A meta-analysis of the effects of episodic future thinking on delay discounting

Abstract

Delay discounting (DD) refers to the phenomenon in which the subjective value of future rewards is reduced over time. There are individual differences in the DD rate, and increased discounting has been observed in those with various psychiatric disorders. Episodic future thinking (EFT) is the act of vividly imagining events that may happen in the future. Studies have shown that EFT could reduce DD, although inconsistent results have been reported. The aim of this meta-analysis was to clarify the efficacy with which EFT reduces DD and to identify potential moderators. Forty-seven studies (including 63 contrasts) were included in the final analysis. EFT was found to significantly reduce DD (Hedges’ g = 0.52). Moderator analysis showed that positive EFT (g = 0.64) was more effective in reducing DD than EFT with the valence not specifically mentioned (g = 0.28) and EFT with neutral or negative valence (g = –0.03). In addition, several factors related to the control task and DD task were related to the efficacy of EFT to reduce DD. These findings have implications for using EFT to reduce DD in the future.

Keywords

Delay discounting episodic future thinking meta-analysis

Introduction

In daily life, people often need to make choices between short-term and long-term rewards, such as choosing to enjoy smoking now or stop smoking and be healthier in the future (Lempert & Phelps, 2016). The phenomenon that the subjective value of future rewards is reduced over time is known as delay discounting (DD), which is also called time discounting or temporal discounting (Frederick et al., 2002; Sellitto et al., 2011). That is, during decision-making, people often prefer to choose an immediate but smaller reward than a delayed but larger reward (Logue, 1988). An increased DD rate has been highly associated with many maladaptive behaviours, such as smoking (MacKillop et al., 2011), alcohol addiction (Bobova et al., 2009), and gambling problems (Reynolds, 2006). It is also a transdiagnostic process in psychiatric disorders, and a meta-analysis demonstrated that increased DD has been observed across psychiatric disorders, including major depressive disorder, schizophrenia, bipolar disorder, and eating disorders (Amlung et al., 2019; Lempert et al., 2019). Given that DD is a relatively stable characteristic of individuals and its association with unhealthy behaviours and psychiatric disorders, a growing number of studies have examined ways to reduce the DD rate and change maladaptive behaviours (Scholten et al., 2019).

Two studies (Rung & Madden, 2018; Scholten et al., 2019) summarised different types of manipulations or trainings used to decrease DD, including mindfulness-based interventions (Hendrickson & Rasmussen, 2017; Yao et al., 2017), contingency management (Giles et al., 2014; Stanger et al., 2013), acceptance and commitment therapy (Morrison et al., 2014), visualisation training (Parthasarathi et al., 2017), and episodic future thinking (EFT; Bulley & Gullo, 2017). Of these methods, EFT is one of the promising manipulations for decreasing the DD rate (Scholten et al., 2019). EFT is the act of projecting oneself into the future and vividly imagining possible future events (Schacter et al., 2007, 2017). EFT is engaged in adaptive behaviours, including far-sighted decision-making and emotion regulation (Brocas & Carrillo, 2018; Schacter et al., 2017; Schacter & Madore, 2016).

To examine the effect of EFT on DD, researchers usually compare discounting rates between conditions in which participants were asked to imagine future events before making choices and while participants made choices under a control condition (e.g., nontemporal mental imagery before making choices) (Bulley et al., 2019). Previous studies found that EFT could reduce DD in both healthy individuals (Scholten et al., 2019) and special populations such as smokers (Chiou & Wu, 2017), obese individuals (Daniel et al., 2013a), and people with alcohol use disorders (Snider et al., 2016). In addition to reducing money-based DD rates, EFT also works on maladaptive behaviours, such as reducing demand for cigarettes (Stein et al., 2016), alcohol (Bulley & Gullo, 2017), and calorie intake (Daniel et al., 2013b).

Generally, EFT is effective in reducing DD, and effect sizes have ranged from 0.00 to 1.49 (Hollis-Hansen et al., 2019; Rung & Madden, 2018). A few studies failed to find an effect of EFT on DD (Hu, Kleinschmidt, et al., 2017; Liu et al., 2013; Palombo et al., 2016; Zhang et al., 2018). Liu et al. (2013) and Zhang et al. (2018) found that compared with the control condition, negative EFT did not reduce the DD rate and even increased the DD rate. Palombo et al. (2016) and Hu, Kleinschmidt, et al. (2017) found that EFT effectively reduced DD in healthy participants but not in amnestic patients. The large variations in effect size may be attributed to study characteristics, including EFT-related factors, control task–related factors, DD task–related factors, and participant-related factors.

Regarding EFT-related factors, the emotional valence of EFT may be associated with its effect on DD (Liu et al., 2013). Positive EFT reduced DD, while neutral and negative EFT did not affect DD and even increased the DD rate (Zhang et al., 2018). Previous studies also found that the more details and vividness associated with EFT, the greater the reduction in DD (Benoit et al., 2011; Kim et al., 2013; Peters & Büchel, 2010; Zhang et al., 2018). The time remoteness of events was shown to be related to the details and vividness of EFT; specifically, recent EFT showed more details and higher vividness than remote EFT events (Arnold et al., 2011). However, it remains unknown whether the delay time in EFT affects the effect of EFT on DD.

When examining the effect of EFT on DD, studies have used different control tasks as the comparison condition. For example, some studies used episodic recent thinking as the control task, for example, participants were asked to recall events that happened within the past 24 hr before making choices (Daniel et al., 2015, 2016); some studies used other types of control tasks, such as storytelling or imagining routine events (Bulley et al., 2019; Bulley & Gullo, 2017), while some studies did not require participants to imagine or recall any event and to simply complete the DD task in the control condition (Sasse et al., 2015). Whether different contents associated with the control condition lead to different degrees of DD reduction is not yet known (Scholten et al., 2019).

Furthermore, the DD task has varied from study to study. For example, there was a great variation in delay time, ranging from 7 days (Cheng et al., 2012) to 25 years (Stein et al., 2017). The delay time may be a potential variable that influences the effect of EFT on DD. In addition, some studies used area under the curve (AUC) as the index of DD (Athamneh et al., 2020; Jia et al., 2020), while some other studies used k (discounting rate) as the index of DD (Dassen et al., 2016; Wu et al., 2017); it might be useful to examine whether different effects would be found with different indices.

Another group of factors is related to participants, that is, age of participants and population of participants (general healthy participants vs. special populations such as smokers). Mok et al. (2020) suggested that the effect of EFT on the DD rate was smaller in older adults than in young adults, while Sasse et al. (2017) did not find an effect of EFT on DD reduction in older adults. In addition, EFT was found to reduce DD in both healthy individuals (Scholten et al., 2019) and in special populations, such as smokers, obese individuals, and people with cannabis use disorders (Chiou & Wu, 2017; Daniel et al., 2013a; Sofis et al., 2020), and whether EFT shows differential effects on DD in different populations remains unknown.

Taken together, there is growing attention to the effects of EFT on reducing DD, and no published study has systematically examined its efficacy and which factors affect this efficacy. The present study aimed to provide a meta-analysis to examine this issue. We considered four groups of factors, including EFT-related factors, control task–related factors, DD task–related factors, and participant-related factors, and examined whether these factors moderate the effect of EFT on DD.

Method

Literature search

Literature searches in Web of Science, Google Scholar, and Springer were conducted with the following keywords: (“delay discounting” OR “temporal discounting” OR “time discount*” OR “intertemporal choice” OR “inter* decision making”) AND (“future thinking” OR “prospection” OR “episodic future thinking” OR “episodic future thought” OR “imagining the future” OR “episodic simulation” OR “future envisioning”). The literature search was from 1990 to 30 July 2021.

Study selection

The literature search was restricted to peer-reviewed papers published in English. A total of 2,612 potential papers were identified from the literature search, and an additional 37 papers were identified through reference lists and citations from related articles. Studies were selected based on the following criteria: (1) the full-text article was available (not a conference abstract); (2) the study was an empirical study, not a review, comment, or meta-analysis; (3) the study was not a case study; (4) the study measured how EFT impacted DD; and (5) the study reported sufficient data to calculate effect sizes. For the studies fulfilling previous criteria but did not include sufficient data to calculate effect sizes, we contacted the authors to provide additional data. If no further data were provided, the studies were excluded from the final analysis. Finally, 47 papers including 63 contrasts were included in the meta-analysis. We followed the PRISMA standard for systematic review and meta-analysis in the process of study selection (Moher et al., 2009). Title and abstract screening of publications was executed by J.Y. In the case of uncertainty, the full paper was read and discussed by J.Y. and Y.W. until an agreement was reached. The subsequent screening processes were conducted by J.Y. and Y.W. The process of literature screening based on the PRISMA flowchart is shown in Figure 1.

