The association between risky decision making and cocaine conditioned place preference is moderated by sex

Justin R Yates; Matthew J Horchar; Joy L Kappesser; Maria R Broderick; Alexis L Ellis; Makayla R Wright

doi:10.1016/j.drugalcdep.2021.109079

. Author manuscript; available in PMC: 2022 Nov 1.

Published in final edited form as: Drug Alcohol Depend. 2021 Sep 24;228:109079. doi: 10.1016/j.drugalcdep.2021.109079

The association between risky decision making and cocaine conditioned place preference is moderated by sex

Justin R Yates ¹, Matthew J Horchar ¹, Joy L Kappesser ², Maria R Broderick ², Alexis L Ellis ¹, Makayla R Wright ²

PMCID: PMC8595855 NIHMSID: NIHMS1743059 PMID: 34600260

Abstract

Background.

Excessive risk taking is a characteristic trait of several psychiatric conditions, including substance use disorders. High risk-taking (HiR) rats self-administer more cocaine compared to low risk-taking (LoR) rats. However, research has not determined if risk taking is associated with enhanced cocaine conditioned place preference (CPP).

Methods.

Male and female Sprague Dawley rats (n = 48 each sex) were first tested in the risky decision task (RDT), in which a response on one lever resulted in safe delivery of one food pellet, and a response on a different lever resulted in delivery of two pellets and probabilistic delivery of foot shock. Following RDT training, rats were tested for cocaine CPP. The first session was a pretest that measured rats’ preference for three compartments that provided different visual and tactile cues. Rats then learned to associate one compartment with cocaine (either 10.0 mg/kg or 20.0 mg/kg; i.p.) and one compartment with saline (1.0 ml/kg; i.p.) across eight conditioning sessions. Finally, rats explored all three compartments in a drug-free state.

Results.

Sex significantly moderated the association between risky decision making and cocaine CPP. While increased risk aversion was somewhat positively associated with cocaine CPP in males, increased risk taking was positively correlated with cocaine CPP in females.

Conclusions.

These results highlight the moderating role of sex on the relationship between risky decision making and cocaine reward.

Keywords: Risk taking, Risky decision task, Conditioned place preference, Conditioned rewarding effects, Cocaine, Individual differences, Sex differences, Rat

1. Introduction

Risk taking is characteristic of several psychiatric conditions, including gambling disorder (see Brevers et al., 2013), internet gaming disorder (Dong and Potenza, 2016), bipolar disorder (Adida et al., 2008; van Enkhuizen et al., 2014), and substance use disorders (SUDs) (Brand et al., 2008; Brevers et al., 2014; Leland and Paulus, 2005; Schutter et al., 2011). Substance misuse can be viewed as a form of risk taking, as individuals prefer a reward alternative (drug) that can lead to negative outcomes, such as incarceration, impaired interpersonal relationships, spread of infectious diseases, loss of job, and death (American Psychiatric Association, 2013). Although there is an association between risk-taking behavior and substance use disorders (SUDs), one important question is whether risk taking is a predisposing factor for the development of SUDs or a result of prolonged drug exposure. At the preclinical level, there is some evidence that this relationship is bidirectional. Ethanol increases risk-taking behaviors in rodents (Boutros et al., 2014; Spoelder et al., 2015), and high risk-taking (HiR) rats show increased self-administration (Mitchell et al., 2014), escalation (Orsini et al., 2020), and incubation of craving (Ferland and Winstanley, 2017) of the psychostimulant cocaine compared to low risk-taking (LoR) rats. These studies have increased our understanding of the association between risk taking and substance use; however, more work is needed to further increase this understanding.

Thus far, studies have focused on the association between risk taking and drug self-administration. However, no study has determined if increased risk-taking behavior is linked to conditioned place preference (CPP). Whereas drug self-administration measures the direct reinforcing efficacy of drugs of abuse, CPP measures the conditioned rewarding effects of a drug (see Bardo and Bevins, 2000). In some instances, there is high concordance between self-administration and CPP. For example, high impulsive animals show increased self-administration of psychostimulants (Anker et al., 2009; Belin et al., 2008; Dalley et al., 2007; Marusich and Bardo, 2009) and show enhanced amphetamine CPP (Yates et al., 2012). In other cases, increased drug self-administration does not translate to enhanced CPP. Animals raised in isolation show greater self-administration (Bardo et al., 2001; Green et al., 2002) but decreased CPP (Bardo et al., 1995; Bowling and Bardo, 1994) of psychostimulants relative to rats raised in an enriched environment (access to social cohorts and novel toys). Because risk taking and impulsivity are distinct constructs (Herman et al., 2018; Shimp et al., 2015; Simon et al., 2009; but see Gabriel et al., 2019), there is no guarantee that HiR rats will show enhanced CPP as observed in high impulsive animals. Thus, determining if increased risk taking predicts cocaine CPP will increase our understanding of the association between risky decision making and SUDs.

In the current experiment, rats were first trained in the risky decision task (RDT) (Simon et al., 2009). In the RDT, rats are allowed to choose between two alternatives: one paired with a small magnitude reinforcer and one paired with a larger magnitude reinforcer and probabilistic delivery of foot shock. We chose this task because rats classified as HiR in the RDT self-administer more cocaine relative to LoR rats (Mitchell et al., 2014; Orsini et al., 2020). The goal of this experiment was to determine if HiR rats show greater cocaine CPP. Because sex differences have been observed in the risky decision task (Orsini et al., 2016; Yates et al., 2021) and in cocaine CPP (Russo et al., 2003; Zakharova et al., 2009; but see Hilderbrand and Lasek, 2014; Morris Bobzean et al., 2010), we tested both males and females in the current experiment. In total, we tested four groups of rats as we tested two doses of cocaine in CPP (10.0 and 20.0 mg/kg) in males and in females. We hypothesized that risk taking as assessed in the RDT would be positively correlated with cocaine CPP and that this correlation would only be observed at the lower tested dose of cocaine as individual differences tend to dissipate when higher doses are tested in CPP (e.g., Klebaur & Bardo, 1999; Yates et al., 2012). We also hypothesized that females would show greater CPP compared to males as females tend to show enhanced cocaine CPP relative to males (see citations above).

2. Materials and Methods

2.1. Animals

A total of 96 adult Sprague Dawley rats (48 males weighing 200–224 g upon arrival; 48 females weighing 150–174 g upon arrival) were ordered from Envigo (Indianapolis, IN). Rats were individually housed throughout the experiment and were allowed to acclimate to the housing room for 7 days before behavioral testing occurred. The housing room was maintained on a 12:12-h reverse light/dark cycle (lights off at 700 h), and rats were tested in the dark phase (approximately 1000–1700 h). Rats were housed in clear plastic cages (30 cm wide × 24.6 cm high × 41.2 cm long) with a wire top lid to hold food and a plastic external bottle top to hold two water bottles. Each cage contained one plastic nylon bone (Bio-Serv, Flemington, NJ; product K3581). During the RDT, rats were maintained at approximately 85% of their free feed weight. During CPP testing, rats had ad libitum access to food. All experimental procedures were carried out according to the Current Guide for the Care and Use of Laboratory Animals (National Research Council, 2011) under a protocol approved by the Northern Kentucky University Institutional Animal Care and Use Committee (Protocol #: 2019–05).

2.2. Apparatus

Sixteen operant conditioning chambers (28 × 21 × 21 cm; MED Associates, St. Albans, VT; product ENV-008) located inside sound attenuating chambers (MED Associates; product ENV-018M) were used for the RDT. The front and back walls of the chambers were made of aluminum, while the side walls were made of Plexiglas. A recessed food tray (5 × 4.2 cm) was located 2 cm above the floor in the bottom-center of the front wall between two retractable levers. Each lever (4.8 × 0.55 × 1.9 cm) was located 2.1 cm above the floor and required a force of 0.245 N to depress. An infrared photobeam was used to record head entries into the food tray. A 28-V white stimulus light (2.54 cm diameter) was located 6 cm above each response lever. A 28-V white house light was mounted in the center of the back wall of the chamber. The floor of the operant chamber was composed of steel rods connected to a shock generator (MED Associates; product ENV-414S) that delivered foot shocks. All responses and scheduled consequences were recorded and controlled by a computer interface. A computer controlled each experimental session using Med-V software.

Eight 3-compartment chambers (68 × 21 × 21 cm; MED Associates; product ENV-013) located inside sound-attenuating chambers (MED Associates; product ENV-020M) were used to measure locomotor activity and CPP. The three compartments were separated by sliding guillotine doors. The middle compartment (12 × 21 × 21 cm) had gray walls with a smooth gray PVC floor. The end compartments (28 × 21 × 21 cm) provided different contexts, with one compartment having black walls with a stainless-steel grid rod floor and the other end compartment having white walls with a stainless-steel mesh floor. Recessed trays were located 2 cm below each compartment. A computer controlled the experimental session using Med-V software. A series of infrared photobeams (6 beams in the black and white compartments and 3 beams in the gray compartment) were used to detect the rats’ presence in a particular compartment and record the amount of time spent in that compartment, as well as to record locomotor activity during conditioning sessions.

2.3. Procedure

2.3.1. RDT.

Rats received 1–2 sessions of magazine training. During these sessions, rats received 45-mg food pellets (Bio-Serv; product F0021) non-contingently according to a variable-time (VT) 30 s schedule of reinforcement. The session ended after rats obtained 20 pellets, which took 10 min. Each rat ate all 20 pellets by the second day of testing. Rats then received 1–3 sessions of lever-press training. Each session began with illumination of the house light. A head entry into the food tray resulted in presentation of one lever; each lever was presented pseudo-randomly, with no more than two consecutive presentations of the same lever. A single response (fixed ratio [FR] 1) on the extended lever resulted in the following events: extinguishment of the house light, retraction of the lever, and delivery of one food pellet. After a 5-s intertrial interval (ITI), the house light was illuminated, signaling the start of the next trial. Each lever-press training session ended after a rat earned 40 reinforcers or after 30 min, whichever came first. Each subject earned all 40 reinforcers by the end of lever-press training. Most rats (91 out of 96) earned all 40 reinforcers during the first session.

Rats were then tested in magnitude discrimination training for 2–4 days. Similar to lever-press training, each session consisted of 40 trials, and the beginning of each trial was signaled by illumination of the house light. A head entry into the food tray extended one of the levers, and the order of presentation between the two levers was pseudo-randomized, with no more than two consecutive presentations of the same lever. Completing the response requirement (FR 1) resulted in the following events: extinguishment of the house light, retraction of the lever, and delivery of reinforcement. One lever was associated with immediate delivery of one pellet, whereas the other lever was associated with immediate delivery of two pellets. Note, male rats in the 10.0 mg/kg cocaine CPP experiment initially received four pellets as the large magnitude reinforcer (see below for more details). The lever associated with the large magnitude reinforcer was counterbalanced across rats. If an animal failed to respond within 20 s, the lever was retracted, the light was extinguished, and the trial was scored as an omission. Each trial lasted 30 s. Because each trial lasted 30 s, the ITI was variable. For example, if an animal completed the response requirement within 5 s, there was a 25-s ITI; if an animal completed the trial within 15 s, there was a 15-s ITI.

Following magnitude discrimination training, rats were trained in a version of the RDT modified from Setlow and colleagues (e.g., Simon et al., 2009; Orsini et al., 2016). Each RDT session was composed of five blocks of 14 trials. The first six trials of each block were forced-choice trials, in which only one lever was made available. The final eight trials of a block were free-choice trials, in which both levers were made available. For half of the rats, a response on the left lever was associated with delivery of one food pellet, and a response on the right lever was associated with delivery of two food pellets. These contingencies were reversed for the other half of the rats. Each trial began with illumination of the house light. A head entry into the food tray resulted in the presentation of one lever (if forced-choice trial) or both levers (if free-choice trial). During forced-choice trials, each lever was presented pseudo-randomly, with no more than two consecutive presentations of the same lever. Completing the response requirement (FR 1) on the extended lever resulted in the following events: extinguishment of the house light, retraction of the lever(s), and delivery of reinforcement. Additionally, responses on the lever associated with the large magnitude reinforcer resulted in delivery of probabilistic foot shock that lasted for 1 s (see below for more details). Like magnitude discrimination training sessions, a 20-s limited hold was in effect. Failure to respond within 20 s resulted in the following events: extinguishment of the house light and retraction of the lever(s). Each trial ended after 30 s, regardless of trial outcome; thus, ITIs were variable as described above. Sessions ended after 70 trials.

