Introduction

Attention-deficit/hyperactivity disorder (ADHD) [1] is a common neurodevelopmental disorder characterized by symptoms of inattention, and/or hyperactivity and impulsivity, which often persists into adulthood [2,3]. ADHD leads to impairments in social and cognitive functioning in an array of settings. Several etiological models have been introduced in the literature and although they all have their own focus, they have in common that they consider ADHD as a disorder of self-regulation, comprising dysregulation of behavior, cognition, and emotions [4–8].

Self-regulation is the ongoing, dynamic, and adaptive modulation of internal state (emotion, cognition) or behavior, by oneself, mediated by central and peripheral physiology [9], and involves a complex interplay of psychological and physiological processes. The idea that bodily processes are crucially involved in self-regulation is not new but recently there is a renewed interest among researchers in the contribution of interoceptive awareness (IA) to emotion, cognition and behavior and how it might be linked to psychopathology. Interoception is the perception of autonomic bodily signals, with IA referring to the awareness of these signals [10]. The anterior insula, as the proposed locus of bodily awareness [11], is heavily involved in interoceptive processes [12–15]. IA has been shown to play a key role in the processing and regulation of emotions, with higher IA being associated with more intense feelings during emotion processing and higher activation of underlying brain regions, and better emotion regulation skills [16,17]. IA also contributes to other cognitive functions required for effective self-regulation such as selective attention, decision-making [18], memory [19], and error processing [20–22]. All these cognitive functions as well as emotion regulation have been reported to be disturbed in ADHD [23–28]. Moreover, ADHD has been associated with structural and functional insula abnormalities [29,30]. It is therefore surprising that to date no study has investigated IA in ADHD. IA has been studied and found altered in other neurodevelopmental disorders, such as autism spectrum disorder (ASD) and tic disorders [31,32]. In the latter study, some of the participants had comorbid ADHD, however whether this contributed to the findings was not tested.

The fact that IA has so far been neglected in ADHD research is even more surprising if one realizes that since the early 70’s dysfunctional autonomic regulation has been proposed to contribute to disturbed self-regulation in ADHD. Zentall defined the optimal stimulation theory and argued that the hyperactive behavior is an attempt to increase stimulation by the child with ADHD experiencing insufficient sensory stimulation [33]. Several researchers in the 80’s elaborated on this notion and proposed that ADHD is associated with hypo-arousal or deficient control of arousal levels [34]. In that same decade, Sergeant and van der Meere started to conduct a large number of studies applying the Cognitive Energetic Model: CEM [35] (see Fig 1.) of Sanders to ADHD [5,36], which led to the most recent and most elaborated model on arousal regulation deficits in ADHD, the state regulation deficit (SRD) account, which is nowadays one of the most influential ADHD models [5–7,36]. For a comprehensive explanation of the CEM and SRD, we would like to refer to existing reviews [5,6, 35,36]. In short, the main idea of the model is that behavior and task performance are dependent on the current energetic state of the organism (a distinction is made between phasic (called “arousal”) and tonic aspects of arousal (called “activation”), with arousal defined as a time-locked phasic physiological response to input, whereas activation refers to tonic readiness for action). Optimal performance requires optimal levels of arousal/activation. When the arousal or activation level is too low or too high, as scanned by an evaluation system, additional effort has to be implemented to counteract such a non-optimal energetic state (i.e. state regulation). The SRD account states that individuals with ADHD have difficulty doing so [5,6,36]. There is now ample behavioral and neurocognitive support from numerous studies for the SRD account. It is out of the scope of the current paper to provide an in-depth review of the evidence that supports the SRD account. For this, we would like to refer the reader to published studies, reviews and a meta-analysis [5–7, 36–40]. Together these studies have provided convincing evidence for a difficulty in children and adults with ADHD in adjusting arousal states. However, it is still not fully understood why this is the case.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. The cognitive energetic model of Sanders (1983).

https://doi.org/10.1371/journal.pone.0205221.g001

According to the CEM, a prerequisite for effective state regulation is state monitoring: the evaluation system checks for discrepancies between the current and the required (target) arousal/activation state and in case of a discrepancy an individual can deliberately allocate extra effort in order to restore equilibrium. As can be seen in Fig 1., the CEM includes feedback loops from the energetic pools (arousal and activation) to the evaluation system. Hence, the CEM explicitly points to the importance of monitoring bodily signals for effective state regulation. If the ability to monitor bodily states and to become aware of these signals is impaired, logically, state regulation will be disrupted as well. We reasoned that if the basic ability of becoming aware of one’s bodily state, i.e. interoceptive awareness, is impaired, this may attribute to the weakened self-regulation skills and in particularly the difficulties in arousal regulation in ADHD. As this aspect has been completely overlooked in prior research, the aim of the current study was therefore to test for difficulties in this basic ability of interoceptive awareness in ADHD.