Figure 1.

Flow diagram of article selection.

Data extraction

For each included paper, the following data were extracted: first, the basic information of the study, including the first author and the year of publication; the sample size and mean age of participants; and the type of participants (healthy participants, special populations) were extracted. Second, data for calculating effect sizes of EFT on DD were extracted. For between-subject design studies, the mean and SD on DD measures were extracted; if the mean and SD were not available, other data that could be used to calculate effect sizes, such as t values and sample sizes, were extracted. Sixteen studies included more than one experiment or different groups of participants, and we calculated a separate effect size for each experiment or contrast for these papers. If several studies had an overlap in participants, we only included one of these studies. For example, there were two studies that had a large overlap in patients (Palombo et al., 2015, 2016), and we included only the former paper when calculating the effect size in the patients. Sofis et al. (2020) included DD for gains and DD for losses. Since all other studies were on DD for gains, we included only the DD for gains results in the present meta-analysis. Third, moderators were recorded for a moderator analysis or meta-regression analysis. Moderators were mainly classified into four types. (1) EFT-related moderators included the following: the valence of EFT; the context of EFT (e.g., personally relevant EFT or task-related EFT. For personally relevant EFT, participants were asked to imagine personally relevant future events but not related to rewards regarding the DD task; for task-related EFT, participants were asked to imagine events that were related to the reward, such as spending money at the delayed time); and the longest delay in EFT was also included as a moderator. (2) Control task–related moderators included the following: the context of the control task (e.g., no control task, recalling past event, other types such as storytelling or routine events), the valence of the control task, and the longest time distance in the control task (these analyses only included studies with recall past events as the control task). (3) DD task–related moderators included the following: the reward type (e.g., hypothetical or potentially real. Hypothetical reward means participants would not receive extra reward no matter what they choose in the DD task; potential real reward means participants may receive some reward based on their choices); the DD task type (e.g., whether the choices were prefixed or the choices varied based on participants’ responses); the outcome indices of DD (e.g., AUC, K [including k value, log k value and ln k value], proportion of choosing larger later reward, indifference point); and the longest delay in the DD task. (4) Participant-related moderators included age of the participants and population (healthy individuals or special populations).

Data analyses

The data were analysed using Comprehensive Meta-Analysis (Version 2.0; https://www.meta-analysis.com/; Borenstein et al., 2005). Hedges’ g was used as the index of effect size. We first examined the overall effect of EFT on DD. If more than one effect size could be calculated in an experiment, we used the mean of these effect sizes as the single effect size for this experiment in the analysis of the overall effect of EFT on DD. We then examined whether the effect of EFT on DD performance was related to the following variables using a moderator analysis or meta-regression: (1) valence of EFT, (2) context type of EFT, (3) longest delay of EFT, (4) context type of control task, (5) valence of control task, (6) longest time distance of control task, (7) DD reward type, (8) type of DD paradigm, (9) outcome index of DD, (10) longest delay of DD, (11) age of the participants, and (12) population of participants. To increase the power of the analysis, we only reported the results when the number of contrasts was at least four in each condition of the moderator analysis.

We reported the heterogeneity of the studies with the Q-statistic. If the heterogeneity was significant, we adopted the random effects model to report effect sizes; otherwise, we reported results using the fixed effects model. The moderator analyses adopted the random effects model. All significance levels were set at p < .05 (Hedges & Vevea, 1998). Publication bias was examined with the fail-safe N analysis, which indicated that the number of studies with null results needed to reject the present significant findings.

Results

Overall effect size of EFT on DD

The final analysis included 47 studies including 63 contrasts to examine the effect of EFT on DD. Table 1 provides a summary of these studies. The mean effect size (Hedges’ g) of EFT in reducing DD was 0.52, 95% confidence interval (CI) = [0.42, 0.63], suggesting that EFT manipulations can reduce DD with a medium effect size (see Figure 2). These studies were heterogeneous (Q = 151.30, p < .001). Publication bias analysis revealed that the classic fail-safe N was 3,419, which means 3,419 studies with null results were needed to reject the present significant results, while Orwin’s fail-safe N was 403, which means 403 studies with null results were needed to reduce the effect size by 50%. The fail-safe Ns were much larger than the number of contrasts included in the analysis (N = 63), and similar results were found after adjusting for publication bias, suggesting that publication bias was unlikely to explain the significant results.

Table 1.

Descriptions of studies included in the meta-analysis.