During the first 5–9 sessions of the RDT, no shock was delivered. These sessions were included to ensure animals could complete all 70 trials with minimal omissions. When we initially conducted this experiment with male rats (10.0 mg/kg CPP group), we had to reduce the magnitude of the large reinforcer from four pellets to two pellets after the third session because omissions were high (16.208 ± 0.855 free-choice omissions; 11.583 ± 0.607 forced-choice omissions). The number of omissions significantly decreased when we adjusted the large magnitude reinforcer, even after the first three sessions after making the switch (10.444 ± 0.885 free-choice omissions, p < .001; 9.264 ± 0.637 forced-choice omissions, p < .001). Previous research has shown that reducing the number of pellets for the large magnitude reinforcer does not substantially affect preference for this alternative (Shimp et al., 2015). Following these initial RDT sessions, the probability of receiving a mild foot shock increased across the session (0, 25, 50, 75, 100%) for all rats. On each individual trial within a block of trials, the probability of receiving foot shock was the same (e.g., during the 25% block, the probability of receiving shock was always set to 0.25). Initially, each rat started with a shock intensity of 0.10 mA. The shock intensity was adjusted across sessions (by 20%) until most rats (1) started to discount the large, risky option and (2) responded more for the large magnitude reinforcer when the probability of receiving foot shock was 0 (i.e., the first block of trials). We also wanted to ensure that rats showed variability in their responses for the large magnitude reinforcer (i.e., we did not want all of the rats to show similar preference for this alternative compared to other rats). For some of the experiments, we had to eventually reduce the shock intensity (from between 5–20%) because a large proportion of rats began to stop responding for the large magnitude reinforcer, even when its delivery was not paired with shock. Table 1 shows the shock intensities used for each group and the number of sessions trained at each intensity. Importantly, the shock intensity was adjusted for all rats at the same rate within an individual experiment. That is, if we had to increase the shock intensity by 20%, we increased the shock intensity for all rats in an individual experiment.

Table 1,

Shock intensity (in mA) sequence for each group of rats and the number of sessions spent at each shock intensity.

Sex	Group	Shock Intensity (Number of Sessions)
Male	10.0 mg/kg cocaine CPP	0.10 (1), 0.12 (1), 0.14 (1), 0.17 (1), 0.20 (1), 0.24 (1), 0.29 (19)
	20.0 mg/kg cocaine CPP	0.10 (1), 0.12 (1), 0.14 (1), 0.17 (1), 0.20 (1), 0.24 (2), 0.29 (1), 0.35 (11), 0.32 (1), 0.30 (13)
Female	10.0 mg/kg cocaine CPP	0.10 (1), 0.12 (1), 0.14 (1), 0.17 (1), 0.20 (1), 0.24 (1), 0.20 (5), 0.22 (3), 0.24 (4), 0.27 (4), 0.30 (4), 0.32 (7)
	20.0 mg/kg cocaine CPP	0.10 (1), 0.12 (1), 0.14 (1), 0.17 (1), 0.20 (1), 0.24 (1), 0.20 (5), 0.22 (3), 0.24 (7)

Open in a new tab

At the end of baseline training, rats had to meet the following criteria to be included in the CPP experiment: (1) show greater responding for the large magnitude reinforcer when the odds against receiving foot shock was 0 and (2) show sensitivity to foot shock as a function of odds against. That is, rats had to show a general decrease in their responding for the large magnitude reinforcer across trial blocks (greater than a 20% decrease from the first block of trials to the last block of trials). If a rat showed near exclusive preference for the large magnitude reinforcer across each block of trials or increased their responding for this alternative across two consecutive blocks of trials (i.e., nonsystematic discounting; see Johnson and Bickel, 2008), it was excluded from the experiment. Across all 96 rats, 18 rats failed to meet the inclusion criteria (4–5 rats per experimental condition), leaving sample sizes of 19 for each group of males and 20 for each group of females.

2.3.2. CPP.

Upon completion of the RDT, rats were given ad libitum access to food and began the 10-day CPP experiment 2–3 days after the last RDT session. On the first day of each CPP experiment, rats were given a pretest. Rats were able to travel freely in all three CPP compartments for 15 min. Data were collected by determining how many photobeams were broken in each chamber during the session. The chamber that the rat spent the least amount of time in during the pre-test was the one that was paired with cocaine during the conditioning phase (i.e., biased design). Following the pretest, rats went through 8 days of conditioning, in which rats were confined by the guillotine door to either the black or the white compartment for 30 min. During conditioning sessions, rats received an injection of cocaine (either 10.0 or 20.0 mg/kg; i.p.) and then were immediately placed in the initially non-preferred chamber every other day. On alternating days, rats received an injection of saline (1.0 ml/kg; i.p.) and then were immediately placed in the initially preferred chamber. During the posttest, the guillotine doors were opened, and rats were allowed to explore all three compartments for 15 min in a drug-free state. The time spent in each compartment was recorded. To determine if cocaine was behaviorally active, activity was also recorded during each conditioning session by measuring the total number of non-repeating photobeam breaks (a proxy of horizontal activity) and repeating photobeam breaks (a proxy of stereotypies). Non-repeating photobeam breaks were also recorded during the posttest to determine if horizontal activity influenced the magnitude of CPP.

2.3.3. Amphetamine Drug Discrimination.

Because risky decision making was negatively correlated with cocaine (10.0 mg/kg) CPP in male rats, we wanted to determine if this finding could be explained by differential sensitivity to the interoceptive cues of psychostimulants. As such, we tested this group of rats in a drug discrimination paradigm using amphetamine. Rats were given four lever-press training sessions as described for the RDT experiment. However, the response requirement increased from an FR 1 to an FR 10 over the course of these sessions (FR 1, 3, 5, 10). Rats were then trained in the drug discrimination task. Using a double alternating schedule as described previously (Bevins et al. 1997; Meyer et al. 2011), rats received subcutaneous injections of amphetamine or saline 15 min before each 10-min experimental session. Half of the rats received amphetamine injections first (DDSSDDSS), and half received saline first (SSDDSSDD). The lever associated with amphetamine was counterbalanced to account for the lever associated with the large, risky option during the RDT (i.e., for half of the rats, the amphetamine lever was the same as the lever associated with the large, risky option). Each session began with illumination of the house light and extension of both levers. Only responses on the injection-appropriate lever were reinforced according to an FR 10 schedule. Responses on the incorrect lever were recorded but had no programmed consequences. There was no time-out period or intertrial interval (ITI) following delivery of each reinforcer. Initially, we used a dose of amphetamine (1.0 mg/kg; s.c.) that has been used in previous drug discrimination experiments (Bevins et al. 1997; Oberlender and Nichols 1988). However, none of the rats emitted more than 11 responses following administration of this dose of amphetamine. After three sessions, the dose of amphetamine was decreased to 0.5 mg/kg. We had to further decrease the dose to 0.3 mg/kg following an additional three sessions as 14 out of 19 rats failed to emit more than 50 responses following amphetamine administration. For 34 sessions, rats received the 0.3 mg/kg dose. Most of the rats (n = 16) responded following amphetamine injections. We then increased the dose back to 0.5 mg/kg and tested the rats for an additional 18 sessions. We measured the number of sessions required to meet the following discrimination criteria for the first time on 7 out of 8 sessions: 1) ≥ 80% responding on the injection appropriate lever before earning the first reinforcer and 2) ≥ 85% responding on the injection appropriate lever during the entire session (Meyer et al. 2011). We also measured the total number of sessions in which rats met the discrimination criteria.

2.4. Drugs

(−)-Cocaine hydrochloride and d-amphetamine sulfate were obtained from Sigma Aldrich (St. Louis, MO) and were mixed in 0.9% NaCl (saline). All injections were delivered at a volume of 1.0 ml/kg via the intraperitoneal (i.p.) route (cocaine) or the subcutaneous (s.c.) route (amphetamine). Doses were calculated based on salt weight.

2.5. Statistical Analyses

Most statistical analyses were conducted using Jamovi (version 1.6.18.0), although R (version 3.6.1) was used for nonlinear mixed effects modeling (see below for more specifics). For all analyses, statistical significance was defined as p < .05, except when multiple t tests/Mann Whitney U tests were conducted to probe a significant interaction (see Results section for more details). For ANOVA analyses in which the within-subjects factor had three or more levels, Greenhouse-Geisser corrections were used in the case the assumption of sphericity was violated.

2.5.1. RDT Performance.

Somewhat similar to previous studies (Mitchell et al. 2014; Orsini et al., 2016; Simon et al., 2009), a two-way repeated measures ANOVA was used to determine if stable responding had been achieved during the final three sessions of RDT training across rats for each experimental group, with day and trial block as within-subjects factors. Separate two-way ANOVAs were conducted for each individual condition. To meet stability, the ANOVA had to yield a significant main effect of trial block but no main effect of day or a significant trial block × day interaction.

At the end of training, 11 rats showed near exclusive preference for the large, risky option across all blocks of trials. Because we could not determine if this exclusive preference reflected increased risky decision making or decreased sensitivity to the shock intensity, we excluded these rats from the CPP experiment. Similarly, two rats showed a bias toward the large, risky option across blocks of trials, albeit less pronounced than the other 11 rats that were excluded. However, these two rats displayed a lack of stimulus control. The proportion of responses for the large magnitude reinforcer did not systematically decrease across blocks of trials. Specifically, one male rat in the 20.0 mg/kg group had a somewhat U-shaped discounting function while one female in the 10.0 mg/kg group had a relatively flat discounting function until the final block of trials. An additional five rats showed low responding for the large magnitude reinforcer (less than 50%) during the first block of trials and were thus excluded from the study as their pattern of responding appeared to reflect a loss of stimulus control (i.e., rats should respond more for the large magnitude reinforcer when its delivery is not paired with shock). Overall, 18 rats were excluded from all statistical analyses (18.75% of all rats).

Baseline risky decision making was analyzed one of two ways. First, we used a mixed factor ANOVA, with trial block as a within-subjects factor and sex and cocaine dose as between-subjects factors. Second, hyperbolic discounting functions were fit to each individual rat’s raw proportion of responses for the large, risky option. The hyperbolic discounting function is often used to quantify delay/probability discounting but has not been applied to RDT data, although we previously applied an exponential discounting function (Yates et al., 2021). While both hyperbolic and exponential discounting functions provided excellent fits of the current data (R² values of .989 and .983, respectively), we used the hyperbolic discounting because Akaike’s information criterion (AIC) and Bayesian information criterion (BIC) were lower for the hyperbolic function (−220.990 and −177.362, respectively) compared to the exponential function (−175.186 and −131.558, respectively). The equation for the hyperbolic function is V = A/(1 + hθ), where V is the subjective value of the large magnitude reinforcer (i.e., corresponds to the proportion of responses for this alternative at each probability of receiving a foot shock), A measures how much an animal responds for the large magnitude reinforcer in the absence of shock, h is the rate of discounting (i.e., measure of risk taking), and θ is odds against delivery of safe reinforcement. The calculation for odds against is θ = (1/p) – 1, where p is the probability of receiving the large magnitude reinforcer without concurrent foot shock. Across each block of trials, the odds against safe delivery of the large magnitude reinforcer were 0, 0.33, 1, 3, and 10. Because the odds against cannot be calculated for the final block of trials ([1/0] – 1 is undefined), they were arbitrarily set to 10 (see Yates et al., 2021).

GraphPad Prism 9 was used to fit the hyperbolic discounting function to individual subject data. Due to convergence issues¹, a constant of 0.04 was added to each value. The value of 0.04 was chosen because it was less than the lowest non-zero value (.042). To compare baseline discounting across groups of rats, the hyperbolic discounting function was fit to the data via nonlinear mixed effects (NLME) modeling using the nlme package in R (Pinheiro et al., 2019). The NLME model included sex and cocaine dose as nominal, between-subjects factors, A and h as free parameters, and subject as a random effect. Specifically, A and h parameter estimates were allowed to vary across subjects.

Because h parameter estimates were not normally distributed, we used log-transformed h values. Additionally, we found that A parameter and log-transformed h parameter estimates were significantly correlated across all rats, r = −.252, p = .027; therefore, we normalized each rat’s raw proportion of responses for the large magnitude reinforcer, such that the proportion of responses for the 0 odds against block was 1. We then divided the proportion of responses for each subsequent block to the proportion of responses in the first block. See Figure 1 for more details. We used normalized, log-transformed h parameter estimates for all subsequent analyses.

Figure 1. — Illustration depicting how discounting functions were normalized. In this example, the raw proportion of responses for the large, risky option are depicted by the closed circles and solid line. Across each block of trials, the proportion of responses are 0.753, 0.253, 0.043, 0, and 0. To normalize the discounting curve (open circles and dashed line), the proportion of responses is set to 1 for the first block of trials. The proportion of responses for blocks 2–5 are divided by the raw proportion of responses during the first block of trials (0.753), resulting in the normalized values of 0.336, 0.058, 0, and 0.

A Spearman correlation was performed to determine if normalized, log-transformed h parameter estimates were correlated with mean choice for the large, risky option (proportion of responses for the large, risky option during trial blocks 2–5 - trial blocks in which shock was administered). Mean choice for the large, risky option has been used previously to correlate risky decision making with cocaine self-administration (Orsini et al., 2020). We used a Spearman correlation because visual inspection of the data showed that the relationship between these two variables was nonlinear.

Omissions were analyzed with Poisson generalized linear models, with trial block as a nominal, within-subjects factor and sex and cocaine dose as nominal, between-subjects factors. Separate models were used for free-choice trials and forced-choice trials.

