Steep temporal reward discounting in ADHD-Combined type: Acting upon feelings

Abstract

Difficulty waiting plays a primary role in symptoms of attention-deficit/hyperactivity disorder (ADHD), in particular, impulsivity. Current theories suggest that relatively strong preferences for small immediate rewards as observed in ADHD-Combined type are the result of delay-related negative feelings. However, the measurement of difficulty waiting is typically limited to objective choices between small immediate and large delayed rewards. This study aimed at extending the measurement of difficulty waiting in ADHD-Combined type with ratings about subjective feelings. Children and adolescents (ages 6-17) with ADHD-Combined type (n=25), ADHD-Inattentive type (n=20) and matched typically developing participants (n=37) performed temporal reward discounting tasks, and completed a Visual Analogue Scale of subjectively experienced ease/difficulty waiting. Although those with ADHD-Combined type demonstrated relatively steep temporal reward discounting, as reported elsewhere (Scheres et al., 2010), there were no group differences for subjectively experienced ease/difficulty waiting. Additionally, correlations between subjective and objective measures of difficulty waiting were significantly higher in the ADHD-Combined type group than in the control group. These findings suggest that (a) those with ADHD-Combined type do not choose impulsively because they have more negative feelings about waiting than controls; (b) choices in the ADHD-Combined type group are more in accordance with/driven by their feelings than choices made by participants in the control group.

Introduction

Attention-deficit/hyperactivity disorder (ADHD) is one of the most common psychiatric conditions with onset in childhood (American Psychiatric Association, 2000). It is characterized by impairing levels of inattention and/or hyperactivity–impulsivity. Decades of research have focused on identifying causal mechanisms associated with these symptom domains. Influential theoretical models attempt to explain symptoms as presented in those with ADHD-Combined type (ADHD-C), and not in those with ADHD-Inattentive type (ADHD-I) (e.g., Barkley, 1997, Quay, 1997, Sonuga-Barke, 2002, Sonuga-Barke, 2003, Sonuga-Barke, 2005, Sagvolden et al., 2005, Johansen et al., 2009). These models focus on underlying mechanisms that are closely, but not exclusively, related to the symptom domain of impulsivity, such as executive disinhibition (Barkley), motivational disinhibition (Sonuga-Barke), and a steep delay-of-reinforcement gradient (Sagvolden). Likewise, a large number of experimental studies examining cognitive and motivational processes in ADHD include participants with ADHD-C only, while fewer researchers have included those with ADHD-I, or have compared the various subtypes (see for reviews Nigg, 2001, Adams et al., 2008).

It is worth noting that the DSM-IV lists only three symptoms of impulsivity, while inattention is represented by nine items, and hyperactivity by six. The symptom domain of impulsivity is represented in the DSM-IV by the following three behavior descriptions: “often has difficulty awaiting turn”, “often interrupts or intrudes on others”, “often blurts out answers before questions have been completed”. Here, we will focus on difficulty waiting as one important expression of impulsivity. Based on general theories of impulsivity (Monterosso and Ainslie, 1999, Cardinal et al., 2004, Kalenscher et al., 2006, Winstanley et al., 2006, Dalley et al., 2008), specific ADHD theories (Barkley, 1997, Sonuga-Barke, 2002, Sonuga-Barke, 2003, Sonuga-Barke, 2005, Tripp and Wickens, 2008, Johansen et al., 2009, Luman et al., 2010, Sagvolden et al., 2005), and clinical descriptions of ADHD (American Psychiatric Association, 2000), difficulty waiting seems to play an important role in impulsivity in general, and in ADHD-related impulsivity specifically.