Study name		Participant	EFT group age	EFT-related moderators			Control task–related moderators			DD-related moderators
			EFT group age	EFT valence	EFT event type	EFT longest time	Control event type	Control valence	Control longest time	DD task money	DD task	DD index type	DD longest delay
Athamneh et al. (2020)	E1	Smoker	34.8	Positive	Personally relevant	16 years	Past events	Positive	3 days	HYP	Variable	AUC	16 years
		Smoker	34.8	Positive	Task-related	16 years	Past events	Positive	3 days	HYP	Variable	AUC	16 years
	E2	Obesity	39.2	Positive	Personally relevant	16 years	Past events	Positive	3 days	HYP	Variable	AUC	16 years
		Obesity	35.7	Positive	Task-related	16 years	Past events	Positive	3 days	HYP	Variable	AUC	16 years
Benoit et al. (2011)		HC	27.3		Task-related		Objective usage			HYP	Fixed	LLR%	360 days
Bromberg et al. (2017)		HC	13–16	Positive or neutral	Personally relevant	7 months	Visual perceptual			PR	Variable	AUC	7 months
												Log k
Bulley & Gullo (2017)		HC	20.67	Positive	Personally relevant	1 year	Storytelling			HYP	Variable	AUC	1 year
Bulley et al. (2019)	E1	HC	19.56	Negative	Personally relevant		Routine events	Neutral		HYP	Fixed	K	182 days
												Log k
												LLR
	E2	HC	19.56	Positive	Personally relevant		Routine events	Neutral		HYP	Fixed	K	182 days
												Log k
												LLR
Calluso et al. (2019)		HC	23.87	Negative	Personally relevant	180 days	No			HYP	Fixed	K	180 days
				Neutral
				Positive
Cheng et al. (2012)		HC		Positive	Personally relevant	4 years	Routine events	Neutral		HYP	Variable	Interest rate (%)	7 days
Chiou & Wu (2017)		Smoker	32.5	Positive	Task-related	1 year	Benefits of quitting smoke	Positive	1 year	HYP	Fixed	K	1 year
							No
Ciaramelli et al. (2019)		HC	35.4	Positive or neutral	Personally relevant	3 years	Past events	Positive or neutral	3 years	HYP	Variable	AUC	1 year
							Scene
							Present
Ciaramelli et al. (2021)		vmPFC lesion	57.4	Positive or neutral	Personally relevant	5 years	No			HYP	Variable	AUC	10 years
		HC	61.1
Daniel et al. (2013a)		Obesity		Positive	Personally relevant	2 years	Storytelling			HYP	Variable	AUC	2 years
Daniel et al. (2013b)		Obesity		Positive	Personally relevant	2 years	Blog events			HYP	Variable	AUC	2 years
Daniel et al. (2015)		Obesity	9	Positive	Personally relevant	6 months	Past events	Positive	1 day	HYP	Variable	AUC	6 months
Daniel et al. (2016)		HC	26.07	Positive	Personally relevant	6 months	Past events	Positive	1 day	HYP	Variable	AUC	6 months
Dassen et al. (2016)	E1	HC			Personally relevant	6 months	Past events		30 days	HYP	Fixed	k	186 days
	E2	HC			Task-related		Routine events	Neutral		HYP	Fixed	k	186 days
Hollis-Hansen et al. (2019)		HC	29	Positive	Personally relevant	1 year	Past events		6 days	HYP	Variable	AUC	2 years
							Standardised episodic thinking
Hu, Kleinschmidt, et al. (2017)	E1	HC	68.29		Personally relevant		No			HYP	Fixed	AUC	1 year
												ln k
	E2	Subjective cognitive decline	66.49		Personally relevant		No			HYP	Fixed	AUC	1 year
												ln k
Hu, Uhle, et al. (2017)		HC	24	Positive or neutral	Personally relevant	1 year	No			HYP	Variable	AUCln k percentage	1 year
Jia et al. (2020)		Schizotypy (HC)	20.94	Positive	Personally relevant	1 year	Past events	Positive	1 day	HYP	Variable	AUC	1 year
													1 year
Carr et al. (2021)		HC	32	Positive	Personally relevant (writing)	5 years	Past events	Positive	5 days	HYP	Variable	AUC	5 years
		HC	31.4	Positive	Personally relevant (drawing)
Lin et al. (2014)		HC	42.8	Positive or neutral	Personally relevant	6 months	Next day	Neutral	1 day	HYP	Fixed	ln k	186 days
Liu et al. (2013)	E1	HC	20.62	Positive	Personally relevant		No			PR	Variable	Ratio of choose larger reward	30 days
	E2	HC	20.74	Negative	Personally relevant		No			PR	Variable	Ratio of choose larger reward	30 days
	E3	HC	20	Neutral	Personally relevant		No			PR	Variable	Ratio of choose larger reward	30 days
Mansouri et al. (2021)		Overweight or obesity	28.39	Positive	Personally relevant	6 months	Past events	Positive	6 days	HYP	Variable	AUC	6 months
Mok et al. (2020)	E1	HC	20	Positive or neutral	Personally relevant	10 years	No			HYP	Variable	AUC	10 years
	E2	HC	64.1	Positive or neutral	Personally relevant	10 years	No			HYP	Variable	AUC	10 years
O’Donnell et al. (2017)		HC	22.2	Positive	Personally relevant	2 years	Past events	Positive	86 hr	HYP	Variable	AUC	2 years
O’Donnell et al. (2018)		HC	35	Positive	Personally relevant	2 years	Past events	Positive	86 hr	HYP	Variable	AUC	2 years
O’Donnell et al. (2019)	E1	HC	41.54	Positive	Personally relevant	2 years	Past events	Positive	86 hr	HYP	Variable	AUC	2 years
	E2	HC	35.17	Positive	Personally relevant	2 years	Past events	Positive	86 hr	HYP	Variable	AUC	2 years
Palombo et al. (2015)	E1	Amnesia	61.2		Task-related	2 years	No			HYP	Fixed	LLR%	2 years
	E2	HC	65		Task-related	2 years	No			HYP	Fixed	LLR%	2 years
Palombo et al. (2016)		HC	60.2		Task-related	2 years	No			HYP	Fixed	LLR%	2 years
Peter & Büchel (2010)		HC	25		Personally relevant	7 months	No			PR	Variable	Log k	233 days
Patel & Amlung (2020)		Alcohol-dependent	41.36	Positive	Personally relevant	6 months	Past events	Positive	7 days	HYP	Variable	Log k	6 months
Rung & Madden (2018)		HC	33.53	Positive	Personally relevant	1 year	Past events		1 day	HYP	Fixed	Ratio of choose larger reward	1 year
Rung & Madden (2019)	E1	HC		Positive	Personally relevant	5 years	Past events	Positive	1 day	HYP	Variable	Indifference points	6 months1 year
	E2	HC		Positive	Personally relevant	1 year	Past events	Positive	1 day	HYP	Variable	Indifference points	1 year
Sasse et al. (2015)		HC	24.96		Task-related	190 days	No			PR	Variable	k	190 days
Sasse et al. (2017)		HC	66.55		Task-related	190 days	No			PR	Variable	k	190 days
Snider et al. (2016)		Alcohol-dependent	38	Positive	Personally relevant	1 year	Past events	Positive	1 day	HYP	Variable	AUC	1 year
Sofis et al. (2020)		Cannabis users	35.1		Personally relevant	1 year	Past events		1 day	HYP	Variable	AUC	1 year
Stein et al. (2016)		Obesity	38.65	Positive	Personally relevant	1 year	Past events	Positive	1 day	HYP	Variable	AUC	1 year
Stein et al. (2017)		Obesity	36.7	Positive	Personally relevant	1 year	Past events	Positive	12 days	HYP	Variable	ln k	25 years
Sze et al. (2017)	E1	Obesity		Positive	Personally relevant	6 months	Past events	Positive	6 days	HYP	Fixed	Log k	182 days
Voss et al. (2021)		Heavy drinking	18.9	Positive	Personally relevant	3 years	Storytelling			HYP	Variable	k
Wiehler et al. (2017)		Gamblers	29.7		Personally relevant	~200 days	No			PR	Variable	Log k	~200 days
		HC	28.5
Craft et al. (2020)		Chronic pain	33.5	Positive	Personally relevant	25 years	Past events	Positive	3 days	HYP	Variable	ln k	25 years
Wu et al. (2017)	E1	HC		Positive	Personally relevant	1 year	Ideal self			PR	Fixed	k	1 year
							No
	E2	HC		Positive	Personally relevant	1 year	Ideal self			PR	Fixed	k	1 year
							Blog events
Yi et al. (2017)	E2	HC	20.4		Personally relevant	5 years	Present events		1 day	HYP	Variable	ln k	5 years
		HC	19.6
Zhang et al. (2018)	E1	HC	21.53	Positive	Personally relevant		No			PR	Fixed	Log k	4 months
	E2	HC	21.53	Negative	Personally relevant		No			PR	Fixed	Log k	4 months
	E3	HC	21.19	Neutral	Personally relevant		No			PR	Fixed	Log k	4 months

Note: EFT: episodic future thinking; DD: delay discounting; HC: healthy population; HYP DD: hypothetical delay discounting task; PR DD: potentially real delay discounting task; LLR: larger later reward.

Figure 2.

The funnel plot of the overall effect of EFT on DD.

Moderator and meta-regression analyses

EFT-related moderators

The effect of EFT valence

Positive EFT (k = 34, g = 0.64, p < .001), EFT with valence not particularly mentioned (k = 16, g = 0.28, p = .004), and positive or neutral EFT (k = 8, g = 0.66, p < .001) all reduced DD, while negative EFT (k = 4, g = –0.21, p = .524) or neutral EFT (k = 3, g = 0.21, p = .195) did not reduce DD. Considering the number of studies, we included studies with positive EFT and EFT with valence not mentioned in the moderator analysis, and the results indicated that the moderator effect of EFT valence was significant (Q = 10.77, p = .001). We further combined studies using negative EFT and neutral EFT (k = 7, g = –0.03, p = .892), and compared these studies to studies with positive EFT, the moderator effect was also significant (Q = 10.66, p = .001). These results suggested that positive EFT had a larger effect (see Table 2).

Table 2.

The effect of EFT-related moderators on DD.

	K	g	Z	p	95% CI	Q	p	Classic fail-safe N
The EFT valence
Positive	34	0.64	12.89	<.001	[0.54, 0.73]			2,342
Not particularly mentioned	16	0.28	2.86	.004	[0.09, 0.47]			58
Total between						10.77	.001**
The context type of EFT
Personally relevant	55	0.50	8.38	<.001	[0.38, 0.62]			3,765
Task-related	10	0.47	5.46	<.001	[0.30, 0.64]			58
Total between						0.09	.764
The longest delay in EFT
⩽180 days	10	0.51	6.58	<.001	[0.36, 0.66]			100
180 < X ⩽ 365 days	21	0.50	6.48	<.001	[0.35, 0.65]			455
>365 days	23	0.62	6.69	<.001	[0.44, 0.81]			976
Total between						1.21	0.547

Note: EFT: episodic future thinking; DD: delay discounting; CI: confidence interval.

p < .01.

The effect of context type of EFT

Both personally relevant EFT (k = 55, g = 0.50, p < .001) and task-related EFT (k = 10, g = 0.47, p < .001) reduced DD. The two context types of EFT showed similar effects on DD (Q = 0.09, p = .764; see Table 2).

The effect of longest delay in EFT

Based on the longest delay, we divided the studies into three groups: no more than 180 days (⩽180 days), between 180 and 365 days (180 < X ⩽ 365 days), and more than 365 days (> 365 days). The results demonstrated that EFT reduced DD irrespective of the length of delay: the longest EFT delay was no more than 180 days (the longest EFT delay was 180 days for all studies in this group; k = 10, g = 0.51, p < .001), between 180 and 365 days (k = 21, g = 0.50, p < .001) and more than 365 days (k = 23, g = 0.62, p < .001), and there was no significant difference among these groups (Q = 1.21, p = .547). We further combined the ⩽180 days group and 180 < X ⩽ 365 days group as one group (k = 31, g = 0.51, p < .001), and compared with the >365 days group, and there was no significant difference (Q = 1.15, p = .284).