In the current study, we investigated IA in ADHD, by comparing adults with and without ADHD on objective and subjective indices of IA. IA was objectively measured by means of a well-validated heartbeat perception task [10,15,16], the Mental Tracking Method [41], in which participants are instructed to silently monitor their own cardiac activity during three separate intervals. This task is well-validated and the most frequently used task to assess interoceptive ability, which has been successfully applied in several research domains [10,16,18, 22]. Previous research has shown that this task is sensitive to interindividual differences in IA, and sensitive enough to uncover differences in IA in other neurodevelopmental disorders (e.g., autism spectrum disorder (ASD), tic disorder) [31,32]. Moreover, neuroimaging research further validated this task by showing enhanced activity of the anterior insula during the heartbeat perception task and local gray matter volume in the anterior insula being positively associated with IA [14]. In addition to the heartbeat task, we administered the Body Perception Questionnaire (BPQ) [42], which contains a subscale gauging the overall awareness of several bodily signals, which has frequently been applied in other experimental and clinical studies assessing interoceptive awareness (e.g., [43,44]). The reliability and validity of the awareness scale has been proven to be good as tested in three large independent samples collected in different countries [45].

Recapitulating, ADHD is considered a self-regulation disorder. IA has been shown to be of crucial importance for self-regulation. Weakened self-regulation in ADHD has for a long time been linked to arousal regulation difficulties. It is not known whether these may be due to impaired IA in ADHD. We therefore studied for the first time IA in adults with ADHD, by means of both objective and subjective measures. Based on existing findings, indicating impaired self-regulation and state regulation, we hypothesized lower IA in adults with ADHD.

Method

Results

Performance on the heartbeat perception task

For both groups, the heartbeat perception score obtained at the beginning of the session correlated significantly with the heartbeat perception score acquired at the end (ADHD: r = .80, p < .001; control: r = .92, p < .001), indicating that the estimate of IA was reliable. Moreover, mean heartbeat perception scores obtained were comparable to previous studies. For the ADHD group, the mean heartbeat perception score was .81 (range: .45 - .97), while it was .83 (range: .45 - .97) in the control group. The between-group comparison in mean heartbeat perception score yielded no significant results (F(1, 45) = 0.23, p = .634), with a striking similar standard deviation (.15 for both groups). The Data are reported in Table 1, separately for the two groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Scores on the objective and subjective measure of interoceptive awareness, separated per group.

https://doi.org/10.1371/journal.pone.0205221.t001

As previous research has indicated worse IA as indexed by the heartbeat perception task in females versus males [57], an additional ANOVA was performed with group (ADHD vs. control) and gender as between-subjects factors. This did not change the findings, as neither the main group effect (F(1, 43) = 0.22, p = .641), nor the group by gender effect (F(1, 43) = 0.02, p = .902) was significant. The main effect of gender was also not significant (F(1, 43) = 0.03, p = .859).

No significant correlations between alexithymia, anxiety or depression scores and the mean heartbeat perception score were observed, neither in the ADHD group (alexithymia: r = .05, p = .805; depression: r = -.04, p = .866; anxiety: r = -.10, p = .649), nor in the control group (alexithymia: r = -.18, p = .412; depression: r = .02, p = .922; anxiety: r = -.05, p = .815).

Discussion

As information regarding the momentary bodily state is of crucial importance for effective self-regulation, and in particular state regulation, ADHD may be associated with lower IA. Surprisingly, this hypothesis has not yet been tested. The aim of the present study was thus to investigate IA in adult ADHD by means of an objective and subjective measure. The heartbeat perception task was administered to gain an objective measure of IA, while a questionnaire was used (the awareness subscale of the BPQ) to assess a self-report measure of IA. Performance on the heartbeat perception task was strikingly similar in adults with ADHD compared to healthy controls. Moreover, adults with ADHD and typically developed adults also did not differ on the self-report measure of IA. Findings therefore suggest preserved monitoring of bodily state in adult ADHD, which tentatively suggests that the state regulation deficit and related self-regulatory difficulties in ADHD may not be due to an inability to monitor the current bodily state.

Several possible reasons for this null-result can be formulated. For instance, it could be related to the paradigm used in the current study. However, this suggestion is doubtful since the heartbeat perception task that we applied is a well validated and widely used paradigm to assess IA in different domains [10,15,16,18, 22] and has previously been shown to be sensitive enough to uncover differences in IA in other developmental disorders (autism spectrum disorder, tic disorder) [31,32] and other clinical groups (anxiety and depression) [52]. Moreover, it has been validated in neuroimaging research that showed enhanced activity of the anterior insula during this task [14]. Also, the scores are comparable to scores from previous studies using the same paradigm [16,22] and these scores furthermore indicate that both groups were able to perform well above chance level. Furthermore, corroborating our finding is the striking similarity in variance of the mean heartbeat perception scores between both groups. In addition, comparing only the heartbeat perception score acquired at the beginning of the session between groups, gave the same null result, excluding the possibility that learning effects across both sessions of the task may explain the findings. Also, important to note is that all necessary steps were taken to rule out using any tricks to estimate one’s heartbeats (see methods section).