Response latencies were analyzed with a mixed factor ANOVA, with trial type (free choice vs. forced-choice) and lever identity (large, risky vs. small, safe) as within-subjects factors and sex and cocaine dose as between-subjects factors. Because of a programming error, the free-choice response latencies for the small, safe alternative during the third block of trials are largely incorrect. This programming error also affected some of the response latencies for the large, risky option during the third block of trials and affect the first free-choice trial response latency (for either the large, risky or small, safe alternative) of the fourth block of trials. As such, we excluded the values affected by this error. We also performed a linear regression analysis to determine if response latencies (large, risky trials subtracted by small, safe trials) were associated with normalized, log-transformed h parameter estimates. This regression model included interactions between normalized, log-transformed h parameter estimates and sex and cocaine dose to determine if the relationship between response latencies and risky decision making were moderated by either of these variables. Separate regression analyses were performed for free-choice trials and forced-choice trials.

2.5.2. Locomotor Activity During CPP Conditioning Sessions.

Due to a technical issue, the data for three female rats in the 10.0 mg/kg CPP are missing for the first session (two saline sessions and one cocaine session). As such, non-repeating and repeating photobeam breaks were analyzed separately using linear mixed effects (LME) modeling. LME models account for partially missing data unlike ANOVA. Each LME model included treatment (cocaine vs. saline) and session as nominal, within-subjects factors, and sex and cocaine dose as nominal, between-subjects factors. Bonferroni post hoc tests were used to probe any significant interactions.

One multiple linear regression analysis was conducted to determine if normalized, log-transformed h parameter estimates, sex, and cocaine dose (as well as all possible interactions) predicted non-repeating photobeam breaks during conditioning sessions. Specifically, non-repeating photobeam breaks in the cocaine-paired compartment were subtracted by non-repeating photobeam breaks in the saline-paired compartment to create a single difference score. We included sex and cocaine dose in this analysis because the LME analyses described above showed that sex and cocaine dose interacted with treatment. Thus, we wanted to determine if sex and/or cocaine dose moderated the relationship between risky decision making and locomotor activity. A similar analysis was performed with repeating photobeam breaks as the criterion variable.

2.5.3. Cocaine CPP.

The time spent in each compartment during the pretest was analyzed with a mixed factorial ANOVA, with compartment as a within-subjects factor and sex and cocaine dose as between-subjects factors. This analysis was conducted to determine if the CPP apparatus used in the current study was biased or unbiased. Tukey’s post hoc test was used to examine the main effect of compartment.

To determine if rats in each condition developed cocaine CPP, the time spent in the cocaine-paired chamber during the pretest was compared to the time spent in this compartment during the posttest using a mixed factorial ANOVA. The ANOVA included test (pre vs. post) as a within-subjects factor and sex and cocaine dose as between-subjects factors. Tukey’s post hoc tests were used to probe any significant interactions.

Separate multiple linear regression analyses were conducted to determine if response latencies during free-choice trials or during forced-choice trials predicted cocaine CPP. These regression models also included sex, cocaine dose, and all possible interactions in order to determine if sex and/or cocaine dose moderated the relationship between response latencies and CPP.

Because locomotor activity during conditioning sessions has been shown to predict CPP (Mathews et al., 2010), we first averaged non-repeating photobeam breaks across each conditioning session and calculated a difference score (cocaine sessions - saline sessions). We used these difference scores as a predictor variable for CPP in a multiple linear regression analysis. The regression model also included sex, cocaine dose, and all possible interactions. A similar analysis was conducted for repeating photobeam breaks.

We performed an additional analysis in which we calculated non-repeating photobeam breaks/s in the cocaine-paired compartment and the saline-paired compartment during the posttest. We used photobeam breaks/s to account for the fact that rats do not spend an equal amount of time in each compartment during the posttest. As with CPP data, we calculated a difference score (photobeam breaks/s in cocaine compartment – photobeam breaks/s in saline compartment). We then performed a multiple linear regression analysis to determine if non-repeating photobeam breaks during the posttest predicted CPP difference scores. The regression model also included sex, cocaine dose, and all possible interactions to determine if either one of these variables moderated the relationship between non-repeating photobeam breaks/s and cocaine CPP.

2.5.4. Association between RDT Performance and Cocaine CPP.

We fit a multiple linear regression model using CPP difference scores (posttest - pretest scores) as the criterion variable, and sex, cocaine dose, and normalized, log-transformed h parameter estimates as the predictor variables. Each possible interaction was included in the regression model as well. Because we found a significant interaction between normalized, log-transformed h parameter estimates and sex, we performed separate Pearson’s correlations for each sex to probe this interaction.

To be able to better directly compare the results of the current experiment with previous studies assessing the association between risky decision making and cocaine self-administration (i.e., Orsini et al., 2020), we performed a similar analysis as above, except that we used mean choice for the large, risky option during blocks 2–5 in place of normalized, log-transformed h parameter estimates.

2.5.5. Amphetamine Drug Discrimination.

We performed linear regression to determine if normalized, log-transformed h parameter estimates predicted the number of individual sessions in which rats met stability criteria (≥ 80% responses on injection-appropriate lever before earning the first reinforcer and ≥ 85% responses on injection-appropriate lever during the entire session for 7 out of 8 sessions). Separate analyses were conducted for each dose of amphetamine.

The total number of responses were averaged across the final eight discrimination sessions (final four amphetamine injection days and final four sessions of saline injections). First, a repeated measures ANOVA was conducted to determine if responses were higher for amphetamine relative to saline and to determine if responses differed across each dose of amphetamine. Second, linear regressions were used to determine if normalized, log-transformed h parameter estimates predicted the number of responses (amphetamine - saline) during drug discrimination sessions.

3. Results

3.1. Baseline RDT Performance

At the end of baseline training, each group of rats met the criteria for stability: main effect of trial block, all F’s ≥ 47.143, all p’s < .001; no significant effect of day, all F’s ≤ 2.164, all p’s ≥ .142; no significant interaction, all F’s ≤ 1.819, all p’s ≥ .076.

Figure 2 shows the raw proportion of responses for the large, risky option averaged across the final three sessions of RDT training for all 96 rats (panels a–d). A total of 18 rats (18.75% of total sample) were excluded from further statistical analyses because they failed to meet the stability criteria. For the remaining 78 rats, we first performed a mixed factor ANOVA to compare baseline RDT performance across sex and across cocaine dose (Fig. 2e). There was a main effect of trial block, F(2.204, 160.867) = 541.726, p < .001, as well as significant interactions between sex and cocaine dose, F(1, 73) = 9.159, p = .003, and trial block, sex, and cocaine dose, F(2.204, 160.867) = 6.017, p = .002. To probe the significant three-way interaction, we first performed separate mixed factor ANOVAs for each sex, with trial block as a within-subjects factor and cocaine dose as a between-subjects factor. For males, there was a main effect of trial block only, F(1.886, 67.892) = 251.666, p < .001. However, for females, there were main effects of trial block, F(2.499, 92.470) = 294.427, p < .001, and cocaine dose, F(1, 37) = 5.945, p = .020, as well as a significant interaction between trial block and cocaine dose, F(2.499, 92.470) = 6.971, p < .001. Females assigned to receive the higher dose of cocaine (20.0 mg/kg) in CPP responded more for the large, risky alternative compared to females assigned to receive a lower dose of cocaine (10.0 mg/kg). To probe the interaction, separate independent-samples t tests or Mann-Whitney U tests (when homogeneity of variance was violated) were conducted to compare the proportion of responses for each trial block. To control for Type I error, we adjusted our p value to .01. Female rats responded more for the large, risky option during the 25% trial block, t(37) = 2.755, p = .009.

To examine potential sex differences in baseline discounting, we also probed the significant three-way interaction by conducting separate mixed factor ANOVAs for each cocaine dose, with trial block as a within-subjects factor and sex as a between-subjects factor. When examining the low dose of cocaine, there was a main effect of trial block only, F(1.776, 65.702) = 272.127, p < .001. However, there were clear sex differences for rats assigned to receive the higher dose of cocaine. Not only was there a main effect of trial block, F(2.576, 92.753) = 271.658, p < .001, but there was a main effect of sex, F(1, 36) = 5.958, p = .020, and a significant trial block × interaction, F(2.576, 92.753) = 5.326, p = .003. Females responded significantly more for the large, risky alternative across trial blocks compared to males. In short, females assigned to receive the high dose of cocaine displayed significantly higher levels of risky decision making compared to the other experimental groups. To probe the interaction, separate independent-samples t tests or Mann-Whitney U tests were conducted to compare the proportion of responses for each trial block. As before, we adjusted our p value to .01 to control for Type I error. Female rats responded more for the large, risky option during the 50% trial block, U(36) = 84.000, p = .004, and during the 100% trial block, U(36) = 104.000, p = .007.

Concerning baseline h parameter estimates, NLME modeling showed a significant sex × cocaine dose interaction only, F(1, 305) = 29.467, p < .0001 (Fig. 2f). For rats in the 10.0 mg/kg CPP experiment, females had higher h parameter estimates (i.e., greater risk aversion) compared to males. Conversely, male rats assigned to the 20.0 mg/kg CPP experiment had higher h parameter estimates compared to females. A parameter estimates did not differ across sex or cocaine dose, all F’s ≤ 0.335, all p’s ≥ .563 (Fig. 2g).

A Spearman correlation showed that mean choice and normalized, log-transformed h parameter estimates were significantly correlated, r_s(76) = .990, p < .001 (Fig. 2h).

Figure 3 shows omissions across each block of trials during free-choice trials and forced-choice trials averaged across the final three baseline training sessions. Results of a Poisson generalized linear model showed that omissions increased across blocks of trials, χ²(4) = 59.191, p < .001. Females had more omissions during free-choice trials compared to males, χ²(1) = 18.953. When examining forced-choice trials, there was a main effect of trial block, χ²(4) = 259.568, p < .001, and a significant three-way interaction between trial block, sex, and cocaine dose, χ²(4) = 9.586, p = .048. To probe this interaction, we first ran separate Poisson generalized linear models for each sex. For males, there was a main effect of trial block only, χ²(4) = 124.661, p < .001. For females, there were main effects of trial block, χ²(4) = 137.656, p < .001, and cocaine dose, χ²(1) = 6.005, p = .014. We also ran separate Poisson generalized linear models for each cocaine dose. In addition to a main effect of trial block, χ²(4) = 122.786, p < .001, female rats assigned to the 10.0 mg/kg cocaine CPP experiment had more omissions compared to males, χ²(1) = 9.681, p = .002.

Concerning response latencies at the end of baseline training, a mixed factor ANOVA revealed significant main effects of sex, F(1, 74) = 6.246, p = .015, trial type, F(1, 74) = 81.023, p < .001, and lever identity, F(1, 74) = 53.896, p < .001. Overall, males responded more quickly than females (difference of 0.181 s). Averaged across lever identity, rats responded more quickly during free-choice trials. Averaged across trial type, rats responded more quickly on the lever associated with the small, safe option. However, there was a significant trial type × lever identity interaction, F(1, 74) = 164.545, p < .001). Tukey’s post hoc tests revealed that response latencies did not differ between the large, risky option and the small, safe option during free choice trials (p = .140), but response latencies were significantly shorter for the small, safe option compared to the large, risky option (p < .001). Furthermore, response latencies for the large, risky option were significantly shorter during free-choice trials compared to forced-choice trials (p < .001), but response latencies did not differ for the small, safe option across trial type (p = .097). Figure 4a depicts this interaction.

Figure 4. — (a) Mean (± SEM) response latencies on the large, risky lever and the small, safe lever during free-choice and forced-choice trials. Panels b and c show the association between response latencies and normalized, log-transformed h parameter estimates for free-choice trials and forced-choice trials, respectively. The best-fitting regression is denoted by the solid line. *p < .05, compared to free-choice trials. #p < .05, compared to the small, safe lever.

Normalized, log-transformed h parameter estimates, as well as its interactions with sex and cocaine dose, did not predict response latencies during free-choice trials, R² = .025, F(4, 73) = 0.469, p = .758). Even if normalized, log-transformed h parameter estimates were entered into the regression model without the interaction terms, there was no association between normalized, log-transformed h parameter estimates and response latencies (R² = .004, p = .579; Fig. 4b). Conversely, when normalized, log-transformed h parameter estimates, as well as its interactions with sex and cocaine dose, were used to predict response latencies during forced-choice trials, the overall model was statistically significant, R² = .126, F(4, 73) = 2.642, p = .040. Normalized, log-transformed h parameter estimates were significantly associated with forced-choice response latencies only, B = 0.454, p = .013 (Fig. 4c).