Control task–related moderators

The effect of context type of control task

EFT showed medium effect sizes with all types of control tasks (no control task, k = 24, g = 0.40, p = .001; past event, k = 24, g = 0.64, p < .001; other types such as storytelling and routine event, k = 19, g = 0.38, p < .001). There was a significant difference among the three types of control tasks (Q = 7.57, p = .023). Further pairwise comparisons revealed that studies using past events as a control task showed larger effect sizes in reducing DD compared with studies using other types of control tasks (Q = 6.68, p = .010); other comparisons were nonsignificant (see Table 3).

Table 3.

The effect of control task–related moderators on DD.

	K	g	Z	p	95% CI	Q	p	Classic fail-safe N
The context type of control task
Past event	24	0.64	9.57	<.001	[0.51, 0.77]			1,188
No control task	24	0.40	3.28	.001	[0.16, 0.64]			334
Others	19	0.38	4.95	<.001	[0.23, 0.52]			253
Total between						7.57	.023*
The control task valence
Positive	19	0.68	9.42	<.001	[0.53, 0.82]			851
Not particularly mentioned	4	0.55	4.91	<.001	[0.33, 0.78]			24
Total between						0.82	.365
The longest time distance of control task
⩽24 hr	10	0.51	7.96	<.001	[0.38, 0.63]			150
>24 hr	14	0.70	6.98	<.001	[0.50, 0.89]			479
Total between						2.54	.111

Note: DD: delay discounting; CI: confidence interval.

p < .05.

The effect of control task valence

EFT reduced DD in studies with either a positive control task (k = 19, g = 0.68, p < .001) or a control task with valence not particularly mentioned (k = 4, g = 0.55, p < .001). The valence of the control task did not influence the EFT effect on DD (Q = 0.82, p = .365).

The effect of longest time distance of control task

In 24 contrasts, the control task mentioned the time distance, and EFT reduced DD in these studies (g = 0.64, p < .001). We further divided the time limit of control events into those occurring within 24 hr or over 24 hr (⩽24 hr, k = 10, g = 0.51, p < .001; >24 hr, k = 14, g = 0.70, p < .001). These two groups of studies showed similar effects (Q = 2.54, p = .111).

DD task–related moderators

The effect of DD reward type

When the reward was hypothetical, EFT reduced DD (k = 49, g = 0.59, p < .001), and when the reward was potentially real, EFT did not significantly reduce DD (k = 14, g = 0.17, p = .209). Moderator analysis revealed that there was a significant difference between the two groups; when the reward was hypothetical, EFT had a larger effect (Q = 7.83, p = .001) (see Table 4).

Table 4.

The effect of DD task–related moderators on DD.

	K	g	Z	p	95% CI	Q	p	Classic fail-safe N
The DD money type
Hypothetical money	49	0.59	10.82	<.001	[0.48, 0.69]			3,760
Potentially real money	14	0.17	1.26	.209	[–0.10, 0.44]			13
Total between						7.83	.001**
The DD task type
Fixed money	20	0.47	4.89	<.001	[0.28, 0.65]			320
Variable money	43	0.50	7.37	<.001	[0.37, 0.64]			2,303
Total between						0.11	.746
The outcome index of DD
Area under curve	28	0.64	9.18	<.001	[0.50, 0.78]			1,366
K	30	0.37	4.67	<.001	[0.21, 0.52]			653
Total between						6.84	.009**
The longest delay of DD
⩽180 days	13	0.25	1.71	.087	[–0.04, 0.53]			44
180 < X ⩽ 365 days	28	0.45	10.27	<.001	[0.38, 0.54]			787
>365 days	22	0.74	7.06	<.001	[0.54, 0.95]			1,107
Total between						9.26	.010**

Note: DD: delay discounting; CI: confidence interval.

p < .05; **p < .01.

The effect of DD paradigm type

EFT reduced the DD rate when the rewards were fixed in the DD task (k = 20, g = 0.47, p < .001) and when the rewards varied depending on the participants’ responses (k = 43, g = 0.50, p < .001). The results indicated that the moderator effect of the DD paradigm was not significant (Q = 0.11, p = .746).

The effect of outcome index of DD task

Studies using AUC (k = 28, g = 0.64, p < .001) and K (including k value, log k value and ln k value) (k = 30, g = 0.37, p < .001) indices showed a significant effect of EFT in reducing DD. Other indices included the ratio of choosing a larger later reward, the ratio of choosing a smaller sooner reward, and indifference points. Studies using these indices also exhibited a significant effect of EFT in reducing DD (k = 14, g = 0.38, p < .001). Considering the great heterogeneity of other indices, we compared the indices of AUC and K, and the moderator effect was significant (Q = 6.84, p = .009).

The effect of the longest delay in the DD task

EFT did not reduce DD when the longest delay in the DD task was ⩽180 days (k = 13, g = 0.25, p = .087), but EFT reduced DD when the longest delay in the DD task was longer (180 < X ⩽ 365 days, k = 28, g = 0.45, p < .001; >365 days, k = 22, g = 0.74, p < .001). There was a significant difference among these studies (Q = 9.26, p = .010). We further combined the ⩽180 days group and 180 < X ⩽ 365 days group into one group (k = 41, g = 0.38, p < .001), and when we compared it with the >365 days group, there was also a significant difference (Q = 9.24, p = .002).

Participant-related moderators

The effects of age

We divided the studies into two groups: mean age of participants no more than 40 years old (k = 43, g = 0.44, p < .001) and more than 40 years (k = 12, g = 0.62, p < .001), and the moderator effect was not significant (Q = 1.43, p = .231). Furthermore, meta-regression analysis revealed that age was not significantly related to the EFT effect (Z = 1.72, p = .085) (see Table 5).

Table 5.

The effect of information of participants-related moderators on reducing DD.

	K	g	Z	p	95% CI	Q	p	Classic fail-safe N
The age of participants
⩽40 years old	43	0.44	6.48	<.001	[0.31, 0.57]			1,949
>40 years old	12	0.62	4.60	<.001	[0.35, 0.88]			141
Total between						1.43	0.231
The population of participants
Healthy individuals	45	0.45	6.70	<.001	[0.32, 0.58]			1,927
Special populations	18	0.60	6.65	<.001	[0.43, 0.78]			486
Total between						1.79	0.181

Note: DD: delay discounting; CI: confidence interval.

The population of participants

EFT reduced the DD rate in both healthy individuals (k = 45, g = 0.45, p < .001) and in those in special populations (k = 18, g = 0.60, p < .001), and there was no significant heterogeneity among these studies (Q = 1.79, p = .181).

Discussion

In the present meta-analysis, 47 studies including 63 contrasts were included to examine the effects of EFT in reducing DD. The results indicated that EFT reduced DD with a medium effect size (Hedges’ g = 0.52), and publication bias was unlikely to account for this significant finding. Moderator analyses revealed that the EFT valence, context type of the control task, delayed money reward type, outcome index of DD, and longest delay in the DD task were significantly associated with the efficacy of EFT in reducing the DD rate.

Overall effects of EFT in reducing DD

The present meta-analysis revealed that EFT significantly reduced DD. There may be several potential mechanisms. According to construal level theory (Trope & Liberman, 2003), EFT may provide a chance for participants to transfer their representation of abstract future events from a high construal level to a low construal level by imaging future events with concrete details. Imagining and representing rich details may make future outcomes more attractive and increase the possibility of choosing the delayed option (Bulley et al., 2019; Ozdes, 2021).

Alternatively, EFT may strengthen self-regulation to revalue the delayed reward and inhibit impulsive behaviour (Baumeister, 2014; Boyer, 2008; Bulley & Gullo, 2017). These processes are regulated by the prefrontal cortex, which is related to self-control function (Figner et al., 2010). Peters and Büchel (2011) summarised previous studies and suggested that DD was related to three neurocognitive systems, that is, the valuation network, the cognitive control network, and the medial temporal lobe (imagery/prospection) network. EFT was shown to be related to particular brain regions, including the ventromedial prefrontal cortex (vmPFC), medial temporal lobe, and amygdala (Schacter et al., 2017). During EFT, the activation and connectivity between these regions were strengthened, which would have reduced the DD rate (Peters & Büchel, 2011). Specifically, one of the key regions is the vmPFC, which is not only related to EFT but also involved in valuation judgement and cognitive control (Hare et al., 2009; Peters & Buechel, 2009). The vmPFC regulates connections among the three networks; specifically, the vmPFC is involved in the representation of future events during imagination, and the vmPFC regulates the judgement of subjective value for the delayed reward. In addition, the vmPFC inspires self-control ability and self-regulation to help individuals choose delayed options (Benoit et al., 2011; Jenkins & Hsu, 2017). These results supported the idea that EFT may regulate cognitive control to inhibit impulsive choices (Sasse et al., 2017).