In research studies of ADHD, difficulty waiting is typically assessed with objective measures of choice preference when presented with two options: a small immediate reward and a large delayed reward (but see Antrop et al., 2005, for an interesting exception). Earlier studies made use of single-choice paradigms such as the choice delay task (CDT), in which children were presented with repeated choices between one point that they could receive right away, or two points that they could receive after 30 s (each point being worth 5 cents, with the money being paid to the participants at the end of the block or the entire task). Sonuga-Barke et al. (1992) were the first to show with this task that 6- to 8-year-old hyperactive–impulsive boys demonstrated unusually strong preferences for small immediate rewards. Overall, more recent studies with single-choice paradigms have replicated this finding (Scheres et al., 2010, Schweitzer and Sulzer-Azaroff, 1995, Kuntsi et al., 2001, Solanto et al., 2001, Tripp and Alsop, 2001, Antrop et al., 2006, Bitsakou et al., 2009, Marco et al., 2009, Vloet et al., 2010; but see Solanto et al., 2007; for a review, see Luman et al., 2005). Most recently, studies of ADHD have employed temporal reward discounting tasks (TD tasks) in which the delay preceding the large reward as well as the magnitude of the immediate reward are varied (Barkley et al., 2001, Scheres et al., 2006, Plichta et al., 2009, Shiels et al., 2009, Paloyelis et al., 2010, Wilson et al., 2011, Demurie et al., 2012). Four out of five studies using hypothetical tasks supported the notion that those with ADHD have relatively strong preferences for small immediate rewards, expressed as relatively steep temporal reward discounting (Barkley et al., 2001, Paloyelis et al., 2010, Wilson et al., 2011, Demurie et al., 2012; but not Plichta et al., 2009). One out of three studies using real TD tasks supported this notion too (Scheres et al., 2010; but not Scheres et al., 2006, Paloyelis et al., 2010). Of note is that these findings may depend on contextual factors such as magnitude of the large reward (Barkley et al., 2001, Scheres et al., 2010) and/or type of task (real vs. hypothetical; see for a direct comparison and discussion Scheres et al., 2008), may be specifically related to the symptom domain of hyperactivity/impulsivity (Thorell, 2007, Scheres et al., 2008, Scheres et al., 2010, Scheres and Hamaker, 2010; but see Paloyelis et al., 2009, Wilson et al., 2011), and may be partially mediated by intelligence levels (Wilson et al., 2011).

Relatively strong preferences for small immediate rewards as observed in ADHD may be interpreted as being a result of negative feelings associated with waiting (Sonuga-Barke, 2002, Sonuga-Barke, 2003, Sonuga-Barke, 2005). However, alternative interpretations are available too, and it may be challenging to filter out the unique contribution of delay-related negative feelings to the choices made in these paradigms: relatively strong preferences for small immediate rewards may also be explained by a hypersensitivity to the immediacy of the small reward (Tripp and Alsop, 2001), or by an insensitivity to reward magnitude or maximizing one's total winnings (see Scheres et al., 2010). Some efforts have been made to tease apart these various explanations: for example, varying the magnitude of the delayed reward and of the total maximum gain (e.g., Barkley et al., 2001, Scheres et al., 2010) can shed light on the potential role of insensitivity to reward maximization in strong preferences for small immediate rewards. Additionally, the use of post-immediate-reward delays in order to control for inter-choice interval and overall task duration, in addition to a condition in which no post-immediate-reward delays are used permit making a distinction between sensitivity to reward immediacy and delay aversion as two explanations for relatively strong preferences for small immediate rewards (Sonuga-Barke et al., 1992, Scheres et al., 2006, Marco et al., 2009, Paloyelis et al., 2009).

However, although the use of post-reward delays makes good sense theoretically, and is common practice in animal studies (see Logue, 1988), it is of limited use in human research, especially when money is used as the reward. For example, previous research has shown that TD tasks with post-reward delays often result in ceiling effects: an overwhelming majority of participants choose the large delayed reward on most trials when task duration is kept constant, especially when they are 12 years of age or older (e.g., Logue et al., 1986, Jackson and Hackenberg, 1996, Scheres et al., 2006, Marco et al., 2009). Additionally, and likely related to the previous point, tasks with post-reward delays have been shown to be relatively insensitive to inter-individual differences in symptoms of ADHD (Sonuga-Barke et al., 1992, Scheres et al., 2006, Marco et al., 2009, Paloyelis et al., 2009).

Therefore, in order to gain more insight into the role difficulty waiting may play in relatively strong preferences for small immediate rewards as observed in those with ADHD, additional measures are needed (see for interesting discussions Bitsakou et al., 2009, Marco et al., 2009, Plichta et al., 2009). After all, it is as of yet an empirical question whether feelings of difficulty waiting are associated with choice preferences for small immediate reward in those with ADHD-C. Here, we propose that adding subjective measures to the objective choice measures will shed more light on the motivation that drives children and adolescents with ADHD to prefer the small immediate reward relatively often.

Therefore, we will focus here on a subjective measure of difficulty waiting that we collected in addition to objective choices on TD tasks (note that the objective choices on TD tasks have been reported by Scheres et al., 2010), and answer the following questions: (1) do participants with ADHD-C differ from typically developing controls (TC) and those with ADHD-I in terms of subjectively reported ease/difficulty waiting?; (2) does subjectively reported ease/difficulty waiting correlate with TD rate for each of the groups, and may this association differ between groups? We hypothesized that (1) the ADHD-C group would report more subjectively experienced difficulty waiting than the ADHD-I group and the control group, and (2) subjectively experienced difficulty waiting would correlate significantly with TD rate, to a similar extent for the three groups.