3.2. Locomotor Activity During CPP Conditioning Sessions

Non-repeating photobeam breaks and repeating photobeam breaks were recorded during each conditioning session. Concerning non-repeating photobeam breaks (measure of horizontal activity) (Fig. 5a–b), the LME model revealed main effects of session, F(3, 515.306) = 3.282, p = .021, sex, F(1, 74.115) = 79.126, p < .001, cocaine dose, F(1, 74.115) = 16.906, p < .001, and treatment, F(1, 515.342) = 643.879, p < .001. There were several significant interactions: sex × cocaine dose, F(1, 74.115) = 11.689, p = .001, session × treatment, F(3, 515.340) = 4.789, p = .003, sex × treatment, F(1, 515.342) = 117.523, p < .001, cocaine dose × treatment, F(1, 515.342) = 10.941, p = .001, and sex × cocaine dose × treatment, F(1, 515.342) = 5.528, p = .019. Because the three-way interaction qualifies the main effects of sex, cocaine dose, and treatment, as well as the sex × cocaine dose, sex × treatment, and cocaine dose × treatment interactions, this interaction will be discussed in more detail. Bonferroni post hoc tests showed the following: (1) cocaine significantly increased locomotor activity compared to saline for both males and females treated with each dose of cocaine (all p’s < .001); (2) females were more active following each dose of cocaine compared to males (all p’s < .001); (3) females in the 20.0 mg/kg cocaine CPP experiment were more active than males following saline injections (p = .015); and (4) males were more active following a low dose of cocaine compared to a high dose of cocaine (p < .001).

Figure 5c depicts the session × treatment interaction. This interaction can be explained by the finding that cocaine-induced locomotor activity was higher on sessions 2–4 compared to session 1 (all p’s ≤ .026) whereas locomotor activity did not change across sessions following saline injections (all p’s = 1.000). Additionally, non-repeating photobeam breaks were higher following cocaine administration relative to saline administration for each session (all p’s < .001).

When examining repeating photobeam breaks (proxy for stereotypic behavior) (Fig. 5d–e), there were main effects of sex, F(1, 74.068) = 35.749, p < .001, cocaine dose, F(1, 74.068) = 12.247, p < .001, and treatment, F(1, 515.211) = 97.477, p < .001. There were significant sex × cocaine dose, F(1, 74.068) = 29.920, p < .001, sex × treatment, F(1, 515.211) = 28.551, p < .001, and sex × cocaine dose × treatment, F(1, 515.211) = 28.003, p < .001, interactions. As with non-repeating photobeam breaks, the significant three-way interaction qualifies each main effect and significant two-way interaction. Thus, only the three-way interaction will be discussed. Bonferroni post hoc tests showed the following: (1) both doses of cocaine increased repeating photobeam breaks relative to saline in female rats (p’s ≤ .002) while the lower dose of cocaine (10.0 mg/kg) increased repeating photobeam breaks in males (p = .003); (2) females in the 20.0 mg/kg cocaine CPP experiment had higher repeating photobeam breaks following saline and cocaine administration compared to males (p’s < .001); and (3) males in the 10.0 mg/kg cocaine CPP experiment had more repeating photobeam breaks following cocaine and saline administration compared to males in the 20.0 mg/kg cocaine CPP experiment (p’s < .001).

To examine the association between baseline risky decision making and locomotor activity, we first computed difference scores for both non-repeating photobeam breaks and repeating photobeam breaks (cocaine sessions - saline sessions). We then ran separate regression analyses (one for each type of photobeam break measurement), which included normalized, log-transformed h parameter estimates, sex, and cocaine dose as predictor variables (as well as all interactions). For non-repeating photobeam breaks, the overall regression model significantly predicted horizontal activity, R² = .464, F(7, 62) = 7.667, p < .001. However, the only variable that significantly predicted non-repeating photobeam breaks was sex (B = 745.387, p = .022) (data not shown). For repeating photobeam breaks, the overall regression model only trended toward statistical significance, R² = .190, F(7, 62) = .060 (data not shown).

3.3. Cocaine CPP

Across all four experimental conditions, rats spent more time in the white compartment (389.944 ± 5.864 s) compared to the gray compartment (186.932 ± 5.492 s) and to the black compartment (323.189 ± 5.279 s), F(2, 148) = 230.455, p < .001. Rats spent more time in the black compartment compared to the gray compartment. There were no other main effects or interactions, all F’s ≤ 1.719, all p’s ≥ .183. These results indicate that our CPP apparatus was biased in the current experiment.

Figure 6 shows cocaine CPP for each individual experiment (panels a–d). Overall, rats developed CPP to cocaine as evidenced by a main effect of test session, F(1, 74) = 59.331, p < .001. Females spent more time in the cocaine-paired compartment compared to males, F(1, 74) = 8.714, p = .004 (main effect of sex). Furthermore, there was a sex × test session interaction, F(1, 74) = 14.117, p < .001. Averaged across cocaine dose, both males and females spent more time in the cocaine-paired compartment during the posttest compared to the pretest (p’s of .036 and < .001, respectively). While males and females spent a near equivalent time in the cocaine-paired compartment during the pretest (p = .971) females spent significantly more time in this compartment during the posttest (p < .001). Finally, there was a cocaine dose × test session interaction, F(1, 74) = 5.174, p = .026. This interaction can be explained by the finding that rats treated with a higher dose of cocaine (20.0 mg/kg) developed a greater CPP (difference score of 112.269 s) compared to rats treated with a lower dose of cocaine (difference score of 61.079 s).

Response latencies during forced-choice trials, as well as interactions between response latencies and sex and cocaine dose, predicted CPP difference scores, R² = .123, F(4, 73) = 2.553, p = .046. However, none of the individual variables significantly predicted CPP difference scores, although there was a trend toward a significant interaction between response latencies and sex (B = 38.901, p = .058) (data not shown). Response latencies during free choice trials did not predict CPP difference scores, nor did sex or cocaine dose moderate the association between response latencies and cocaine CPP, R² = .037, F(4, 73) = .696, p = .597 (data not shown).

To determine if non-repeating photobeam breaks, sex, cocaine dose, and all possible interactions predicted CPP scores, these variables were entered into a multiple linear regression model. This model significantly predicted CPP difference scores, R² = .500, F(7, 69) = 9.876, p < .001. Sex was the only variable to predict non-repeating photobeam breaks, B = 880.189, p < .001, although there was a trend for cocaine dose to predict non-repeating photobeam breaks, B = −439.805, p = .057. Similar to the results obtained with non-repeating photobeam breaks, a regression model containing repeating photobeam breaks, sex, cocaine dose, and all possible interactions significantly predicted CPP difference scores, R² = .261, F(7, 69) = 3.489, p = .003. Sex was a significant predictor of repeating photobeam breaks, B = 490.996, p = .048. There was also a significant interaction between normalized, log-transformed h parameter estimates and sex, B = −515.726, p = .010.

The overall linear regression model with movement difference score (non-repeating photobeam breaks/s in the cocaine compartment - non-repeating photobeam breaks/s in the saline compartment), sex, and cocaine dose (as well as all interactions) significantly predicted CPP difference scores, R² = .494, F(7, 69) = 9.625, p < .001. Movement difference scores significantly predicted CPP differences scores, B = −85.549, p = .049, as did the three-way interaction, B = 178.205, p = .035. To examine the three-way interaction, bivariate correlations were conducted for each individual experimental group. The association between movement difference scores and CPP difference scores was negative for males treated with either dose of cocaine (10.0 mg/kg: r = −.522, p = .022; 20.0 mg/kg: r = −.742, p < .001) (Fig. 6e–6f). In females, a statistically significant negative correlation was only observed for females treated with the lower dose of cocaine, r = −.783, p < .001 (Fig. 6g), although a trend toward a negative correlation was observed for females treated with the higher dose of cocaine, r = −.392, p = .088 (Fig. 6h). These results show that greater horizontal activity (per s) in the saline-paired compartment relative to the cocaine-paired compartment was associated with increased cocaine CPP.

3.4. Association between Risk-Taking Behavior and Cocaine CPP

The multiple linear regression model with normalized, log-transformed h parameter estimates, sex, and cocaine dose (as well as all possible interactions) was a significant predictor of CPP difference scores, R² = .265, F(7, 69) = 3.551, p = .003. Sex, B = 188.474, p < .001, and the sex × normalized, log-transformed h parameter estimate interaction, B = −105.851, p = .012, significantly predicted CPP difference scores. Figure 6a depicts the significant sex × normalized, log-transformed h parameter estimate interaction. Corroborating the results reported in Section 3.3, females had higher difference scores compared to males, indicating greater cocaine CPP. Interestingly, risk averse males developed greater CPP compared to risk-taking males whereas risk-taking females developed greater CPP compared to risk averse females. Because there was a significant interaction between sex and normalized h parameter estimates, we performed separate correlation analyses for males and for females. When collapsed across cocaine dose, there was no significant correlation between normalized, log-transformed h parameter estimates and CPP difference scores for males (r(36) = .267, p = .102), but there was a significant negative correlation for females (r(38) = −.361, p = .022).

The regression model with mean choice for the large magnitude reinforcer, sex, and cocaine dose (as well as all possible interactions) was a significant predictor of difference scores, R² = .230, F(7, 70) = 2.990, p = .008. While the overall model significantly predicted difference scores, none of the variables were significant predictors of difference scores. Figure 6b shows the correlation between mean choice and CPP difference scores for males and females.

Because we found that non-repeating photobeam breaks/s (cocaine - saline) predicted CPP difference scores, we performed a hierarchical regression analysis to determine if adding this variable better accounts for the proportion of variance in CPP difference scores compared to the interaction between sex and normalized, log-transformed h parameter estimates. The first step of the hierarchical regression model included normalized, log-transformed h parameter estimates, sex, and their interaction. The second step included non-repeating photobeam breaks/s, the interaction between non-repeating photobeam breaks/s and normalized, log-transformed h parameter estimates, and a three-way interaction between non-repeating photobeam breaks/s, normalized, log-transformed h parameter estimates, and sex. Adding non-repeating photobeam breaks/s to the regression analysis did not better explain the proportion of variance in cocaine CPP difference scores, ΔR² = .025, p = .493. Even with the additional variables, the interaction between normalized, log-transformed h parameter estimates and sex remained statistically significant, B = −134.449, p = .012.

3.5. Amphetamine Drug Discrimination

Because low risk-taking male rats tended to show decreased cocaine (10.0 mg/kg) CPP, we conducted a drug discrimination experiment to determine if these findings could be potentially related to differential sensitivity to the interoceptive cues of psychostimulants. Normalized, log-transformed h parameter estimates did not predict the number of sessions meeting discrimination criteria for either dose of amphetamine, R²’s ≤ .016, p’s ≥ .606 (Fig. 8a–8b).

Figure 8. — Results of the amphetamine drug discrimination experiment for males originally tested in the 10.0 mg/kg cocaine CPP experiment. Panels a and b show correlations between normalized, log-transformed h parameter estimates and the number of individual sessions meeting stability criteria for each dose of amphetamine. Note, for each panel there were two rats that had the same normalized, log-transformed h parameter estimate and the same number of sessions meeting stability criteria. Thus, these points overlap with one another. Panels c and d show the number of responses averaged across the final four amphetamine sessions and the final four saline sessions for each dose of amphetamine. Panels e and f show correlations between normalized, log-transformed h parameter estimates and responses (amphetamine – saline) for each dose of amphetamine. *p < .05, compared to amphetamine. #p < .05, compared to saline responses from the 0.3 mg/kg amphetamine experiment. @p < .05, compared to 0.3 mg/kg amphetamine. The best-fitting regression is denoted by the solid line in panels a, b, e, and f.

During drug discrimination, responses were higher following saline treatment compared to amphetamine treatment, F(1, 18) = 161.769, p < .001. There was also a significant interaction between treatment and amphetamine dose, F(1, 18) = 25.869, p < .001. Tukey’s post hoc tests showed that responses for saline were higher for the 0.5 mg/kg experiment compared to the 0.3 mg/kg experiment (p = .002) and that responses for amphetamine (0.5 mg/kg) were lower compared to responses for amphetamine (0.3 mg/kg) (p = .001) (Fig. 8c–8d). Normalized, log-transformed h parameter estimates did not predict the number of lever responses (amphetamine sessions - saline sessions) for either dose of amphetamine, R²’s ≤ .105, p’s ≥ .176 (Fig. 8e–8f).

4. Discussion

The RDT we used in the current experiment is similar to the one used by Orsini et al. (2016), although we made some minor alterations. We reduced the number of trials from 90 to 70 and used a 20-s limited hold as opposed to a 10-s limited hold. Despite these alterations, rats acquired the RDT at a similar rate as reported in previous studies (Orsini et al., 2020; Simon et al., 2009). Similar to previous work (e.g., Gabriel et al., 2019; Mitchell et al., 2014; Orsini et al., 2020; Simon et al., 2009), we observed large individual differences in the proportion of responses for the large, risky option. Importantly, the individual differences observed in the current study do not appear to be attributed to differences in reward motivation, anxiety, and pain tolerance, as Simon et al. (2009) did not observe any correlations between RDT performance and these other constructs.