EFT-related moderators

Based on the present results, the majority of studies found that EFT significantly reduced DD. However, several factors moderated this effect. One factor was the valence of EFT. Positive EFT significantly reduced DD with a medium effect size (Bulley et al., 2019; Calluso et al., 2019; Zhang et al., 2018). Boyer (2008) suggested that the emotion associated with EFT may act as a “motivational brake,” which means that positive EFT helps individuals counteract the impulsiveness of choosing the immediate option. Specifically, EFT makes people pre-experience future events and generate anticipatory feelings (Suddendorf & Moore, 2011). Vivid EFT may add weight to the value of the delayed reward, which would provide a stronger motivation for goal pursuit (Bulley et al., 2019; Miloyan et al., 2016; Ozdes, 2021; Renner et al., 2019). Anticipation of future events may expand the temporal window, which increases tolerance for delayed rewards, and people would prefer to choose the larger reward, as it brings more good feelings (Snider et al., 2016). Although previous studies showed inconsistent findings on negative EFT and the number of studies involving negative EFT was small (Bulley et al., 2019; Calluso et al., 2019; Liu et al., 2013; Zhang et al., 2018), the results of the moderator analysis provided direct and indirect evidence that neutral or negative EFT would not reduce DD rate, since in these studies that did not mention the valence of EFT, participants may have imagined neutral or negative future events, and these types of EFT showed a significantly smaller or no effect in reducing DD (Bulley & Schacter, 2020). These results suggested that the emotional valence of EFT is an essential factor in reducing DD, and positive EFT is more effective.

Regarding the context type of EFT, both personally related and task-related EFT reduced DD to a similar degree. The time distance of EFT was not related to the efficacy of EFT in reducing DD. One potential reason is that thinking about future events that might happen in a few months could broaden the time horizon to a similar degree to those in a few years; thus, the effect on DD might be the same regardless of how far into the future the participants imagine. From the neural perspective, it might be regardless of whether participants imagine personally relevant events or task-related events and regardless of whether they imagine a relatively distant future or relatively near future, similar brain regions are activated, such as the vmPFC and hippocampus (Bulley & Schacter, 2020; D’Argembeau et al., 2008), and DD is reduced to a similar degree (Benoit et al., 2011; Palombo et al., 2015).

Control task–related moderators

The present meta-analysis demonstrated that the context type of the control task was related to the efficacy of episodic recent thinking in reducing DD. There are several types of control tasks, including no control tasks, episodic recent thinking, and other tasks, such as storytelling or describing routine events (Calluso et al., 2019; Cheng et al., 2012; Hollis-Hansen et al., 2019). Compared with all types of control tasks, EFT showed a significant effect on DD. The results also suggested that the EFT effect on DD was larger when the control task was episodic recent thinking than other types. Episodic recent thinking requires individuals to vividly describe recent events, and this task controls scene construction and self-relevance (Hollis-Hansen et al., 2019); for storytelling tasks (belonging to the “other type” of control task), it controls for an individual’s verbal ability and scene construction (Bulley & Gullo, 2017). However, the other types of control tasks included different tasks, such as storytelling (Bulley & Gullo, 2017), recalling travel blogs (Daniel et al., 2013b), and episodic thinking through recalling mobile application games (Hollis-Hansen et al., 2019). Considering the number of contrasts, they cannot be divided for further analysis. However, these results suggested the importance of standardising the control task when examining the effect of EFT in reducing DD in future studies (Hollis-Hansen et al., 2019).

Regarding the events recalled in the control task, the retrospective time period ranged from the past 24 hr to 3 years. The present results demonstrated that the time distance of the control task was not associated with the EFT effect (O’Donnell et al., 2018). One of the reasons might be that in most studies, the participants were required to recall events within several days in the control task. Moreover, the valence of the control task did not influence the EFT effect either, which suggests that only the emotional valence of EFT is critical in reducing DD.

DD-related moderators

DD has been investigated with various tasks (Mok et al., 2020; Scholten et al., 2019). The present study examined several characteristics of the DD task. We found that EFT reduced DD to a larger degree in studies using hypothetical money rewards than in studies using potential real money rewards, which is inconsistent with previous studies suggesting that both hypothetical and potential real rewards yielded similar EFT effects in reducing DD (Lawyer et al., 2011; Madden et al., 2003, 2004). One possible reason is that when using potential real reward, the amount of reward was relatively small, while when using hypothetical reward, the amount of reward was larger, which is consistent with previous findings that reward amount has an effect on DD (Białaszek & Ostaszewski, 2012; Green & Myerson, 2004). One factor that needs to be mentioned is that the number of the two types of studies showed a large difference; almost 75% of the studies used hypothetical rewards, and future studies need to focus more on potential real money rewards to compare the effects of EFT on DD.

The DD paradigm can be divided into two types: the fixed task in which the amount of reward is fixed, not influenced by participants’ choices, and the varied task in which the amount of reward changed according to participants’ responses. Both paradigms showed medium and similar sized effects of EFT on DD (Bromberg et al., 2017; Jia et al., 2020). These results suggested that the effect of EFT was stable across DD paradigms.

AUC- and K-related indices are two main types of indices of DD. Consistent with Bromberg et al. (2017) and Hu, Kleinschmidt, et al. (2017), the present results indicated that EFT reduced DD using both AUC and K as indices, suggesting that the effects of EFT on DD were stable across indices. In addition, we found that EFT had a larger effect on AUC-related than K-related indices; however, there was no single study that directly compared the effect of EFT in reducing DD with different indices (Hamilton et al., 2015). These two indices are based on different mathematical models and are usually derived from different measurement methods (Odum, 2011). Further studies are needed to examine the meaning and differences between AUC- and K-related indices.

The delay time used in the DD task varied widely from 7 days to 25 years (Cheng et al., 2012; Stein et al., 2017), and the present results indicated that the effect of EFT on DD was stronger when the delay time was longer. This might be because a longer delay was associated with a higher DD. EFT could broaden the temporal window and shorten the subjective duration, and this effect was stronger during longer delays (Jia et al., 2020; Snider et al., 2016; Sze et al., 2017).

Participant-related moderators

Very few studies have compared the effectiveness of EFT in reducing DD across age groups. Sasse et al. (2017) indicated that EFT did not reduce DD in older groups, while Mok et al. (2020) indicated that the effect of EFT also reduced DD in older groups. The present results demonstrated that age did not moderate the effect of EFT in reducing DD. However, only a few studies have examined this effect in participants more than 60 years (Hu, Uhle, et al., 2017) or in adolescents (Bromberg et al., 2015; Daniel et al., 2015), and further studies in older adults and adolescents are needed. Moreover, we found that EFT reduced DD in both healthy individuals and special populations, consistent with previous studies which suggested that the EFT effect is widespread, involving different populations and ranging from financial rewards to maladaptive behaviours (Dassen et al., 2016).

Limitations and implications

There are several limitations in the present study. First, the underlying mechanisms of EFT on DD still need to be examined. Second, some studies considered other outcomes, such as cigarette smoking, energy intake, and alcohol demand (Daniel et al., 2015; Stein et al., 2016), but the present study did not analyse these outcomes due to the limited number of studies. Third, studies have shown that episodic past thinking could also reduce DD (Lempert et al., 2017); however, the number of studies was not enough to be included as a subgroup analysis. Future studies need to examine whether this effect is stable and whether there are differential effects of EFT and episodic past thinking in reducing DD. Fourth, most studies examined the effect of EFT on DD for gains, and whether EFT could reduce DD for losses needs further investigation. Fifth, for several moderator analyses, the number of studies included in each subgroup was imbalanced, and future studies are needed in several subgroups. For example, more studies are needed in clinical patients, since an increased DD rate is a transdiagnostic process in psychiatric disorders (Amlung et al., 2019).

Notwithstanding the above limitations, there are some implications of this study. First, EFT is an effective manipulation to reduce DD. One future direction is to apply this method in clinical populations with impulsive behaviours. Second, the present results suggest that the standardisation of the control task is necessary. Third, given that positive EFT had a larger effect in reducing DD, when applying EFT, requiring participants to imagine future positive events is an optimal option.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the National Science Foundation of China (32071062).