Moreover, and importantly, preserved IA in ADHD was confirmed by the self-report measure of IA, the awareness scale of the BPQ [42]. The BPQ assesses sensitivity of interoceptive signals in daily life. The findings indicate also no difference between groups for this index. Recent research has shown good reliability and validity indices for the awareness scale, as tested in three large independent samples collected in different countries [45]. In order to stay as close as possible to the original BPQ, we used an official back-translated Dutch version, approved by the authors of the original BPQ. Nevertheless, future research may be needed to validate the Dutch version of the BPQ as used in the current study.

Other factors previously shown to be related to IA could have potentially obscured our findings. Alexithymia, anxiety and depression have been (differently) related to IA [52–55]. In the current study, groups did not significantly differ for symptoms of anxiety, but adults with ADHD reported more symptoms of depression and alexithymia. However, correlations between indices of these factors and scores on both measures of IA were negligible in both groups, suggesting no association between IA and those constructs in our sample. Moreover, as alexithymia has previously been shown to be negatively associated with IA [55], higher alexithymia symptoms in ADHD would result in lower IA in adults with ADHD and cannot explain the absence of a difference in IA between groups. The same reasoning holds for the elevated depression symptoms in ADHD. For mild to moderate levels of symptoms of depression, a negative relation has been reported between IA and symptoms of depression [58]. Hence, more symptoms of depression within our ADHD sample could contribute to lower IA in ADHD, but cannot explain similar IA ability. Anxiety on the other hand has been positively correlated with IA [53]. Thus, elevated levels of anxiety could have resulted in elevated IA in the ADHD group. This factor is however highly unlikely to explain the findings, because the groups did not differ on anxiety symptoms and anxiety was not found to be correlated with IA.

ADHD is a heterogeneous neurodevelopmental disorder and we are aware that the findings may not generalize to all individuals with ADHD. Both adults with ADHD predominantly inattentive and combined subtypes were included in the study. It would be of interest to test for differences in IA between subgroups or subtypes of ADHD, however, in our study, this was not possible as separate groups were not large enough. Also, both men and women with ADHD were included. Groups were however carefully matched on gender and additional analyses showed no influence of gender. With regard to severity of ADHD symptoms, we feel confident that we tested a representative sample of adults with ADHD. All participants had a formal clinical diagnosis provided by a multidisciplinary team including a psychiatrist and this diagnosis was confirmed by a clinical interview. Also, although adults with ADHD did not differ on the awareness subscale of the BPQ, they were found to have more difficulties in autonomic reactivity, stress response and stress style (the other subscales of the BPQ), which is in line with previous research in adults with ADHD, in which elevated physiological stress responses and higher self-reported stress were observed [59–61]. Nonetheless, as to our knowledge this is the first study on IA in ADHD, further research is warranted to replicate our findings in other samples as well as in children with ADHD, preferably with alternative measures as well, before final conclusions on IA in ADHD can be formulated.

The findings suggest that the basic skill of IA is intact in ADHD during a simple heartbeat perception task, but this does not rule out the possibility that becoming aware of bodily signals might be disrupted in other situations in daily life, perhaps as a result of reduced attention or distraction. In other words, in the current situation, when distraction was very limited, adults with ADHD were equally well in monitoring their interoceptive signals, but it has to be tested whether this is still the case when other (distracting) events are happening and additional tasks have to performed, which often is the case in daily life. Similarly, our finding of intact basic ability to become aware of bodily signals, does not exclude the possibility that during situations that require state regulation, this ability is not disrupted. This could be addressed in future research. Alternatively, it could be that IA in ADHD is preserved also in other situations, but that interoceptive information is wrongly applied or interpreted. In other words, individuals with ADHD might be able to monitor the momentary body state but are not able to use the available information sufficiently. This can be further explored in future studies. The finding of preserved IA in ADHD as measured in the current study, instigates the debate on whether the self-regulatory difficulties in ADHD reflect a difficulty in allocating the required effort or are related to a general altered motivational style [6,62]. Further research is needed to examine these hypotheses.

Some limitations have to be mentioned. First, as this is the first study investigating IA in ADHD and ADHD is known to be a heterogeneous disorder, generalization to other samples with ADHD is difficult. Replication of these findings is thus warranted. Second, although the heartbeat perception task we used is a well-validated task that has been extensively applied in previous research, it would be informative if results hold for other heartbeat perception paradigms, such as a heartbeat discrimination task. Third, individuals with ADHD with and without use of medication (methylphenidate) were included in the study. Although we asked participants to stop their medication 48 hours prior to participation, in line with a lot of existing studies, it cannot be fully excluded that medication use may have influenced our findings. An exploratory data check did however not indicate a difference in IA scores between individuals with ADHD who were or were not taking methylphenidate in daily life.

In summary, following recent research indicating IA to contribute to self-regulation and previous research inspired by the SRD model that unequivocally provided support for deficient regulation of energetic state in ADHD, the hypothesis was put forward that the ability to monitor the momentary bodily state may be impaired in ADHD, a hypothesis that had been overlooked in previous research. However, no support was rendered for altered IA in adults with ADHD. The finding of preserved monitoring of bodily state tentatively suggests that self-regulatory difficulties in ADHD may not be related to a lack of awareness about the current bodily state.

Acknowledgments

We would like to thank everybody who participated in the experiment.