Not only were individual differences observed within each condition, but h parameter estimates differed across individual conditions. Males treated with the higher dose of cocaine (20.0 mg/kg) and females treated with the lower dose of cocaine (10.0 mg/kg) showed greater discounting of the large, risky reinforcer compared to males treated with the lower dose of cocaine and to females treated with the higher dose of cocaine. This discrepancy may be due to the differential training history observed across experimental conditions. Males treated with 20.0 mg/kg cocaine and females treated 10.0 mg/kg received 33 total sessions of the RDT once shock was introduced, whereas rats in the other conditions received 25 and 21 sessions, respectively. There is support to this argument as discounting rates increase as a function of training in delay- and probability-discounting tasks (e.g., Aparicio et al., 2015; St. Onge & Floresco, 2009; Yates et al., 2017). As the RDT is structured similarly to the Evenden and Ryan (1996) delay-discounting task and the Cardinal and Howes (2005) probability-discounting task, the increased sensitivity to the large, risky option observed in the conditions with extended training is not surprising.

In the current experiment, we performed a regression analysis to determine if normalized, log-transformed h parameter estimates, sex, and cocaine dose predicted CPP. We found that females developed greater cocaine CPP compared to males. These results are consistent with previous studies that have observed sex differences in cocaine CPP (Russo et al., 2003; Zakharova et al., 2009). Of direct interest to the present study was the finding that sex moderated the association between risky decision making and cocaine CPP. Whereas CPP scores increased as the level of risky decision making increased in female rats, the opposite association was observed for male rats. One caveat is that individual correlation analyses did not reveal a statistically significant association between normalized, log-transformed h parameter estimates and CPP difference scores for males. Thus, these data need to be interpreted with some caution. Because previous studies using the RDT used mean choice for the large, risky option as a predictor variable for abuse-related behavior (e.g., Mitchell et al., 2014; Orsini et al., 2020), we performed a second regression analysis using this variable instead of normalized, log-transformed h parameter estimates. While sex moderated the association between normalized, log-transformed h parameter estimates and cocaine CPP, no such finding was observed when we used the mean proportion of responses for the large, risky option. This finding is interesting as mean choice and normalized, log-transformed h parameter estimates were highly correlated. By only considering mean choice during blocks 2–5, we had a large proportion of rats that had values near 0 (see Fig. 7b), which may have precluded our ability to detect significant correlations between mean choice and CPP. By normalizing log-transformed h parameter estimates, we were able to generate values that were more centered (see Fig. 7a). As such, we performed a supplemental analysis in which we centered mean choice values before performing the regression analysis. After centering mean choice, the interaction between risky decision making and sex only trended toward significance (p = .098). Despite failing to detect a significant interaction when using mean choice, the pattern of results obtained with this analysis was similar to those obtained when using normalized, log-transformed h parameter estimates.

Figure 7. — (a) Correlation between normalized, log-transformed h parameter estimates and CPP difference scores for males (open circles and dashed line) and for females (closed squares and solid line). (b) Correlation between the mean proportion of responses for the large, risky option and CPP difference scores for males (open circles and dashed line) and for females (closed squares and solid line). The best-fitting regression is denoted by the solid/dashed lines.

The finding that increased risky decision making in males was not predictive of cocaine CPP contrasts with previous studies showing increased drug abuse vulnerability in HiR males (Gabriel et al., 2019; Mitchell et al., 2014; Orsini et al., 2020). One major analytic difference between the current study and previous studies examining the relationship between RDT performance and drug abuse vulnerability merits attention. We chose to exclude rats from statistical analyses if they failed to meet certain stability criteria (see Section 2.5.1). Specifically, we excluded rats that failed to discount the large, risky option because we could not determine if their performance was indicative of increased risky decision making or an insensitivity to foot shock. We also excluded animals that chose the large, risky option in less than 50% of free-choice trials during the first block of trials as this indicates that the animal fails to understand the contingencies of reinforcement. Excluding these animals may have prevented us from replicating the positive association that has been observed between risky decision making and drug self-administration in male rats (Mitchell et al., 2014; Orsini et al., 2020). Yet, this explanation does not appear to fully account for the differences obtained in the current study and in the Orsini et al. (2020) study. When examining the correlations reported in Orsini et al. (2020), there appears to be large variability in cocaine self-administration in rats that showed exclusive preference for the large, risky option. For example, two of the highest risk-taking rats self-administered less cocaine on day 14 compared to day 6 of an escalation paradigm. The significant correlation reported by Orsini et al. (2020) does not appear to be dependent on the performance of the most risk-taking rats.

Another important consideration that may account for the discrepancies across studies is that CPP and drug self-administration measure distinct aspects of addiction-like behavior (Bardo and Bevins, 2000), and as such, there have been cases in which a group of animals will self-administer more drug but will show decreased CPP relative to another group of animals (e.g., rats raised in isolation self-administer more drug but show decreased CPP compared to rats raised in an enriched environment; Bardo et al., 1995; Bardo et al., 2001; Bowling and Bardo, 1994; Green et al., 2002). At face value, the current results suggest that LoR male rats tend to be more sensitive to the conditioned rewarding effects of cocaine compared to HiR male rats. Importantly, cocaine (10.0 mg/kg) was behaviorally active in HiR and LoR rats, as non-repeating photobeams (i.e., horizontal activity) increased following cocaine administration relative to saline administration. The low variability observed across both cocaine doses and across each sex suggests that cocaine produced similar increases in locomotor activity for all rats. Additionally, the increased CPP observed in LoR rats cannot be explained by differential sensitivity to the locomotor-stimulant effects of cocaine, as non-repeating photobeams across conditioning sessions was uncorrelated with normalized, log-transformed h parameter estimates (data not shown).

One potential explanation for the discrepant findings observed in the current study and those using drug self-administration is that increased risk taking in males is associated with higher reward thresholds. In self-administration, higher reward thresholds mean that animals need to respond more drug infusions to reach their threshold. This view is consistent with the increased drug self-administration observed in HiR male rats compared to LoR male rats (Gabriel et al., 2019; Mitchell et al., 2014; Orsini et al., 2020). In CPP, drug doses below the reward threshold will fail to produce CPP, as was the case with the lower dose of cocaine (10.0 mg/kg) in the current experiment. In females, the relationship between reward thresholds and cocaine self-administration and CPP may be reversed. That is, LoR females may have higher reward thresholds compared to HiR females. There is some support to this argument as LoR females self-administer more cocaine during acquisition compared to HiR females (Orsini et al., 2020) while risky decision making in females was positively correlated with cocaine CPP in the current experiment. Differential reward thresholds observed between HiR males and females may result from alterations in punishment-induced activation of the hypothalamic-pituitary-adrenal (HPA) axis. For example, females exposed to chronic unpredictable mild foot shocks have greater blood plasma corticosterone levels compared to males (Lu et al., 2015). Given that corticosterone is an important mediator of drug self-administration (see Goeders, 2002), probabilistic delivery of foot shock during the RDT could potentially cause alterations in the HPA axis in HiR females that could manifest as increased cocaine CPP but decreased cocaine self-administration.

While differential reward thresholds provide the most parsimonious explanation for the discrepant relationships between risky decision making and cocaine CPP/drug self-administration, we wanted to rule out the possibility that the increased CPP observed in LoR male rats could be attributed to differential sensitivity to the interoceptive cues of the drug, we conducted an amphetamine drug discrimination experiment. Because rats had previously been treated with cocaine during the CPP experiment, we wanted to use a different psychostimulant drug that rats had not been exposed to previously. HiR rats and LoR rats learned to discriminate the amphetamine-paired and saline-paired levers at a similar rate, and the percentage of rats in each group that learned to make this discrimination was similar. These data suggest that LoR rats are not necessarily more sensitive to the interoceptive cues of the drug. Additional work examining why LoR male rats are more sensitive to the conditioned rewarding effects of psychostimulants relative to HiR male rats is merited. Even though previous research has shown no differences in working memory as measured in a delayed response task (Shimp et al., 2015), perhaps alterations in long-term memory may explain, at least in part, the results observed in the current study. Future studies can test this potential alternative explanation by testing HiR rats and LoR rats in a radial arm maze, which can measure both working memory and reference (i.e., long-term) memory (Halas et al., 1983; Mei et al., 2020). This additional experiment could provide further insights on the behavioral mechanisms underlying enhanced cocaine CPP in LoR male rats.

One interesting observation is that increased discounting across different tasks predicts CPP for psychostimulants in male rats. In the current study, LoR rats showed greater discounting of the large, risky option. In a previous study, male rats that showed greater delay discounting (i.e., increased impulsive choice) showed enhanced amphetamine CPP (Yates et al., 2012). Pilot work from our lab has replicated the results of Yates et al. (2012) using a probability-discounting task; that is, rats that show greater discounting of probabilistic reinforcement (i.e., increased risk aversion) develop greater amphetamine CPP compared to rats that show shallower discounting (Yates et al. unpublished results). There has been some debate as to whether these discounting tasks measure dissociable constructs (e.g., impulsive choice vs. risky decision making) or measure a similar behavioral mechanism (Green et al., 1999; Green et al., 2010; Myerson and Green, 1995; Rachlin et al., 1991; see Green and Myerson, 2004 for a review). Performance in the RDT is correlated with probability discounting (Simon et al., 2009) but is not correlated with delay discounting (Shimp et al., 2015; Simon et al., 2009). Inconsistencies have also been reported concerning the relationship between probability discounting and delay discounting, as some studies have found positive correlations between each task (Baumann and Odum, 2012; Jones and Rachlin, 2009; Richards et al., 1999) while others have failed to observe any correlations (Andrade and Petry, 2012; Johnson et al., 2015; Takahashi et al., 2007). Despite these inconsistencies, one potential hypothesis is that for males, the rate of discounting may be a better predictor of CPP compared to constructs such as impulsivity or risky decision making. This account may also explain why increased risky decision making in a rat gambling task predicts methamphetamine CPP in males (Takahashi et al., 2020), as this task does not measure discounting of a reinforcer like the RDT.

Although we did not observe differences in cocaine-induced increases in locomotor activity across HiR and LoR rats, previous research has demonstrated that HiR rats show increased locomotor activity following first-time nicotine administration (Gabriel et al., 2019). Because we originally averaged across all conditioning sessions, we performed an additional analysis in which we examined the relationship between risky decision making and horizontal activity during the first cocaine conditioning session. We did not observe any association between normalized, log-transformed h parameter estimates or non-repeating photobeam breaks (data not shown). There are important methodological differences between the current study and the one conducted by Gabriel et al. (2019) that make direct comparisons difficult. First, in the current experiment, rats were housed individually, whereas Gabriel et al. (2019) pair-housed rats. Previous research has shown that pair-housing animals can augment drug-induced locomotor sensitization (Trujillo and Heller, 2020). By isolating rats in our experiment, we may have precluded the ability to observe individual differences in cocaine-induced increases in locomotor activity. Additionally, locomotor activity was interpreted as the number of non-repeating photobeam breaks during 30-min sessions in the current experiment; however, Gabriel et al. (2019) recorded total distance traveled (in cm) in an open field for a total of 2 hr (30-min baseline, 30 min following saline injection, and 1 hr following nicotine injection). Measuring locomotion in an open field can better determine if there are subtle differences in sensitivity to the acute and/or long-term stimulant effects of cocaine between HiR rats and LoR rats.

One limitation to the current experiment needs to be addressed. We originally conducted CPP (10.0 mg/kg) with male rats and then conducted a follow-up experiment with a separate group of rats to test a higher dose of cocaine (20.0 mg/kg). Although we ran both groups of females simultaneously, we did not counterbalance CPP groups based on baseline discounting rates. Instead, we randomly assigned rats to each condition before conducting the experiment. Given the differences in baseline discounting observed across groups, interpreting the correlations between risky decision making and cocaine CPP should be made with some caution. However, this limitation is mitigated by the orderly data observed in the current study. Individual differences in risky decision making predicted cocaine CPP when a lower dose was tested, as originally hypothesized. We also found that females developed greater cocaine CPP independent of baseline risky decision making, another finding we had hypothesized. Regardless, we acknowledge that a better approach would have been to match each rat according to baseline h parameter estimates before assigning them to each CPP experiment.

In conclusion, we found that increased risky decision making is positively associated with CPP to a moderate dose if cocaine (10.0 mg/kg) in female rats, which is consistent with past findings showing increased vulnerability in this population. However, we showed that decreased risk taking is somewhat positively associated with increased cocaine (10.0 mg/kg) CPP in male rats, and this preference does not appear to be mediated by the locomotor stimulant effects or the interoceptive cues of psychostimulant drugs. Because Pavlovian conditioning plays a pivotal role in SUDs (Hinson et al., 1986; Martin-Fardon and Weiss, 2013), using the CPP paradigm is important for increasing our understanding of the relationship between risk-taking behavior and addiction. Given that environmental stimuli play a prominent role in relapse (Fuchs et al., 2008), determining if excessive risk taking is a risk factor for relapse-like behavior is an important topic that deserves investigation.

Research Highlights.

Sex moderates the association between risky decision making and cocaine CPP
High risk taking in male rats tended to be negatively associated with cocaine CPP
High risk taking in female rats was positively associated with cocaine CPP
Females developed greater cocaine CPP compared to males

Role of Funding Source

None of the funding sources were involved in the study design, analysis, interpretation, or writing of the current manuscript.