ORCID iD

Ya Wang

References

Amlung

Marsden

Holshausen

Morris

Patel

Vedelago

. . .McCabe

R. E.

(2019). Delay discounting as a transdiagnostic process in psychiatric disorders: A meta-analysis. JAMA Psychiatry, 76(11), 1176–1186. https://doi.org/10.1001/jamapsychiatry.2019.2102

Arnold

K. M.

McDermott

K. B.

Szpunar

K. K.

(2011). Imagining the near and far future: The role of location familiarity. Memory & Cognition, 39(6), 954–967. https://doi.org/10.3758/s13421-011-0076-1

Athamneh

L. N.

Stein

M. D.

Lin

E. H.

Stein

J. S.

Mellis

A. M.

Gatchalian

K. M.

. . .Bickel

W. K.

(2020). Setting a goal could help you control: Comparing the effect of health goal versus general episodic future thinking on health behaviors among cigarette smokers and obese individuals. Experimental and Clinical Psychopharmacology, 29(1), 59–72. https://doi.org/10.1037/pha0000351

Baumeister

R. F.

(2014). Self-regulation, ego depletion, and inhibition. Neuropsychologia, 65, 313–319. https://doi.org/10.1016/j.neuropsychologia.2014.08.012

Benoit

R. G.

Gilbert

S. J.

Burgess

P. W.

(2011). A neural mechanism mediating the impact of episodic prospection on farsighted decisions. Journal of Neuroscience, 31(18), 6771–6779. https://doi.org/10.1523/jneurosci.6559-10.2011

Białaszek

Ostaszewski

(2012). Discounting of sequences of delayed rewards of different amounts. Behavioural Processes, 89(1), 39–43. https://doi.org/org/10.1016/j.beproc.2011.10.014

Bobova

Finn

P. R.

Rickert

M. E.

Lucas

(2009). Disinhibitory psychopathology and delay discounting in alcohol dependence: Personality and cognitive correlates. Experimental and Clinical Psychopharmacology, 17(1), 51–61. https://doi.org/10.1037/a0014503

Borenstein

Hedges

L. V.

Higgins

J. P.

Rothstein

H. R.

(2005). Comprehensive meta-analysis (Version 2). Englewood Cliffs.

Boyer

(2008). Evolutionary economics of mental time travel? Trends in Cognitive Sciences, 12(6), 219–224. https://doi.org/org/10.1016/j.tics.2008.03.003

10.

Brocas

Carrillo

J. D.

(2018). A neuroeconomic theory of mental time travel. Frontiers in Neuroscience, 12, Article 658. https://doi.org/10.3389/fnins.2018.00658

11.

Bromberg

Lobatcheva

Peters

(2017). Episodic future thinking reduces temporal discounting in healthy adolescents. PLOS ONE, 12(11), Article e0188079. https://doi.org/10.1371/journal.pone.0188079

12.

Bromberg

Wiehler

Peters

(2015). Episodic future thinking is related to impulsive decision making in healthy adolescents. Child Development, 86(5), 1458–1468. https://doi.org/10.1111/cdev.12390

13.

Bulley

Gullo

M. J.

(2017). The influence of episodic foresight on delay discounting and demand for alcohol. Addictive Behaviors, 66, 1–6. https://doi.org/org/10.1016/j.addbeh.2016.11.003

14.

Bulley

Miloyan

Pepper

G. V.

Gullo

M. J.

Henry

J. D.

Suddendorf

(2019). Cuing both positive and negative episodic foresight reduces delay discounting but does not affect risk-taking. Quarterly Journal of Experimental Psychology, 72(8), 1998–2017. https://doi.org/10.1177/1747021818819777

15.

Bulley

Schacter

D. L.

(2020). Deliberating trade-offs with the future. Nature Human Behaviour, 4(3), 238–247. https://doi.org/10.1038/s41562-020-0834-9

16.

Calluso

Tosoni

Cannito

Committeri

(2019). Concreteness and emotional valence of episodic future thinking (EFT) independently affect the dynamics of intertemporal decisions. PLOS ONE, 14(5), Article e0217224. https://doi.org/10.1371/journal.pone.0217224

17.

Carr

K. A.

Hollis-Hansen

Austin

Epstein

L. H.

(2021). Written or drawn episodic future thinking cues improves delay discounting in adults. Learning and Motivation, 74, 101727. https://doi.org/10.1016/j.lmot.2021.101727

18.

Cheng

Y. Y.

Shein

P. P.

Chiou

W. B.

(2012). Escaping the impulse to immediate gratification: The prospect concept promotes a future-oriented mindset, prompting an inclination towards delayed gratification. British Journal of Psychology, 103(1), 129–141. https://doi.org/10.1111/j.2044-8295.2011.02067.x

19.

Chiou

W. B.

W. H.

(2017). Episodic future thinking involving the nonsmoking self can induce lower discounting and cigarette consumption. Journal of Studies on Alcohol and Drugs, 78(1), 106–112. https://doi.org/10.15288/jsad.2017.78.106

20.

Ciaramelli

De Luca

Kwan

Mok

Bianconi

Knyagnytska

. . .Rosenbaum

R. S.

(2021). The role of ventromedial prefrontal cortex in reward valuation and future thinking during intertemporal choice. eLife, 10, e67387. https://doi.org/10.7554/eLife.67387

21.

Ciaramelli

Sellitto

Tosarelli

di Pellegrino

(2019). Imagining Events Alternative to the Present Can Attenuate Delay Discounting. Frontiers in Behavioral Neuroscience, 13, 269. https://doi.org/10.3389/fnbeh.2019.00269

22.

Craft

W. H.

Tegge

A. N.

Bickel

W. K.

(2020). Episodic future thinking reduces chronic pain severity: A proof of concept study. Drug and Alcohol Dependence, 215, 108250.https://doi.org/10.1016/j.drugalcdep.2020.108250

23.

Daniel

T. O.

Said

Stanton

C. M.

Epstein

L. H.

(2015). Episodic future thinking reduces delay discounting and energy intake in children. Eating Behaviors, 18, 20–24. https://doi.org/org/10.1016/j.eatbeh.2015.03.006

24.

Daniel

T. O.

Sawyer

Dong

Y. L.

Bickel

W. K.

Epstein

L. H.

(2016). Remembering versus imagining: When does episodic retrospection and episodic prospection aid decision making? Journal of Applied Research in Memory and Cognition, 5(3), 352–358. https://doi.org/org/10.1016/j.jarmac.2016.06.005

25.

Daniel

T. O.

Stanton

C. M.

Epstein

L. H.

(2013a). The future is now: Comparing the effect of episodic future thinking on impulsivity in lean and obese individuals. Appetite, 71, 120–125. https://doi.org/10.1016/j.appet.2013.07.010

26.

Daniel

T. O.

Stanton

C. M.

Epstein

L. H.

(2013b). The future is now: Reducing impulsivity and energy intake using episodic future thinking. Psychological Science, 24(11), 2339–2342. https://doi.org/10.1177/0956797613488780

27.

D’Argembeau

Xue

Z. L.

Van der Linden

Bechara

(2008). Neural correlates of envisioning emotional events in the near and far future. NeuroImage, 40(1), 398–407. https://doi.org/org/10.1016/j.neuroimage.2007.11.025

28.

Dassen

F. C. M.

Jansen

Nederkoorn

Houben

(2016). Focus on the future: Episodic future thinking reduces discount rate and snacking. Appetite, 96, 327–332. https://doi.org/org/10.1016/j.appet.2015.09.032

29.

Figner

Knoch

Johnson

E. J.

Krosch

A. R.

Lisanby

S. H.

Fehr

Weber

E. U.

(2010). Lateral prefrontal cortex and self-control in intertemporal choice. Nature Neuroscience, 13(5), 538–539. https://doi.org/10.1038/nn.2516

30.

Frederick

Loewenstein

O’Donoghue

(2002). Time discounting and time preference: A critical review. Journal of Economic Literature, 40(2), 351–401. https://doi.org/10.1257/002205102320161311

31.

Giles

E. L.

Robalino

McColl

Sniehotta

F. F.

Adams

(2014). The effectiveness of financial incentives for health behaviour change: Systematic review and meta-analysis. PLOS ONE, 9(3), Article e90347. https://doi.org/10.1371/journal.pone.0090347

32.

Green

Myerson

(2004). A discounting framework for choice with delayed and probabilistic rewards. Psychological Bulletin, 130(5), 769–792. https://doi.org/10.1037/0033-2909.130.5.769

33.

