H, Enggasser JL, Richards JB (2002) Acute administration

of

D

-amphetamine decreases impulsivity in healthy volunteers.

Neuropsychopharmacology 27:813

–825

Glimcher PW, Rustichini A (2004) Neuroeconomics: the consilience

of brain and decision. Science 306:447

–452

Johnson MW, Bickel WK (2002) Within-subject comparison of real

and hypothetical money rewards in delay discounting. J Exp
Anal Behav 77:129

–146

Kausch O (2003) Patterns of substance abuse among treatment-

seeking pathological gamblers. J Subst Abuse Treat 25:263

–

270

Kirby KN, Petry NM (2004) Heroin and cocaine abusers have

higher discount rates for delayed rewards than alcoholics or
non-drug-using controls. Addiction 99:461

–471

Kirby KN, Petry NM, Bickel WK (1999) Heroin addicts have higher

discount rates for delayed rewards than non-drug-using con-
trols. J Exp Psychol Gen 128:78

–87

Knutson B, Westdorp A, Kaiser E, Hommer D (2000) FMRI

visualization of brain activity during a monetary incentive delay
task. NeuroImage 12:20

–27

Knutson B, Fong GW, Bennett SM, Adams CM, Hommer D (2003)

A region of mesial prefrontal cortex tracks monetarily re-
warding outcomes: characterization with rapid event-related
fMRI. NeuroImage 18:263

–272

Liu ZH, Jin WQ (2004) Decrease of ventral tegmental area do-

pamine neuronal activity in nicotine withdrawal rats. Neuro-
report 15:1479

–1481

Martin-Solch C, Magyar S, Kunig G, Missimer J, Schultz W,

Leenders KL (2001) Changes in brain activation associated
with reward processing in smokers and nonsmokers. A positron
emission tomography study. Exp Brain Res 139:278

–286

Martin-Solch C, Missimer J, Leenders KL, Schultz W (2003) Neural

activity related to the processing of increasing monetary reward
in smokers and nonsmokers. Eur J Neurosci 18:680

–688

Mazur JE (1987) An adjusting procedure for studying delayed

reinforcement. In: Commons ML, Mazur JE, Nevin JA, Rachlin
H (eds) Quantitative analysis of behavior, vol 5. The effects
of delay and of intervening events on reinforcement value.
Erlbaum, Hillsdale, NJ, pp 55

–73

McClure SM, Laibson DI, Loewenstein G, Cohen JD (2004)

Separate neural systems value immediate and delayed monetary
rewards. Science 306:503

–507

Mitchell SH (1999) Measures of impulsivity in cigarette smokers

and non-smokers. Psychopharmacology 146:455

–464

Mitchell SH (2004) Effects of short-term nicotine deprivation on

decision-making: delay, uncertainty and effort discounting. Nic-
otine Tob Res 6:819

–828

Montague PR, Berns GS (2002) Neural economics and the bio-

logical substrates of valuation. Neuron 36:265

–284

Munafo M, Clark T, Johnstone E, Murphy M, Walton R (2004) The

genetic basis for smoking behavior: a systematic review and
meta-analysis. Nicotine Tob Res 6:583

–597

Myerson J, Green L, Warusawitharana M (2001) Area under the

curve as a measure of discounting. J Exp Anal Behav 76:235

–

243

Myerson J, Green L, Hanson JS, Holt DD, Estle SJ (2003)

Discounting delayed and probabilistic rewards: processes and
traits. J Econ Psychol 24:619

–635

Odum AL, Madden GJ, Bickel WK (2002) Discounting of delayed

health gains and losses by current never- and ex-smokers of
cigarettes. Nicotine Tob Res 4:295

–303

Perry JL, Larson EB, German JP, Madden GJ, Carroll ME (2005)

Impulsivity (delay discounting) as a predictor of acquisition of
IV cocaine self-administration in female rats. Psychopharma-
cology 178:193

–201

Petry NM (2001) Delay discounting of money and alcohol in ac-

tively using alcoholics currently abstinent alcoholics and con-
trols. Psychopharmacology 154:243

–250

Pietras CJ, Cherek DR, Lane SD, Tcheremissine OV, Steinberg JL

(2003) Effects of methylphenidate on impulsive choice in adult
humans. Psychopharmacology 170:390

–398

Rachlin H, Raineri A, Cross D (1991) Subjective probability and

delay. J Exp Anal Behav 55:233

–244

Rahman S, Zhang J, Engleman EA, Corrigall WA (2004) Neu-

roadaptive changes in the mesoaccumbens dopamine system
after chronic self-administration: a microdialysis study. Neuro-
science 129:415

–424

Reynolds B (2004) Do high rates of cigarette consumption increase

delay discounting? A cross-sectional comparison of adolescent
smokers and young-adult smokers and nonsmokers. Behav
Processes 67:545

–549

Reynolds B, Karraker K, Horn K, Richards JB (2003) Delay and

probability discounting as related to different stages of adoles-
cent smoking and non-smoking. Behav Processes 64:333

–344

Reynolds B, Richards JB, Horn K, Karraker K (2004) Delay

discounting and probability discounting as related to cigarette
smoking status in adults. Behav Processes 65:35

–42

Richards JB, Sabol KE, de Wit H (1999a) Effects of methamphet-

amine on the adjusting amount procedure, a model of impulsive
behavior in rats. Psychopharmacology 146:432

–439

Richards JB, Zhang L, Mitchell SH, de Wit H (1999b) Delay or

probability discounting in a model of impulsive behavior: effect
of alcohol. J Exp Anal Behav 71:121

–143

Schultz W (2004) Neural coding of basic reward terms of animal

learning theory game theory microeconomics and behavioural
ecology. Curr Opin Neurobiol 14:139

–147

Simpson CA, Vuchinich RE (2000) Reliability of a measure of

temporal discounting. Psychol Rec 50:3

–16

Takahashi T (2004) Cortisol levels and time-discounting of mon-

etary gain in humans. Neuroreport 15:2145

–2147

Thaler R (1981) Some empirical evidence on dynamic inconsis-

tency. Econ Lett 8:201

–207

Tversky A, Kahneman D (1981) The framing of decisions and the

psychology of choice. Science 211:453

–458

Ueda K, Kawachi I, Nakamura M, Nogami H, Shirokawa N, Masui

S, Okayama A, Oshima A (2002) Cigarettes nicotine yields and
nicotine intake among Japanese male workers. Tob Control
11:55

–60

Vuchinich RE, Simpson CA (1998) Hyperbolic temporal discount-

ing in social drinkers and problem drinkers. Exp Clin Psycho-
pharmacol 6:292

–305

Wade TR, de Wit H, Richards JB (2000) Effects of dopaminergic

drugs on delayed reward as a measure of impulsive behavior in
rats. Psychopharmacology 150:90

–101

515