The current study was supported by NIH grant R15DA047610 and NIGMS grant P20GM103436.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Convergence issues arose when animals showed extremely steep discounting. Specifically, when animals showed an “L-shaped” discounting curve (i.e., exclusive preference for the small, safe option during blocks 2–5), GraphPad could not fit the hyperbolic function. By adding a constant of 0.04, the function could be fit to the data for each rat.

Conflict of Interest

None of the authors have any conflicts of interest.

References

Adida M, Clark L, Pomietto P, Kaladijian A, Besnier N, Azorin J-M, et al. (2008). Lack of insight may predict impaired decision making in manic patients. Bipolar Disord 10, 829–837. doi: 10.1111/j.1399-5618.2008.00618.x [DOI] [PubMed] [Google Scholar]
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author. [Google Scholar]
Andrade L, and Petry NM (2012). Delay and probability discounting in pathological gamblers with and without a history of substance use problems. Psychopharmacology 219, 491–499. doi: 10.1007/s00213-011-2508-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
Anker JJ, Perry JL, Gliddon LA, and Carroll ME (2009). Impulsivity predicts the escalation of cocaine self-administration in rats. Pharmacol. Biochem. Behav 93, 343–348. doi: 10.1016/j.pbb.2009.05.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
Aparicio CF, Elcoro M, and Alonso-Alvarez B (2015). A long-term study of the impulsive choices of Lewis and Fischer 344 rats. Learn. Behav 43, 251–271. 10.3758/s13420-015-0177-y [DOI] [PubMed] [Google Scholar]
Bardo MT, and Bevins RA (2000). Conditioned place preference: What does it add to our preclinical understanding of drug reward? Psychopharmacology 153, 31–43. doi: 10.1007/s002130000569 [DOI] [PubMed] [Google Scholar]
Bardo MT, Bowling SL, Rowlett JK, Manderscheid P, Buxton ST, and Dwoskin LP (1995). Environmental enrichment attenuates locomotor sensitization, but not in vitro dopamine release, induced by amphetamine. Pharmacol. Biochem. Behav 51, 397–405. doi: 10.1016/0091-3057(94)00413-d [DOI] [PubMed] [Google Scholar]
Bardo MT, Klebaur JE, Valone JM, and Deaton C (2001). Environmental enrichment decreases intravenous self-administration of amphetamine in female and male rats. Psychopharmacology 155, 278–284. doi: 10.1007/s002130100720 [DOI] [PubMed] [Google Scholar]
Baumann AA, and Odum AL (2012). Impulsivity, risk taking, and timing. Behav. Processes 90, 408–414. doi: 10.1016/j.beproc.2012.04.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
Belin D, Mar AC, Dalley JW, Robbins TW, and Everitt BJ (2008). High impulsivity predicts the switch to compulsive cocaine-taking. Science 320, 1352–1355. doi: 10.1126/science.1158136 [DOI] [PMC free article] [PubMed] [Google Scholar]
Bevins RA, Klebaur JE, and Bardo MT (1997). Individual differences in response to novelty, amphetamine-induced activity and drug discrimination in rats. Behav. Pharmacol 8, 113–123. [PubMed] [Google Scholar]
Boutros N, Semenova S, Liu W, Crews FT, and Markou A (2014). Adolescent intermittent ethanol exposure is associated with increased risky choice and decreased dopaminergic and cholinergic neuron markers in adult rats. Int. J. Neuropsychopharmacol, 18, pyu003. doi: 10.1093/ijnp/pyu003 [DOI] [PMC free article] [PubMed] [Google Scholar]
Bowling SL, and Bardo MT (1994). Locomotor and rewarding effects of amphetamine in enriched, social, and isolate reared rats. Pharmacol. Biochem. Behav 48, 459–464. doi: 10.1016/0091-3057(94)90553-3 [DOI] [PubMed] [Google Scholar]
Brand M, Roth-Bauer M, Driessen M, and Markowitsch HJ (2008). Executive functions and risky decision-making in patients with opiate dependence. Drug Alcohol Depend 97, 64–72. doi: 10.1016/j.drugalcdep.2008.03.017 [DOI] [PubMed] [Google Scholar]
Brevers D, Bechara A, Cleeremans A, Kornreich C, Verbanck P, and Noël X (2014). Impaired decision-making under risk in individuals with alcohol dependence. Alcohol. Clin. Exp. Res 38, 1924–1931. doi: 10.1111/acer.12447 [DOI] [PMC free article] [PubMed] [Google Scholar]
Brevers D, Bechara A, Cleeremans A, and Noël X (2013). Iowa Gambling Task (IGT): Twenty years after – Gambling disorder and IGT. Front. Psychol 4, 665. doi: 10.3389/fpsyg.2013.00665 [DOI] [PMC free article] [PubMed] [Google Scholar]
Cardinal RN, and Howes NJ (2005). Effects of lesions of the nucleus accumbens core on choice between small certain rewards and large uncertain rewards in rats. BMC Neurosci 6, 37. 10.1186/1471-2202-6-37 [DOI] [PMC free article] [PubMed] [Google Scholar]
Dalley JW, Fryer TD, Brichard L, Robinson ESJ, Theobald DEH, Lääne K, et al. (2007). Nucleus accumbens D2/3 receptors predict trait impulsivity and cocaine reinforcement. Science 315, 1267–1270. doi: 10.1126/science.1137073 [DOI] [PMC free article] [PubMed] [Google Scholar]
Dong G, and Potenza MN (2016). Risk-taking and risky decision-making in Internet gaming disorder: Implications regarding online gaming in the setting of negative consequences. J. Psychiatr. Res 73, 1–8. doi: 10.1016/j.jpsychires.2015.11.011 [DOI] [PubMed] [Google Scholar]
Evenden JL, and Ryan CN (1996). The pharmacology of impulsive behaviour in rats: the effects of drugs on response choice with varying delays of reinforcement. Psychopharmacology 128, 161–170. 10.1007/s002130050121 [DOI] [PubMed] [Google Scholar]
Ferland J-MN, and Winstanley CA (2017). Risk-preferring rats make worse decisions and show increased incubation of craving after cocaine self-administration. Addict. Biol 22, 991–1001. doi: 10.1111/adb.12388 [DOI] [PubMed] [Google Scholar]
Fuchs RA, Lasseter HC, Ramirez DR, and Xie X (2008). Relapse to drug seeking following prolonged abstinence: the role of environmental stimuli. Drug Discov. Today Dis. Models 5, 251–258. 10.1016/j.ddmod.2009.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
Gabriel DBK, Freels TG, Setlow B, and Simon NW (2019). Risky decision-making is associated with impulsive action and sensitivity to first-time nicotine exposure. Behav. Brain Res 359, 579–588. doi: 10.1016/j.bbr.2018.10.008 [DOI] [PubMed] [Google Scholar]
Goeders NE (2002). The HPA axis and cocaine reinforcement. Psychoneuroendocrinology 27, 13–33. doi: 10.1016/s0306-4530(01)00034-8 [DOI] [PubMed] [Google Scholar]
Green TA, Gehrke BJ, and Bardo MT (2002). Environmental enrichment decreases intravenous amphetamine self-administration in rats: dose-response functions for fixed-and progressive-ratio schedules. Psychopharmacology 162, 373–378. doi: 10.1007/s00213-002-1134-y [DOI] [PubMed] [Google Scholar]
Green L, and Myerson J (2004). A discounting framework for choice with delayed and probabilistic rewards. Psychol. Bull 130, 769–792. doi: 10.1037/0033-2909.130.5.769 [DOI] [PMC free article] [PubMed] [Google Scholar]
Green L, Myerson J, and Calvert AL (2010). Pigeons’ discounting of probabilistic and delayed reinforcers. J. Exp. Anal. Behav 94, 113–123. doi: 10.1901/jeab.2010.94-113 [DOI] [PMC free article] [PubMed] [Google Scholar]
Green L, Myerson J, and Ostaszewski P (1999). Amount of reward has opposite effects on the discounting of delayed and probabilistic outcomes. J. Exp. Anal. Psychol. Learn. Mem. Cogn 25, 418–427. doi: 10.1037//0278-7393.25.2.418 [DOI] [PubMed] [Google Scholar]
Halas ES, Eberhardt MJ, Diers MA, and Sandstead HH (1983). Learning and memory impairment in adult rats due to severe zinc deficiency during lactation. Physiol. Behav 30, 371–381. doi: 10.1016/0031-9384(83)90140-3 [DOI] [PubMed] [Google Scholar]
Herman AM, Critchley HD, and Duka T (2018). Risk-taking and impulsivity: The role of mood states and interoception. Front. Psychol 9, 1625. doi: 10.3389/fpsyg.2018.01625 [DOI] [PMC free article] [PubMed] [Google Scholar]
Hilderbrand ER, and Lasek AW (2014). Sex differences in cocaine conditioned place preference in C57BL/6J mice. Neuroreport 25, 105–109. doi: 10.1097/WNR.0000000000000053 [DOI] [PMC free article] [PubMed] [Google Scholar]
Hinson RE, Poulos CX, Thomas W, and Cappell H (1986). Pavlovian conditioning and addictive behavior: Relapse to oral self-administration of morphine. Behav. Neurosci 100, 368–375. doi: 10.1037//0735-7044.100.3.368 [DOI] [PubMed] [Google Scholar]
Johnson MW, and Bickel WK (2008). An algorithm for identifying nonsystematic delay-discounting data. Exp. Clin. Psychopharmacol 16, 264–274. 10.1037/1064-1297.16.3.264 [DOI] [PMC free article] [PubMed] [Google Scholar]
Johnson MW, Johnson PS, Herrmann ES, and Sweeny MM (2015). Delay and probability discounting of sexual and monetary outcomes in individuals with cocaine use disorders and matched controls. PLoS One 10, e0128641. doi: 10.1371/journal.pone.0128641 [DOI] [PMC free article] [PubMed] [Google Scholar]
Jones BA, and Rachlin H (2009). Delay, probability, and social discounting in a public goods game. J. Exp. Anal. Behav 91, 61–73. doi: 10.1901/jeab.2009.91-61 [DOI] [PMC free article] [PubMed] [Google Scholar]
Klebaur JE, and Bardo MT (1999). Individual differences in novelty seeking on the playground maze predict amphetamine conditioned place preference. Pharmacol. Biochem. Behav 63, 131–136. doi: 10.1016/s0091-3057(98)00258-5 [DOI] [PubMed] [Google Scholar]
Leland DS, and Paulus MP (2005). Increased risk-taking decision-making but not altered response to punishment in stimulant-using young adults. Drug Alcohol Depend 78, 83–90. doi: 10.1016/j.drugalcdep.2004.10.001 [DOI] [PubMed] [Google Scholar]
Lu J, Wu XY, Zhu QB, Li J, Shi LG, Wu JL, Zhang QJ, Huang ML, and Bao AM (2015). Sex differences in the stress response in SD rats. Behav. Brain Res 284, 231–237. doi: 10.1016/j.bbr.2015.02.009 [DOI] [PubMed] [Google Scholar]
Martin-Fardon R, and Weiss F (2013). Modeling relapse in animals. Curr. Top. Behav. Neurosci 13, 403–432. doi: 10.1007/7854_2012_202 [DOI] [PMC free article] [PubMed] [Google Scholar]
Marusich JA, and Bardo MT (2009). Differences in impulsivity on a delay-discounting task predict self-administration of a low unit dose of methylphenidate in rats. Behav. Pharmacol 20, 447–454. doi: 10.1097/FBP.0b013e328330ad6d [DOI] [PMC free article] [PubMed] [Google Scholar]
Mathews IZ, Morrissey MD, and McCormick CM (2010). Individual differences in activity predict locomotor activity and conditioned place preference to amphetamine in both adolescent and adult rats. Pharmacol. Biochem. Behav 95, 63–71. doi: 10.1016/j.pbb.2009.12.007 [DOI] [PubMed] [Google Scholar]
Mei J, Kohler J, Winter Y, Spies C, Endres M, Banneke S, et al. (2020). Automated radial 8-arm maze: A voluntary and stress-free behavior test to assess spatial learning and memory in mice. Behav. Brain Res 381, 112352. doi: 10.1016/j.bbr.2019.112352 [DOI] [PubMed] [Google Scholar]
Meyer AC, Horton DB, Neugebauer NM, Wooters TE, Nickell JR, Dwoskin LP, et al. (2011). Tetrabenaxine inhibition of monoamine uptake and methamphetamine behavioral effects: Locomotor activity, drug discrimination and self-administration. Neuropharmacology 61, 849–856. doi: 10.1016/j.neuropharm.2011.05.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
Mitchell MR, Weiss VG, Beas BS, Morgan D, Bizon JL, and Setlow B (2014). Adolescent risk taking, cocaine self-administration, and striatal dopamine signaling. Neuropsychopharmacology 39, 955–962. doi: 10.1038/npp.2013.295 [DOI] [PMC free article] [PubMed] [Google Scholar]
Morris Bobzean SA, Dennis TS, Addison BD, and Perrotti LI (2010). Influence of sex on reinstatement of cocaine-conditioned place preference. Brain Res. Bull 83, 331–336. doi: 10.1016/j.brainresbull.2010.09.003 [DOI] [PubMed] [Google Scholar]
Myerson J, and Green L (1995). Discounting of delayed rewards: Models of individual choice. J. Exp. Anal. Behav 64, 263–276. doi: 10.1901/jeab.1995.64-263 [DOI] [PMC free article] [PubMed] [Google Scholar]
National Research Council. (2011). Guide for the care and use of laboratory animals (8th edition). Washington, DC: National Academies Press. [Google Scholar]
Oberlender R, and Nichols DE (1988). Drug discrimination studies with MDMA and amphetamine. Psychopharmacology 95, 71–76. doi: 10.1007/BF00212770 [DOI] [PubMed] [Google Scholar]
Orsini CA, Blaes SL, Dragone RJ, Betzhold SM, Finner AM, Bizon JL, et al. (2020). Distinct relationships between risky decision making and cocaine self-administration under short- and long-access conditions. Prog. Neuropsychopharmacol. Biol. Psychiatry 98, 109791. doi: 10.1016/j.pnpbp.2019.109791 [DOI] [PMC free article] [PubMed] [Google Scholar]
Orsini CA, Willis ML, Gilbert RJ, Bizon JL, and Setlow B (2016). Sex differences in a rat model of risky decision making. Behav. Neurosci 130, 50–61. doi: 10.1037/bne0000111 [DOI] [PMC free article] [PubMed] [Google Scholar]
Pinheiro J, Bates D, DebRoy S, Sarkar D, and R Core Team (2019). nlme: Linear and Nonlinear Mixed Effects Models R Foundation for Statistical Computing, Vienna, Austria. https://CRAN.R-project.org/package=nlme [Google Scholar]
Rachlin H, Raineri A, and Cross D (1991). Subjective probability and delay. J. Exp. Anal. Behav 55, 233–244. doi: 10.1901/jeab.1991.55-233 [DOI] [PMC free article] [PubMed] [Google Scholar]
Richards JB, Zhang L, Mitchell SH, and de Wit H (1999). Delay and probability discounting in a model of impulsive behavior: Effect of alcohol. J. Exp. Anal. Behav 71, 121–143. doi: 10.1901/jeab.1999.71-121 [DOI] [PMC free article] [PubMed] [Google Scholar]
Russo SJ, Jenab S, Fabian SJ, Festa ED, Kemen LM, and Quinones-Jenab V (2003). Sex differences in the conditioned rewarding effects of cocaine. Brain Res 970, 214–220. doi: 10.1016/s0006-8993(03)02346-1 [DOI] [PubMed] [Google Scholar]
Schutter DJ, van Bokhoven I, Vanderschuren LJ, Lochman JE, and Matthys W (2011). Risky decision making in substance dependent adolescents with a disruptive behavior disorder. J. Abnorm. Child Psychol 39, 333–339. doi: 10.1007/s10802-010-9475-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
Shimp KG, Mitchell MR, Beas BS, Bizon JL, and Setlow B (2015). Affective and cognitive mechanisms of risky decision making. Neurobiol. Learn. Mem 117, 60–70. doi: 10.1016/j.nlm.2014.03.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
Simon NW, Gilbert RJ, Mayse JD, Bizon JL, and Setlow B (2009). Balancing risk and reward: A rat model of risky decision making. Neuropsychopharmacology 34, 2208–2217. doi:10.1038.npp.2009.48 [DOI] [PMC free article] [PubMed] [Google Scholar]
Spoelder M, Lesscher HM, Hesseling P, Baars AM, Lozeman-van t Klooster JG, Mijnsbergen R, et al. (2015). Altered performance in a rat gambling task after acute and repeated alcohol exposure. Psychopharmacology 232, 3649–3662. doi: 10.1007/s00213-015-4020-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
St. Onge JR, and Floresco SB (2009). Dopaminergic modulation of risk-based decision making. Neuropsychopharmacology 34, 681–697. 10.1038/npp.2008.121 [DOI] [PubMed] [Google Scholar]
Takahashi T, Ikeda K, and Hasegawa T (2007). A hyperbolic decay of subjective probability of obtaining delayed rewards. Behav. Brain Funct 3, 52. doi: 10.1186/1744-9081-3-52 [DOI] [PMC free article] [PubMed] [Google Scholar]
Takahashi K, Toyoshima M, Ichitani Y, and Yamada K (2020). Enhanced methamphetamine-induced conditioned place preference in risk-taking rats. Behav. Brain Res 378, 112299. 10.1016/j.bbr.2019.112299 [DOI] [PubMed] [Google Scholar]
Trujillo KA, and Heller CY (2020). Ketamine sensitization: Influence of dose, environment, social isolation and treatment interval. Behav. Brain Res 378, 112271. doi: 10.1016/j.bbr.2019.112271 [DOI] [PMC free article] [PubMed] [Google Scholar]
Van Enhuizen J, Henry BL, Minassian A, Perry W, Milienne-Petiot M, Higa KK, et al. (2014). Reduced dopamine transporter functioning induces high-reward risk-preference consistent with bipolar disorder. Neuropsychopharmacology 39, 3112–3122. doi: 10.1038/npp.2014.170 [DOI] [PMC free article] [PubMed] [Google Scholar]
Yates JR, Gunkel BT, Rogers KK, Hughes MN, and Prior NA (2017). Effects of N-methyl-D-aspartate receptor ligands on sensitivity to reinforcer magnitude and delayed reinforcement in a delay-discounting procedure. Psychopharmacology 234, 461–473. 10.1007/s00213-016-4469-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
Yates JR, Horchar MJ Ellis AL, Kappesser JL, Mbambu P, Sutphin TG, Dehner DS, Igwe HO, Wright MR (2021). Differential effects of glutamate N-methyl-D-aspartate receptor antagonists on risky choice as assessed in the risky decision task. Psychopharmacology 238, 133–148. doi: 10.1007/s00213-020-05664-z [DOI] [PMC free article] [PubMed] [Google Scholar]
Yates JR, Marusich JA, Gipson CD, Beckmann JS, and Bardo MT (2012). High impulsivity in rats predicts amphetamine conditioned place preference. Pharmacol. Biochem. Behav 100, 370–376. doi: 10.1016/j.pbb.2011.07.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
Zakharova E, Wade D, and Izenwasser S (2009). Sensitivity to cocaine conditioned reward depends on sex and age. Pharmacol. Biochem. Behav 92, 131–134. doi: 10.1016/j.pbb.2008.11.002 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R1] Adida M, Clark L, Pomietto P, Kaladijian A, Besnier N, Azorin J-M, et al. (2008). Lack of insight may predict impaired decision making in manic patients. Bipolar Disord 10, 829–837. doi: 10.1111/j.1399-5618.2008.00618.x [DOI] [PubMed] [Google Scholar]