Hamilton

K. R.

Mitchell

M. R.

Wing

V. C.

Balodis

I. M.

Bickel

W. K.

Fillmore

. . .Moeller

F. G.

(2015). Choice impulsivity: Definitions, measurement issues, and clinical implications. Personality Disorders: Theory Research and Treatment, 6(2), 182–198. https://doi.org/10.1037/per0000099

34.

Hare

T. A.

Camerer

C. F.

Rangel

(2009). Self-control in decision-making involves modulation of the vmPFC valuation system. Science, 324(5927), 646–648. https://doi.org/10.1126/science.1168450

35.

Hedges

L. V.

Vevea

J. L.

(1998). Fixed- and random-effects models in meta-analysis. Psychological Methods, 3(4), 486–504. https://doi.org/10.1037/1082-989x.3.4.486

36.

Hendrickson

K. L.

Rasmussen

E. B.

(2017). Mindful eating reduces impulsive food choice in adolescents and adults. Health Psychology, 36(3), 226–235. https://doi.org/10.1037/hea0000440

37.

Hollis-Hansen

O’Donnell

S. E.

Seidman

J. S.

Brande

S. J.

Epstein

L. H.

(2019). Improvements in episodic future thinking methodology: Establishing a standardized episodic thinking control. PLOS ONE, 14(3), Article e0214397. https://doi.org/10.1371/journal.pone.0214397

38.

X. C.

Kleinschmidt

Martin

J. A.

Han

Thelen

Meiberth

. . .Weber

(2017). A reduction in delay discounting by using episodic future imagination and the association with episodic memory capacity. Frontiers in Human Neuroscience, 10, Article 663. https://doi.org/10.3389/fnhum.2016.00663

39.

X. C.

Uhle

Fliessbach

Wagner

Han

Weber

Jessen

(2017). Reduced future-oriented decision making in individuals with subjective cognitive decline: A functional MRI study. Alzheimer’s & Dementia, 6, 222–231. https://doi.org/10.1016/j.dadm.2017.02.005

40.

Jenkins

A. C.

Hsu

(2017). Dissociable contributions of imagination and willpower to the malleability of human patience. Psychological Science, 28(7), 894–906. https://doi.org/10.1177/0956797617698133

41.

Jia

L. X.

Liu

Cui

J. F.

Ding

Q. Y.

J. Y.

Liu

L. L.

. . .Wang

(2020). Future thinking is related to lower delay discounting than recent thinking, regardless of the magnitude of the reward, in individuals with schizotypy. Australian Psychologist, 55, 572–581. https://doi.org/org/10.1111/ap.12460

42.

Kim

Schnall

White

M. P.

(2013). Similar psychological distance reduces temporal discounting. Personality and Social Psychology Bulletin, 39(8), 1005–1016. https://doi.org/10.1177/0146167213488214

43.

Lawyer

S. R.

Schoepflin

Green

Jenks

(2011). Discounting of hypothetical and potentially real outcomes in nicotine-dependent and nondependent samples. Experimental and Clinical Psychopharmacology, 19(4), 263–274. https://doi.org/10.1037/a0024141

44.

Lempert

K. M.

Phelps

E. A.

(2016). The malleability of intertemporal choice. Trends in Cognitive Sciences, 20(1), 64–74. https://doi.org/10.1016/j.tics.2015.09.005

45.

Lempert

K. M.

Speer

M. E.

Delgado

M. R.

Phelps

E. A.

(2017). Positive autobiographical memory retrieval reduces temporal discounting. Social Cognitive and Affective Neuroscience, 12(10), 1584–1593. https://doi.org/10.1093/scan/nsx086

46.

Lempert

K. M.

Steinglass

J. E.

Pinto

Kable

J. W.

Simpson

H. B.

(2019). Can delay discounting deliver on the promise of RDoC? Psychological Medicine, 49(2), 190–199. https://doi.org/10.1017/S0033291718001770

47.

Lin

Epstein

L. H.

(2014). Living in the moment: Effects of time perspective and emotional valence of episodic thinking on delay discounting. Behavioral Neuroscience, 128(1), 12–19. https://doi.org/10.1037/a0035705

48.

Liu

Feng

T. Y.

Chen

(2013). The value of emotion: How does episodic prospection modulate delay discounting? PLOS ONE, 8(11), Article e81717. https://doi.org/10.1371/journal.pone.0081717

49.

Logue

A. W.

(1988). Research on self-control: An integrating framework. Behavioral and Brain Sciences, 11(4), 665–678. https://doi.org/10.1017/s0140525x00053978

50.

MacKillop

Amlung

M. T.

Few

L. R.

Ray

L. A.

Sweet

L. H.

Munafo

M. R.

(2011). Delayed reward discounting and addictive behavior: A meta-analysis. Psychopharmacology, 216(3), 305–321. https://doi.org/10.1007/s00213-011-2229-0

51.

Madden

G. J.

Begotka

A. M.

Raiff

B. R.

Kastern

L. L.

(2003). Delay discounting of real and hypothetical rewards. Experimental and Clinical Psychopharmacology, 11(2), 139–145. https://doi.org/10.1037/1064-1297.11.2.139

52.

Madden

G. J.

Raiff

B. R.

Lagorio

C. H.

Begotka

A. M.

Mueller

A. M.

Hehli

D. J.

Wegener

A. A.

(2004). Delay discounting of potentially real and hypothetical rewards: II. Between- and within-subject comparisons. Experimental and Clinical Psychopharmacology, 12(4), 251–261. https://doi.org/10.1037/1064-1297.12.4.251

53.

Mansouri

T. H.

Crandall

A. K.

Temple

J. L.

(2021). The effect of repeated episodic future thinking on the relative reinforcing value of snack food. Journal of Health Psychology, 26(13), 2402–2413. https://doi.org/10.1177/1359105320914060

54.

Miloyan

Bulley

Suddendorf

(2016). Episodic foresight and anxiety: Proximate and ultimate perspectives. British Journal of Clinical Psychology, 55(1), 4–22. https://doi.org/10.1111/bjc.12080

55.

Moher

Liberati

Tetzlaff

Altman

D. G.

(2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. BMJ, 339, Article b2535. https://doi.org/10.1136/bmj.b2535

56.

Mok

J. N. Y.

Kwan

Green

Myerson

Craver

C. F.

Rosenbaum

R. S.

(2020). Is it time? Episodic imagining and the discounting of delayed and probabilistic rewards in young and older adults. Cognition, 199, Article 104222. https://doi.org/org/10.1016/j.cognition.2020.104222

57.

Morrison

K. L.

Madden

G. J.

Odum

A. L.

Twohig

M. P.

(2014). Altering impulsive decision making with an acceptance-based procedure. Behavior Therapy, 45(5), 630–639. https://doi.org/10.1016/j.beth.2014.01.001

58.

O’Donnell

Daniel

T. O.

Koroschetz

Kilanowski

Otminski

Bickel

W. K.

Epstein

L. H.

(2019). Do process simulations during episodic future thinking enhance the reduction of delay discounting for middle income participants and those living in poverty? Journal of Behavioral Decision Making, 32(3), 231–240. https://doi.org/10.1002/bdm.2108

59.

O’Donnell

Oluyomi

D. T.

Epstein

L. H.

(2017). Does goal relevant episodic future thinking amplify the effect on delay discounting? Consciousness and Cognition, 51, 10–16. https://doi.org/10.1016/j.concog.2017.02.014

60.

O’Donnell

Hollis-Hansen

Epstein

L. H.

(2018). Mix and match: An investigation into whether episodic future thinking cues need to match discounting delays in order to be effective. Behavioral Sciences, 9(1), Article 1. https://doi.org/10.3390/bs9010001

61.

Odum

A. L.

(2011). Delay discounting: I’m a k, you’re a k. Journal of the Experimental Analysis of Behavior, 96(3), 427–439. https://doi.org/10.1901/jeab.2011.96-423

62.

Ozdes

(2021). How did I do it then? How will I do it later? A theoretical review of the impact of mental time travel on decision-making processes. New Ideas in Psychology, 62, Article 100869. https://doi.org/10.1016/j.newideapsych.2021.100869

63.

Palombo

D. J.

Keane

M. M.

Verfaellie

(2015). The medial temporal lobes are critical for reward-based decision making under conditions that promote episodic future thinking. Hippocampus, 25(3), 345–353. https://doi.org/10.1002/hipo.22376

64.

Palombo

D. J.

Keane

M. M.

Verfaellie

(2016). Using future thinking to reduce temporal discounting: Under what circumstances are the medial temporal lobes critical? Neuropsychologia, 89, 437–444. https://doi.org/org/10.1016/j.neuropsychologia.2016.07.002

65.

Parthasarathi

McConnell

M. H.

Luery

Kable

J. W.

(2017). The vivid present: Visualization abilities are associated with steep discounting of future rewards. Frontiers in Psychology, 8, Article 289. https://doi.org/10.3389/fpsyg.2017.00289

66.

Patel

Amlung

(2020). Acute and extended exposure to episodic future thinking in a treatment seeking addiction sample: A pilot study. Journal of Substance Abuse Treatment, 116, 108046. https://doi.org/10.1016/j.jsat.2020.108046

67.

Peters

Büchel

(2010). Episodic future thinking reduces reward delay discounting through an enhancement of prefrontal-mediotemporal interactions. Neuron, 66(1), 138–148. https://doi.org/org/10.1016/j.neuron.2010.03.026

68.

Peters

Büchel

(2011). The neural mechanisms of inter-temporal decision-making: Understanding variability. Trends in Cognitive Sciences, 15(5), 227–239. https://doi.org/10.1016/j.tics.2011.03.002

69.

Peters

Buechel

(2009). Overlapping and distinct neural systems code for subjective value during intertemporal and risky decision making. Journal of Neuroscience, 29(50), 15727–15734. https://doi.org/10.1523/jneurosci.3489-09.2009

70.

Renner

Murphy

F. C.

J. L.

Manly

Holmes

E. A.

(2019). Mental imagery as a “motivational amplifier” to promote activities. Behaviour Research and Therapy, 114, 51–59. https://doi.org/org/10.1016/j.brat.2019.02.002

71.

Reynolds

(2006). A review of delay-discounting research with humans: Relations to drug use and gambling. Behavioural Pharmacology, 17(8), 651–667. https://doi.org/10.1097/FBP.0b013e3280115f99

72.

Rung

J. M.

Madden

G. J.

(2018). Demand characteristics in episodic future thinking: Delay discounting and healthy eating. Experimental and Clinical Psychopharmacology, 26(1), 77–84. https://doi.org/10.1037/pha0000171

73.

Rung

J. M.

Madden

G. J.

(2019). Demand characteristics in episodic future thinking II: The role of cues and cue content in changing delay discounting. Experimental and Clinical Psychopharmacology, 27(5), 482–495. https://doi.org/10.1037/pha0000260

74.

Rung

J. M.

Madden

G. J.

(2018). Experimental reductions of delay discounting and impulsive choice: A systematic review and meta-analysis. Journal of Experimental Psychology-General, 147(9), 1349–1381. https://doi.org/10.1037/xge0000462

75.

Sasse

L. K.

Peters

Brassen

(2017). Cognitive control modulates effects of episodic simulation on delay discounting in aging. Frontiers in Aging Neuroscience, 9, Article 58. https://doi.org/10.3389/fnagi.2017.00058

76.

Sasse

L. K.

Peters

Buechel

Brassen

(2015). Effects of prospective thinking on intertemporal choice: The role of familiarity. Human Brain Mapping, 36(10), 4210–4221. https://doi.org/10.1002/hbm.22912

77.

Schacter

D. L.

Addis

D. R.

Buckner

R. L.

(2007). Remembering the past to imagine the future: The prospective brain. Nature Reviews Neuroscience, 8(9), 657–661. https://doi.org/10.1038/nrn2213

78.

Schacter

D. L.

Benoit

R. G.

Szpunar

K. K.

(2017). Episodic future thinking: Mechanisms and functions. Current Opinion in Behavioral Sciences, 17, 41–50. https://doi.org/10.1016/j.cobeha.2017.06.002

79.

Schacter

D. L.

Madore

K. P.

(2016). Remembering the past and imagining the future: Identifying and enhancing the contribution of episodic memory. Memory Studies, 9(3), 245–255. https://doi.org/10.1177/1750698016645230

80.

Scholten

Scheres

de Water

Graf

Granic

Luijten

(2019). Behavioral trainings and manipulations to reduce delay discounting: A systematic review. Psychonomic Bulletin & Review, 26(6), 1803–1849. https://doi.org/10.3758/s13423-019-01629-2

81.

Sellitto

Ciaramelli

di Pellegrino

(2011). The neurobiology of intertemporal choice: Insight from imaging and lesion studies. Reviews in the Neurosciences, 22(5), 565–574. https://doi.org/10.1515/rns.2011.046

82.

Snider

S. E.

LaConte

S. M.

Bickel

W. K.

(2016). Episodic future thinking: Expansion of the temporal window in individuals with alcohol dependence. Alcoholism: Clinical and Experimental Research, 40(7), 1558–1566. https://doi.org/10.1111/acer.13112

83.

Sofis

M. J.

Lemley

S. M.

Lee

D. C.

Budney

A. J.

(2020). A web-based episodic specificity and future thinking session modulates delay discounting in cannabis users. Psychology of Addictive Behaviors, 34, 532–540. https://doi.org/10.1037/adb0000557

84.

Stanger

Budney

A. J.

Bickel

W. K.

(2013). A developmental perspective on neuroeconomic mechanisms of contingency management. Psychology of Addictive Behaviors, 27(2), 403–415. https://doi.org/10.1037/a0028748

85.

Stein

J. S.

Sze

Y. Y.

Athamneh

Koffarnus

M. N.

Epstein

L. H.

Bickel

W. K.

(2017). Think fast: Rapid assessment of the effects of episodic future thinking on delay discounting in overweight/obese participants. Journal of Behavioral Medicine, 40(5), 832–838. https://doi.org/10.1007/s10865-017-9857-8

86.

Stein

J. S.

Wilson

A. G.

Koffarnus

M. N.

Daniel

T. O.

Epstein

L. H.

Bickel

W. K.

(2016). Unstuck in time: Episodic future thinking reduces delay discounting and cigarette smoking. Psychopharmacology, 233(21–22), 3771–3778. https://doi.org/10.1007/s00213-016-4410-y

87.

Suddendorf

Moore

(2011). Introduction to the special issue: The development of episodic foresight. Cognitive Development, 26(4), 295–298. https://doi.org/10.1016/j.cogdev.2011.09.001

88.

Sze

Y. Y.

Stein

J. S.

Bickel

W. K.

Paluch

R. A.

Epstein

L. H.

(2017). Bleak present, bright future: Online episodic future thinking, scarcity, delay discounting, and food demand. Clinical Psychological Science, 5(4), 683–697. https://doi.org/10.1177/2167702617696511

89.

Trope

Liberman

(2003). Temporal construal. Psycho-logical Review, 110(3), 403–421. https://doi.org/10.1037/0033-295x.110.3.403

90.

Voss

A. T.

Jorgensen

M. K.

Murphy

J. G.

(2021). Episodic future thinking as a brief alcohol intervention for heavy drinking college students: A pilot feasibility study. Experimental and Clinical Psychopharmacology. https://doi.org/10.1037/pha0000451

91.

Wiehler

Petzschner

F. H.

Stephan

K. E.

Peters

(2017). Episodic Tags Enhance Striatal Valuation Signals during Temporal Discounting in pathological Gamblers. eNeuro, 4(3), ENEURO.0159-0117.2017. https://doi.org/10.1523/ENEURO.0159-17.2017

92.

W. H.

Cheng

Chiou

W. B.

(2017). Episodic future thinking about the ideal self induces lower discounting, leading to a decreased tendency toward cheating. Frontiers in Psychology, 8, Article 287. https://doi.org/10.3389/fpsyg.2017.00287

93.

Yao

Y. W.

Chen

P. R.

C. S. R.

Hare

T. A.

Zhang

J. T.

. . .Fang

X. Y.

(2017). Combined reality therapy and mindfulness meditation decrease intertemporal decisional impulsivity in young adults with Internet gaming disorder. Computers in Human Behavior, 68, 210–216. https://doi.org/10.1016/j.chb.2016.11.038

94.

Stuppy-Sullivan

Pickover

Landes

R. D.

(2017). Impact of construal level manipulations on delay discounting. PLoS One, 12(5), e0177240. https://doi.org/10.1371/journal.pone.0177240

95.

Zhang

S. M.

Peng

Qin

L. L.

Suo

Feng

T. Y.

(2018). Prospective emotion enables episodic prospection to shift time preference. British Journal of Psychology, 109(3), 487–499. https://doi.org/10.1111/bjop.12284