[R2] American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author. [Google Scholar]

[R3] Andrade L, and Petry NM (2012). Delay and probability discounting in pathological gamblers with and without a history of substance use problems. Psychopharmacology 219, 491–499. doi: 10.1007/s00213-011-2508-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] Anker JJ, Perry JL, Gliddon LA, and Carroll ME (2009). Impulsivity predicts the escalation of cocaine self-administration in rats. Pharmacol. Biochem. Behav 93, 343–348. doi: 10.1016/j.pbb.2009.05.013 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] Aparicio CF, Elcoro M, and Alonso-Alvarez B (2015). A long-term study of the impulsive choices of Lewis and Fischer 344 rats. Learn. Behav 43, 251–271. 10.3758/s13420-015-0177-y [DOI] [PubMed] [Google Scholar]

[R6] Bardo MT, and Bevins RA (2000). Conditioned place preference: What does it add to our preclinical understanding of drug reward? Psychopharmacology 153, 31–43. doi: 10.1007/s002130000569 [DOI] [PubMed] [Google Scholar]

[R7] Bardo MT, Bowling SL, Rowlett JK, Manderscheid P, Buxton ST, and Dwoskin LP (1995). Environmental enrichment attenuates locomotor sensitization, but not in vitro dopamine release, induced by amphetamine. Pharmacol. Biochem. Behav 51, 397–405. doi: 10.1016/0091-3057(94)00413-d [DOI] [PubMed] [Google Scholar]

[R8] Bardo MT, Klebaur JE, Valone JM, and Deaton C (2001). Environmental enrichment decreases intravenous self-administration of amphetamine in female and male rats. Psychopharmacology 155, 278–284. doi: 10.1007/s002130100720 [DOI] [PubMed] [Google Scholar]

[R9] Baumann AA, and Odum AL (2012). Impulsivity, risk taking, and timing. Behav. Processes 90, 408–414. doi: 10.1016/j.beproc.2012.04.005 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] Belin D, Mar AC, Dalley JW, Robbins TW, and Everitt BJ (2008). High impulsivity predicts the switch to compulsive cocaine-taking. Science 320, 1352–1355. doi: 10.1126/science.1158136 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R11] Bevins RA, Klebaur JE, and Bardo MT (1997). Individual differences in response to novelty, amphetamine-induced activity and drug discrimination in rats. Behav. Pharmacol 8, 113–123. [PubMed] [Google Scholar]

[R13] Boutros N, Semenova S, Liu W, Crews FT, and Markou A (2014). Adolescent intermittent ethanol exposure is associated with increased risky choice and decreased dopaminergic and cholinergic neuron markers in adult rats. Int. J. Neuropsychopharmacol, 18, pyu003. doi: 10.1093/ijnp/pyu003 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] Bowling SL, and Bardo MT (1994). Locomotor and rewarding effects of amphetamine in enriched, social, and isolate reared rats. Pharmacol. Biochem. Behav 48, 459–464. doi: 10.1016/0091-3057(94)90553-3 [DOI] [PubMed] [Google Scholar]

[R15] Brand M, Roth-Bauer M, Driessen M, and Markowitsch HJ (2008). Executive functions and risky decision-making in patients with opiate dependence. Drug Alcohol Depend 97, 64–72. doi: 10.1016/j.drugalcdep.2008.03.017 [DOI] [PubMed] [Google Scholar]

[R16] Brevers D, Bechara A, Cleeremans A, Kornreich C, Verbanck P, and Noël X (2014). Impaired decision-making under risk in individuals with alcohol dependence. Alcohol. Clin. Exp. Res 38, 1924–1931. doi: 10.1111/acer.12447 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] Brevers D, Bechara A, Cleeremans A, and Noël X (2013). Iowa Gambling Task (IGT): Twenty years after – Gambling disorder and IGT. Front. Psychol 4, 665. doi: 10.3389/fpsyg.2013.00665 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R18] Cardinal RN, and Howes NJ (2005). Effects of lesions of the nucleus accumbens core on choice between small certain rewards and large uncertain rewards in rats. BMC Neurosci 6, 37. 10.1186/1471-2202-6-37 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] Dalley JW, Fryer TD, Brichard L, Robinson ESJ, Theobald DEH, Lääne K, et al. (2007). Nucleus accumbens D2/3 receptors predict trait impulsivity and cocaine reinforcement. Science 315, 1267–1270. doi: 10.1126/science.1137073 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] Dong G, and Potenza MN (2016). Risk-taking and risky decision-making in Internet gaming disorder: Implications regarding online gaming in the setting of negative consequences. J. Psychiatr. Res 73, 1–8. doi: 10.1016/j.jpsychires.2015.11.011 [DOI] [PubMed] [Google Scholar]

[R21] Evenden JL, and Ryan CN (1996). The pharmacology of impulsive behaviour in rats: the effects of drugs on response choice with varying delays of reinforcement. Psychopharmacology 128, 161–170. 10.1007/s002130050121 [DOI] [PubMed] [Google Scholar]

[R22] Ferland J-MN, and Winstanley CA (2017). Risk-preferring rats make worse decisions and show increased incubation of craving after cocaine self-administration. Addict. Biol 22, 991–1001. doi: 10.1111/adb.12388 [DOI] [PubMed] [Google Scholar]

[R23] Fuchs RA, Lasseter HC, Ramirez DR, and Xie X (2008). Relapse to drug seeking following prolonged abstinence: the role of environmental stimuli. Drug Discov. Today Dis. Models 5, 251–258. 10.1016/j.ddmod.2009.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] Gabriel DBK, Freels TG, Setlow B, and Simon NW (2019). Risky decision-making is associated with impulsive action and sensitivity to first-time nicotine exposure. Behav. Brain Res 359, 579–588. doi: 10.1016/j.bbr.2018.10.008 [DOI] [PubMed] [Google Scholar]

[R25] Goeders NE (2002). The HPA axis and cocaine reinforcement. Psychoneuroendocrinology 27, 13–33. doi: 10.1016/s0306-4530(01)00034-8 [DOI] [PubMed] [Google Scholar]

[R26] Green TA, Gehrke BJ, and Bardo MT (2002). Environmental enrichment decreases intravenous amphetamine self-administration in rats: dose-response functions for fixed-and progressive-ratio schedules. Psychopharmacology 162, 373–378. doi: 10.1007/s00213-002-1134-y [DOI] [PubMed] [Google Scholar]

[R27] Green L, and Myerson J (2004). A discounting framework for choice with delayed and probabilistic rewards. Psychol. Bull 130, 769–792. doi: 10.1037/0033-2909.130.5.769 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R28] Green L, Myerson J, and Calvert AL (2010). Pigeons’ discounting of probabilistic and delayed reinforcers. J. Exp. Anal. Behav 94, 113–123. doi: 10.1901/jeab.2010.94-113 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] Green L, Myerson J, and Ostaszewski P (1999). Amount of reward has opposite effects on the discounting of delayed and probabilistic outcomes. J. Exp. Anal. Psychol. Learn. Mem. Cogn 25, 418–427. doi: 10.1037//0278-7393.25.2.418 [DOI] [PubMed] [Google Scholar]

[R30] Halas ES, Eberhardt MJ, Diers MA, and Sandstead HH (1983). Learning and memory impairment in adult rats due to severe zinc deficiency during lactation. Physiol. Behav 30, 371–381. doi: 10.1016/0031-9384(83)90140-3 [DOI] [PubMed] [Google Scholar]

[R31] Herman AM, Critchley HD, and Duka T (2018). Risk-taking and impulsivity: The role of mood states and interoception. Front. Psychol 9, 1625. doi: 10.3389/fpsyg.2018.01625 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R32] Hilderbrand ER, and Lasek AW (2014). Sex differences in cocaine conditioned place preference in C57BL/6J mice. Neuroreport 25, 105–109. doi: 10.1097/WNR.0000000000000053 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] Hinson RE, Poulos CX, Thomas W, and Cappell H (1986). Pavlovian conditioning and addictive behavior: Relapse to oral self-administration of morphine. Behav. Neurosci 100, 368–375. doi: 10.1037//0735-7044.100.3.368 [DOI] [PubMed] [Google Scholar]

[R34] Johnson MW, and Bickel WK (2008). An algorithm for identifying nonsystematic delay-discounting data. Exp. Clin. Psychopharmacol 16, 264–274. 10.1037/1064-1297.16.3.264 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] Johnson MW, Johnson PS, Herrmann ES, and Sweeny MM (2015). Delay and probability discounting of sexual and monetary outcomes in individuals with cocaine use disorders and matched controls. PLoS One 10, e0128641. doi: 10.1371/journal.pone.0128641 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R36] Jones BA, and Rachlin H (2009). Delay, probability, and social discounting in a public goods game. J. Exp. Anal. Behav 91, 61–73. doi: 10.1901/jeab.2009.91-61 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R37] Klebaur JE, and Bardo MT (1999). Individual differences in novelty seeking on the playground maze predict amphetamine conditioned place preference. Pharmacol. Biochem. Behav 63, 131–136. doi: 10.1016/s0091-3057(98)00258-5 [DOI] [PubMed] [Google Scholar]

[R38] Leland DS, and Paulus MP (2005). Increased risk-taking decision-making but not altered response to punishment in stimulant-using young adults. Drug Alcohol Depend 78, 83–90. doi: 10.1016/j.drugalcdep.2004.10.001 [DOI] [PubMed] [Google Scholar]

[R39] Lu J, Wu XY, Zhu QB, Li J, Shi LG, Wu JL, Zhang QJ, Huang ML, and Bao AM (2015). Sex differences in the stress response in SD rats. Behav. Brain Res 284, 231–237. doi: 10.1016/j.bbr.2015.02.009 [DOI] [PubMed] [Google Scholar]

[R40] Martin-Fardon R, and Weiss F (2013). Modeling relapse in animals. Curr. Top. Behav. Neurosci 13, 403–432. doi: 10.1007/7854_2012_202 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R41] Marusich JA, and Bardo MT (2009). Differences in impulsivity on a delay-discounting task predict self-administration of a low unit dose of methylphenidate in rats. Behav. Pharmacol 20, 447–454. doi: 10.1097/FBP.0b013e328330ad6d [DOI] [PMC free article] [PubMed] [Google Scholar]

[R42] Mathews IZ, Morrissey MD, and McCormick CM (2010). Individual differences in activity predict locomotor activity and conditioned place preference to amphetamine in both adolescent and adult rats. Pharmacol. Biochem. Behav 95, 63–71. doi: 10.1016/j.pbb.2009.12.007 [DOI] [PubMed] [Google Scholar]

[R43] Mei J, Kohler J, Winter Y, Spies C, Endres M, Banneke S, et al. (2020). Automated radial 8-arm maze: A voluntary and stress-free behavior test to assess spatial learning and memory in mice. Behav. Brain Res 381, 112352. doi: 10.1016/j.bbr.2019.112352 [DOI] [PubMed] [Google Scholar]

[R44] Meyer AC, Horton DB, Neugebauer NM, Wooters TE, Nickell JR, Dwoskin LP, et al. (2011). Tetrabenaxine inhibition of monoamine uptake and methamphetamine behavioral effects: Locomotor activity, drug discrimination and self-administration. Neuropharmacology 61, 849–856. doi: 10.1016/j.neuropharm.2011.05.033 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] Mitchell MR, Weiss VG, Beas BS, Morgan D, Bizon JL, and Setlow B (2014). Adolescent risk taking, cocaine self-administration, and striatal dopamine signaling. Neuropsychopharmacology 39, 955–962. doi: 10.1038/npp.2013.295 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R46] Morris Bobzean SA, Dennis TS, Addison BD, and Perrotti LI (2010). Influence of sex on reinstatement of cocaine-conditioned place preference. Brain Res. Bull 83, 331–336. doi: 10.1016/j.brainresbull.2010.09.003 [DOI] [PubMed] [Google Scholar]

[R47] Myerson J, and Green L (1995). Discounting of delayed rewards: Models of individual choice. J. Exp. Anal. Behav 64, 263–276. doi: 10.1901/jeab.1995.64-263 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R49] National Research Council. (2011). Guide for the care and use of laboratory animals (8th edition). Washington, DC: National Academies Press. [Google Scholar]

[R50] Oberlender R, and Nichols DE (1988). Drug discrimination studies with MDMA and amphetamine. Psychopharmacology 95, 71–76. doi: 10.1007/BF00212770 [DOI] [PubMed] [Google Scholar]

[R51] Orsini CA, Blaes SL, Dragone RJ, Betzhold SM, Finner AM, Bizon JL, et al. (2020). Distinct relationships between risky decision making and cocaine self-administration under short- and long-access conditions. Prog. Neuropsychopharmacol. Biol. Psychiatry 98, 109791. doi: 10.1016/j.pnpbp.2019.109791 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R52] Orsini CA, Willis ML, Gilbert RJ, Bizon JL, and Setlow B (2016). Sex differences in a rat model of risky decision making. Behav. Neurosci 130, 50–61. doi: 10.1037/bne0000111 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R53] Pinheiro J, Bates D, DebRoy S, Sarkar D, and R Core Team (2019). nlme: Linear and Nonlinear Mixed Effects Models R Foundation for Statistical Computing, Vienna, Austria. https://CRAN.R-project.org/package=nlme [Google Scholar]

[R54] Rachlin H, Raineri A, and Cross D (1991). Subjective probability and delay. J. Exp. Anal. Behav 55, 233–244. doi: 10.1901/jeab.1991.55-233 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R55] Richards JB, Zhang L, Mitchell SH, and de Wit H (1999). Delay and probability discounting in a model of impulsive behavior: Effect of alcohol. J. Exp. Anal. Behav 71, 121–143. doi: 10.1901/jeab.1999.71-121 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R56] Russo SJ, Jenab S, Fabian SJ, Festa ED, Kemen LM, and Quinones-Jenab V (2003). Sex differences in the conditioned rewarding effects of cocaine. Brain Res 970, 214–220. doi: 10.1016/s0006-8993(03)02346-1 [DOI] [PubMed] [Google Scholar]

[R57] Schutter DJ, van Bokhoven I, Vanderschuren LJ, Lochman JE, and Matthys W (2011). Risky decision making in substance dependent adolescents with a disruptive behavior disorder. J. Abnorm. Child Psychol 39, 333–339. doi: 10.1007/s10802-010-9475-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R58] Shimp KG, Mitchell MR, Beas BS, Bizon JL, and Setlow B (2015). Affective and cognitive mechanisms of risky decision making. Neurobiol. Learn. Mem 117, 60–70. doi: 10.1016/j.nlm.2014.03.002 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R59] Simon NW, Gilbert RJ, Mayse JD, Bizon JL, and Setlow B (2009). Balancing risk and reward: A rat model of risky decision making. Neuropsychopharmacology 34, 2208–2217. doi:10.1038.npp.2009.48 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R60] Spoelder M, Lesscher HM, Hesseling P, Baars AM, Lozeman-van t Klooster JG, Mijnsbergen R, et al. (2015). Altered performance in a rat gambling task after acute and repeated alcohol exposure. Psychopharmacology 232, 3649–3662. doi: 10.1007/s00213-015-4020-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R61] St. Onge JR, and Floresco SB (2009). Dopaminergic modulation of risk-based decision making. Neuropsychopharmacology 34, 681–697. 10.1038/npp.2008.121 [DOI] [PubMed] [Google Scholar]

[R62] Takahashi T, Ikeda K, and Hasegawa T (2007). A hyperbolic decay of subjective probability of obtaining delayed rewards. Behav. Brain Funct 3, 52. doi: 10.1186/1744-9081-3-52 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R63] Takahashi K, Toyoshima M, Ichitani Y, and Yamada K (2020). Enhanced methamphetamine-induced conditioned place preference in risk-taking rats. Behav. Brain Res 378, 112299. 10.1016/j.bbr.2019.112299 [DOI] [PubMed] [Google Scholar]

[R64] Trujillo KA, and Heller CY (2020). Ketamine sensitization: Influence of dose, environment, social isolation and treatment interval. Behav. Brain Res 378, 112271. doi: 10.1016/j.bbr.2019.112271 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R65] Van Enhuizen J, Henry BL, Minassian A, Perry W, Milienne-Petiot M, Higa KK, et al. (2014). Reduced dopamine transporter functioning induces high-reward risk-preference consistent with bipolar disorder. Neuropsychopharmacology 39, 3112–3122. doi: 10.1038/npp.2014.170 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R66] Yates JR, Gunkel BT, Rogers KK, Hughes MN, and Prior NA (2017). Effects of N-methyl-D-aspartate receptor ligands on sensitivity to reinforcer magnitude and delayed reinforcement in a delay-discounting procedure. Psychopharmacology 234, 461–473. 10.1007/s00213-016-4469-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R100] Yates JR, Horchar MJ Ellis AL, Kappesser JL, Mbambu P, Sutphin TG, Dehner DS, Igwe HO, Wright MR (2021). Differential effects of glutamate N-methyl-D-aspartate receptor antagonists on risky choice as assessed in the risky decision task. Psychopharmacology 238, 133–148. doi: 10.1007/s00213-020-05664-z [DOI] [PMC free article] [PubMed] [Google Scholar]

[R67] Yates JR, Marusich JA, Gipson CD, Beckmann JS, and Bardo MT (2012). High impulsivity in rats predicts amphetamine conditioned place preference. Pharmacol. Biochem. Behav 100, 370–376. doi: 10.1016/j.pbb.2011.07.012 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R68] Zakharova E, Wade D, and Izenwasser S (2009). Sensitivity to cocaine conditioned reward depends on sex and age. Pharmacol. Biochem. Behav 92, 131–134. doi: 10.1016/j.pbb.2008.11.002 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

The association between risky decision making and cocaine conditioned place preference is moderated by sex

Justin R Yates

Matthew J Horchar

Joy L Kappesser

Maria R Broderick

Alexis L Ellis

Makayla R Wright

Abstract

Background.

Methods.

Results.

Conclusions.

1. Introduction

2. Materials and Methods

2.1. Animals

2.2. Apparatus

2.3. Procedure

2.3.1. RDT.

Table 1,

2.3.2. CPP.

2.3.3. Amphetamine Drug Discrimination.

2.4. Drugs

2.5. Statistical Analyses

2.5.1. RDT Performance.

Figure 1.

2.5.2. Locomotor Activity During CPP Conditioning Sessions.

2.5.3. Cocaine CPP.

2.5.4. Association between RDT Performance and Cocaine CPP.

2.5.5. Amphetamine Drug Discrimination.

3. Results

3.1. Baseline RDT Performance

Figure 2.

Figure 3.

Figure 4.

3.2. Locomotor Activity During CPP Conditioning Sessions

Figure 5.

3.3. Cocaine CPP

Figure 6.

3.4. Association between Risk-Taking Behavior and Cocaine CPP

3.5. Amphetamine Drug Discrimination

Figure 8.

4. Discussion

Figure 7.

Research Highlights.

Role of Funding Source

Footnotes

References

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases