Transcranial direct current stimulation improves time perception in children with ADHD

Abstract

Individuals with ADHD struggle with time perception. The ventromedial prefrontal cortex (vmPFC) and dorsolateral prefrontal cortex (dlPFC) are two distinct cortical areas that are involved in the psychopathology of ADHD, including time perception. In the present study, we aimed to explore if modulation of the excitability of these areas with non-invasive brain stimulation alters time perception in ADHD. Twenty-six children with ADHD performed time discrimination and time estimation/reproduction tasks in five sessions with one week interval and specific tDCS conditions. The tDCS conditions were anodal left dlPFC/cathodal right vmPFC, the reversed electrode positioning, isolated anodal left dlPFC, right vmPFC stimulation with extracranial return electrodes, and sham stimulation. An improved time discrimination was observed in all real stimulation conditions compared to sham stimulation, however the results did not differ between real stimulation conditions. For the time reproduction/estimation task, anodal vmPFC/cathodal dlPFC, isolated anodal dlPFC stimulation and isolated anodal vmPFC stimulation with extracranial return electrodes improved time reproduction, but not estimation, compared to sham stimulation. These results imply that the left dlPFC and right vmPFC are involved in time estimation, whereas the right vmPFC might be more critical for time reproduction. This finding is discussed in light of task and stimulus properties.

Introduction

Attention deficit hyperactivity disorder (ADHD) is the most common psychiatric disorder in childhood with two main behavioral symptoms: impaired attention and hyperactivity-impulsivity¹. Three main cognitive theories explain these behavioral symptoms: the executive theory², the delay aversion theory³, and the dual pathway theory⁴. The executive theory focuses on impaired executive functions, especially inhibitory control. The delay aversion/motivational theory explains ADHD symptoms based on impaired time perception and respective reward processing. The dual pathway theory considers a combination of both concepts to describe ADHD symptoms.

Time perception is the ability to experience the duration of an event and passage of time. The impaired time perception in ADHD has been described as over-/under-estimation and/or reproduction of time, with the perception of an overly quick or slow passage of time. Several meta-analyses describe impaired time perception in individuals with ADHD^5–7. Impaired time perception in ADHD could be explained by both, executive and motivational theories. On the one hand, time perception is a dynamic component of executive functions, and on the other hand it is an important parameter which defines the value of stimuli in motivational/reward processing⁵. Children with ADHD struggle with motor timing (unadjusted behavior or motor responses), time estimation (overestimation of elapsed time especially in tasks with sustained attention demand), discrimination (inability to compare the duration of tasks to priorize them) and temporal foresight (inability to consider the consequence of behavior in the future or planning of goal-directed behavior)^8,9. Individuals with ADHD perceive time as longer than healthy controls, indicating the presence of a faster internal clock¹⁰. This leads to inaccurate and impulsive behaviors¹¹, impaired prospective memory¹¹, and impaired sustained attention¹². Time perception plays a critical role in the daily functioning of individuals with ADHD, affecting their ability to regulate behavior¹³, anticipate future outcomes¹⁴, and maintain attention over time¹². The impairment of time perception in ADHD can lead to difficulties in managing time effectively, resulting in challenges such as procrastination¹⁵ and impulsivity¹⁶.

At the neural level, based on neuroimaging studies, impaired time perception in children with ADHD is associated with abnormal functions of the ventromedial prefrontal cortex (vmPFC)^17,18, the dorsolateral prefrontal cortex (dlPFC)^18–22, the parietal lobe^19,21,23,24, the cerebellum^19,21,22,25, the basal ganglia^19,21,22,26, and the amygdala¹⁸. A meta-analysis of fMRI studies in ADHD described a deficit in the left fronto-parieto-cerebellar network with respect to timing functions and right fronto-striatal dysfunctions during inhibition and attention²⁷.

Beyond the above-mentioned correlational studies, non-invasive brain stimulation (NIBS), including transcranial magnetic stimulation (TMS) and transcranial electrical stimulation (tES), studies are used to provide causal evidence about the role of different cortical areas in time perception. TMS studies in healthy individuals showed a relevance of the right and left dlPFC^28–30 in time perception. Transcranial direct current stimulation (tDCS) studies describe a role of the right and the left posterior parietal cortex (PPC) in time perception^31–34. Also, transcranial random noise stimulation (tRNS) identified a role of the left dlPFC³⁵ and primary visual and auditory cortices³⁵ in time perception of healthy individuals³⁵. No NIBS study has however so far explored a possible role of the vmPFC in time perception, as suggested by above-mentioned neuroimaging studies.

TDCS has been used in a more general sense for amelioration of ADHD symptoms^36,37 and improvement of a variety of cognitive impairments in ADHD, such as inhibitory control^38–43, attention^36,44,45, reward processing⁴⁶, and working memory^47,48, for comprehensive reviews see^49,50.

With respect to the role of time perception for the development of ADHD symptoms, given the well-documented cortical underpinnings of time perception in healthy humans, and the impact of NIBS on time perception, we aimed to evaluate the impact of tDCS on time perception in children with ADHD. We targeted two cortical areas, the dlPFC and the vmPFC, which are involved in the psychopathology of ADHD, and time perception. The ADHD-specific deficits outlined in the main theories of ADHD can be explained by impaired functionality of these areas. The executive theory focuses on a dysfunctional dlPFC, the motivational theory focuses on vmPFC impairments, and the dual pathway theory suggests an impaired interaction between the dlPFC and the vmPFC. The present study used isolated stimulation of these areas with an extracranial reference electrode and bipolar stimulation of both areas, with anodal tDCS over one area combined with cathodal tDCS over the other, to reveal the role of these cortical areas in time perception in children with ADHD.

Materials and methods

Participants

Twenty-six children with ADHD participated in this study. All participants were diagnosed with ADHD according to the Diagnostic and Statistical Manual of Mental Disorders 5th ed.⁵¹ diagnostic criteria by a professional child psychiatrist. Three children were excluded from the study because of missing at least one session. Twenty-three children completed all five sessions of the study and were included in the data analysis. These had an age range between 7–14 years (mean age: 10.087 ± 2.02, 22 boys, 1 girl). Based on the SNAP_IV rating scale⁵², 21 children had the combined subtype, and 2 participants the inattentive subtype of ADHD. Symptom severity based on the SNAP-IV was mild for 2, moderate for 16, and severe for 5 children. Table 1 shows the demographic characteristics of the participants. G*power was used to determine the required sample size⁵³. Based on a power of 0.95, an alpha level of 0.05, and a medium effect size (f = 0.40) suggested for tDCS studies⁵⁴, the required sample size for our study design and a repeated measures ANOVA was 16. As this study included five sessions with one week interval per participant, given our previous experience, we added ten more subjects to compensate for drop-outs, and unforeseen variability. Participants were unaware of the stimulation protocol and aims of the study. Participants were drug-naive except three, who were medicated with methylphenidate. For these participants, the washout period of the substance was 24 h before each stimulation session, and thus more than five half-lives (3.5 h). All participants were right-handed, and had normal or corrected to normal vision, based on self-reports. None of the participants had presence or a past history of head trauma or other neurological or psychiatric disorders, based on their medical records in the psychiatry clinic. The procedures were in accordance with the ethical standards of the Helsinki Declaration of 1975, as revised in 1983. The study was approved by the ethical committee of Shahid Beheshti University (Ethics Code: IR.SBU.REC.1400.033).

Table 1.

Demographic characteristics of the participants.

Variables	M(sd)/ Description
Age	10.08 (2.02)
Education	4.08 (2.02)
ADHD- Score	28.91 (8.21)
ADHD- Subtype	2ADHD-I, 21ADHD-C
Gender (F:M)	1:22

M: mean, sd: standard deviation, ADHD-I: inattentive ADHD, ADHD-C: combined ADHD.

Time estimation/reproduction task

In this task, participants have to estimate an elapsed time interval and then reproduce this time interval⁵⁵. In the present study, the participants were instructed to estimate the duration of a bulb lightening and then lighten the bulb for the same duration by pressing a button (Fig. 1). The “light-on” period included two short intervals (3 and 6 s), and two long intervals (12 and 24 s). Each time interval was presented 16 times. The time intervals were randomized, excluding immediate repetition of the same interval. Correct answers were defined as the extact duration ± 16% (i.e., for the interval of 12 s, the range of 10–14 s was considered as correct answer). According to the instructiuon, the task had two demands: estimation and reproduction. The outcome measures of this task were accuracy (percentage of all stimuli of the respective category) of the short duration (ESD), and long duration (ELD) estimations, reproduction of short duration (RSD), and long duration (RLD), and total accuracy over all conditions.

Fig. 1.

Schematic drawing of the time discrimination task (TDT; left) and time estimation/reproduction task (TERT; right). In the TDT, after a fixation cross, two stimuli with different durations are presented serially and the participant has to decide which of these had a longer duration. In the TERT, after presentation of a fixation cross, a lightening bulp is presented, and the participant has to estimate the lightening duration and reproduce this duration by pressing a button.

Time discrimination task

In this task, two stimulus durations, namely target and comparison, are presented to the participants serially. Participants are instructed to identify the stimulus with the longer duration. In the present study, two visual (black square) or auditory (a tone) stimuli were presented for long (1000–1200 ms) or short (200–220 ms) periods. The square had a size of 100*100 pixel, the frequency of the tone was 1000 Hz at 80 Db. The target duration was fixed in all trails (either 200 or 1000 ms), whereas the duration of the comparison stimulus differed incrementally by 5 ms steps for up to 20 ms (for short duration stimuli) or by 25 ms for up to 200 ms (for long duration stimuli), Fig. 1. The inter-trial interval was 1000 ms with an inter-stimulus interval of 500 ms. This task has been used earlier in children with ADHD⁵⁶. The outcome measure of this task is the percentage of accurate discriminations in percent of all possible answers in a category (auditory short duration (ASD), auditory long duration (ALD), visual short duration (VSD), visual long duration (VLD) stimuli).

tDCS protocol

The neurostim2 transcranial stimulator (Medinateb, Iran) was used for brain stimulation. An electrical direct current of 1.5 mA was applied through a pair of saline-soaked sponge electrodes with a size of 25 cm² (5 × 5 cm) for 15 min. Electrode placement was based on the 10–20 EEG international system. The five montages applied in five separate sessions were: (a) anodal left dlPFC (F3)/cathodal right vmPFC (Fp2), (b) anodal right vmPFC (Fp2)/cathodal left dlPFC (F3), (c) anodal left dlPFC (F3) with the return electrode placed over the right shoulder, (d) anodal right vmPFC (Fp2) with the return electrode over the left shoulder, and (e) sham stimulation. In the sham stimulation condition, the electrode placement was identical with one of the real stimualation conditions, and selected randomly for each participant. The electrical current was ramped up and down in all stimulation conditions for 30 s. In the sham stimulation condition, the current was ramped down immediately after ramping up for 30 s without the participants’ awareness to generate the same sensation as in the real condition. We used SimNIBS [71–75] to examine the distribution of the electric field (EF), as illustrated in Fig. 2.

Fig. 2.

Electric field (EF) distribution generated by the transcranial direct current stimulation (tDCS) montage used in this study. The EF model demonstrates targeted stimulation in the expected regions, supporting the montage choice. Frontal and lateral views of different montage from top to buttom includes: F3/FP2, FP2/F3, F3/EC, FP2/EC. Note: F3: left dorsolateral prefrontal cortex, Fp2: right ventromedial prefrontal cortex, EC: extracranial, areas pre and post the dash indicate anodal and cathodal electrode placement respectively.

Procedure

After checking for the inclusion criteria and explanation of the study procedure, the parents of the participants signed a written informed consent form and filled in the ADHD rating scale, SNAP_IV. Then, the participants were introduced to the study procedure and the time perception tasks. The stimulation sessions were performed in a in a quiet room in the Omid rehabilitation clinic (Kerman, Iran). There was a one-week interval between each stimulation session to prevent carry-over effects. The order of stimulation conditions was randomized across participants. For this purpose, we defined a list of all possible session sequences and dedicated a number to each sequence. Participants were instructed to select a number unaware of the respective sequence assigned. The selected number was deleted from the list after selection. The participants were blind to the stimulation conditions in the study. In each session, five minutes after the start of stimulation, the time discrimination task and the time estimation/reproduction task were performed in random order, and took about 10 min, Fig. 3. After each stimulation session, participants answered a questionnaire of side-effects (Brunoni et al., 2011) and had to guess the stimulation condition (real or sham). The introduction of the parents and participants to the study procedures and tasks, tDCS, the explanation of task performance, and completion of the questionnaires were performed by one researcher/author (F.M.).

Fig. 3.

Experimental procedures. In each study, participants received one of the tDCS protocols in randomized order. The tasks was performed in each session 5 min after stimulation and finally completed the side effect survey. Note: F3: left dorsolateral prefrontal cortex, Fp2: right ventromedial prefrontal cortex, EC: extracranial, areas pre and post the dash indicate anodal and cathodal electrode placement respectively. TERT: time estimation/reproduction task, TDT: time discrimination task, SES: side effect survey.

Data analysis

This study had a single-blind, complete crossover design. Data analyses were conducted using the statistical package SPSS for Windows, version 21. Normality and homogeneity of variance of the data collected in each stimulation condition were confirmed using the Levene test. To explore the effect of tDCS on task performance, two repeated measure analyses of variance (ANOVA) were conducted. For the time discrimination task, the within-subject factors were: intervention (5 stimulation conditions) task modality (visual or auditory), and stimulus duration (short or long). The dependent variable was percent of correct responses. For the time estimation/reproduction task, the within-subject factors were: stimulation conditions (5 conditions) stimulus duration (2 values: long and short), and the task demand (2 values: estimation and reproduction). Accuracy of performance served as dependent variable. Mauchly’s Test of Sphericity was conducted to test for sphericity of the data, and degrees of freedom were corrected using the Greenhouse–Geisser method, if required. Post hoc analyses were carried out in case of significant results in the ANOVAs by Fisher’s Least Significant Difference (LSD) tests. Session order and task order were added as covariates in additional ANCOVAs. One factorial repeated measure ANOVAs and and chi square tests were conducted for evaluation of side effects and blinding data. A significance level of p < 0.05 was used for all statistical comparisons.

Results

All participants tolerated the stimulation without major side effects, except one who was excluded after the first session (which was the sham stimulation condition), because of severe headache. Blinding was successful based on participants’ guesses about real or sham conditions after the respective experimental sessions. Guesses of real stimulation for anodal left dlPFC (F3)/cathodal right vmPFC (Fp2) were 95%, for anodal right vmPFC (Fp2)/cathodal left dlPFC (F3) 82%, for anodal left dlPFC (F3) with the return electrode placed over the right shoulder 82%, for anodal right vmPFC (Fp2) with the return electrode over the left shoulder 95%, and for sham stimulation 95%. These guesses did not differ between stimulation conditions (χ2(1) = 0.905, p = 0.342). Table 2 shows the mean values and standard deviations of side effects in the different stimulation conditions. The one-way ANOVAs showed no significant differences between conditions for pain (F_4,88 = 1.43, p = 0.24), burning (F_4,88 = 0.55, p = 0.63), itching (F_4,88 = 1, p = 0.412), fatigue (F_4,88 = 1.235, p = 0.304), vertigo (F_4,88 = 1.92, p = 0.159) and nausea (F_4,88 = 1.11, p = 0.350), but a significant difference of confusion (F_4,88 = 3.32, p = 0.037). The post hoc analysis showed significantly larger confusion for the anodal vmPFC with an extracranial reference electrode tDCS condition compared to all other stimulation conditions.

Table 2.

Side effects of tDCS (means and sd) in the different stimulation conditions and the respective ANOVAs results.

Measures	Conditions, M(SD)					Statistics
	Sham	F3/FP2	FP2/F3	F3/EC	FP2/EC	df	F	P	ηp2
Pain	.13(.34)	.08(.28)	.30 (.63)	.26(.68)	.08 (.28)	4,88	1.43	.244	.061
Confusion	.04(.20)	0(0)	.08 (.28)	.04 (.20)	.26(.44)	4,88	3.37	.037	.133
Burning	.21(.42)	.43(.72)	.26(.68)	.39(.78)	.30(.70)	4,88	.55	.63	.025
Itching	.13(.34)	.08 (.28)	.08(.28)	.13 (.34)	.26 (.44)	4,88	1	.412	.043
Fatigue	0(0)	.08 (.28)	.08 (.28)	.04 (.20)	.17(.38)	4,88	1.23	.304	.053
Vertigo	.13(.62)	.13(34)	.17(.49)	.13 (.34)	.17 (.38)	4,88	.072	.975	.003
Nausea	0 (0)	.04(.20)	.08 (.28)	0 (0)	.08 (.28)	4,88	1.11	.343	.048

F3: left dorsolateral prefrontal cortex, Fp2: right ventromedial prefrontal cortex, EC: extracranial, areas pre and post the dash indicate anodal and cathodal electrode placement respectively.

In the time discrimination task, the ANOVA showed a significant main effect of intervention (F_4.88 = 4.861, p = 0.001, partial η2 = 0.181). The main effects of duration (F_1.22 = 2.245, p = 0.148, partial η2 = 0.093), modality (F_1.22 = 1.960, p = 0.175, partial η2 = 0.082) and the interactions of intervention and duration (F_4.88 = 1.012, p = 0.406, partial η2 = 0.044), intervention and modality (F_4.88 = 0.365, p = 0.833, partial η2 = 0.016), and the interaction of intervention, duration, and modality (F_4.88 = 0.679, p = 0.608, partial η2 = 0.030) were non-significant. The LSD post hoc analysis showed that accuracy in the sham stimulation condition was lower than in the anodal left dlPFC/cathodal right vmPFC (mean difference(MD) = 5.97, p = 0.001), anodal right vmPFC/cathodal left dlPFC (MD = 8.15, p = 0.001), anodal left dlPFC with extracranial return electrode (MD = 7.04, p < 0.001), and anodal right vmPFC with extracranial return electrode (MD = 7.50, p = 0.001) conditions, Table 3.

Table 3.

Descriptive statistics, means (+ /-standard deviation), of cognitive task performance in percentage of accuracy.

Tasks	Sham	F3/FP2	FP2/F3	F3/EC	FP2/EC
TDT
VLD	45.00 (11.45)	50 .04 (13.04)	56.43 (16.23)	56.08 (15.49)	52.43 (15.02)
VSD	48.86 (10.56)	53.00 (18.84)	51.08 (19.13)	51.21 (15.00)	52.13 (18.05)
ALD	43.56 (14.51)	56.08 (21.68)	56.34 (17.90)	54.73 (14.91)	55.56 (22.46)
ASD	42.82 (12.06)	45.00 (14.12)	49.00 (13.12)	46.39 (12.48)	50.13 (13.61)
Total	45.08 (5.22)	50.91(6.99)	53.00 (9.23)	52.04 (8.48)	52.60 (9.03)
TERT
RLD	37.39 (15.51)	50.86 (17.55)	57.82 (15.94)	56.08 (16.44)	52.60 (17.63)
RSD	30.65 (17.85)	33.91 (18.27)	39.78 (18.73)	31.73 (20.14)	36.08 (19.24)
ELD	7.82 (23.15)	9.56 (20.55)	9.56 (20.10)	5.21 (12.01)	14.78 (24.09)
ESD	5.22 (17.01)	10.86 (19.2)	10.86 (23.91)	6.52 (12.28)	8.69 (22.62)
Total	23.21 (17.00)	26.43 (13.69)	29.65 (14.57)	24.95 (9.50)	28.17 (14.53)

F3: left dorsolateral prefrontal cortex, Fp2: right ventromedial prefrontal cortex, EC: extracranial, areas pre and post the dash indicate anodal and cathodal electrode placement respectively. TERT: time estimation/reproduction task, TDT: time discrimination task, VLD: visual long duration, VSD: visual short duration, ALD: auditory long duration, ASD: auditory short duration, RLD: reproduction of long duration, RSD: reproduction of short duration, ELD: estimation of long duration, ESD: estimation of short duration.

In the time estimation/reproduction task, significant main effects of intervention (F_4.88 = 4.607, p = 0.002, partial η2 = 0.173), task demand (F_1.22 = 103.532, p < 0.001, partial η2 = 0.825), stimulus duration (F_1.22 = 35.571, p < 0.001, partial η2 = 0.619), the interaction of intervention and task demand (F_4.88 = 2.616, p = 0.040, partial η2 = 0.106), the interaction of duration and task demand (F_4.88 = 26.783, p < 0.001, partial η2 = 0.549), and the interaction of intervention, task demand and stimulus duration (F_4.88 = 2.535, p = 0.046, partial η2 = 0.103) were significant. The interaction of intervention and duration (F_4.88 = 1.174, p = 0.328, partial η2 = 0.051) was non-significant.

The LSD post hoc tests conducted for the intervention, independent of task demand and stimulus duration, revealed larger accuracy during anodal vmPFC/cathodal dlPFC stimulation (MD = 9.238, p = 0.008) and anodal vmPFC/extracranial refrence electrode (MD = 7.771, p = 0.017) stimulation, as compared to the sham stimulation condition. For task demand, independent of stimulation condition and stimulus duration, the LSD pairwise comparisions showed larger accuracy for time reproduction compared to time estimation (MD = 33.782, p < 0.001).

For stimulus duration, independent of intervention and task demand, the post hoc analysis showed larger accuracy for the long compared to the short stimulus durations (MD = 8.739, p < 0.001). The LSD post hoc comparisions conducted for the interaction of intervention and task demand revealed larger accuracy of time reproduction in the anodal vmPFC/cathodal dlPFC (MD = 14.783, p = 0.001) and anodal vmPFC/extracranial reference electrode (MD = 10.32, p = 0.022) conditions, as compared to sham stimulation. The LSD post hoc tests conducted for the interaction of intervention, task demand, and stimulus duration revealed only larger accuracy of time reproduction in the long duration stimulus condition during anodal vmPFC/cathodal dlPFC stimulation (MD = 20.435, p = 0.001), anodal dlPFC/ extracranial reference electrode (MD = 18.696, p = 0.001), and anodal vmPFC/ extracranial reference electrode (MD = 15.217, p = 0.027) conditions, as compared to the sham stimulation condition, Fig. 4. Additional ANCOVAs that included session order and task order as covariates revealed no significant effects of either task or session order on the tDCS outcomes.

Fig. 4.

The effects of tDCS on accuracy of time estimation/reproduction (above) and time discrimination (below) are shown. Note: Different tDCS conditions (anodal/cathodal in order) are shown in the X axis, and labeled based on the 10–20 EEG system, the Y axis refers to accuracy of performance. F3: left dorsolateral prefrontal cortex, Fp2: right ventromedial prefrontal cortex, EC: extracranial, areas pre and post the dash indicate anodal and cathodal electrode placement respectively. VLD: visual long duration, VSD: visual short duration, ALD: auditory long duration, ASD: auditory short duration, RLD: reproduction of long duration, RSD: reproduction of short duration, ELD: estimation of long duration, ESD: estimation of short duration. Total refers to the both estimation and production components in the time estimation/reproduction task and both auditory and visual signal in the time discrimination task.The brackets show significance at a p-level of .05 based on the results of the pairwise comparisons between stimulation conditions. The bars are showing the means, error bars represent standard error of means.

Discussion

The present study aimed to evaluate the impact of tDCS on time perception in children with ADHD. The results show a positive impact of tDCS over the dlPFC and vmPFC on time discrimination, and reproduction. This effect was domain-specific. In the following section we discuss the impact of the different intervention conditions, task modalities, demands, and stimulus durations on performance.

The relevance of intervention conditions

In the present study, we targeted the left dlPFC and the right vmPFC in five different stimulation conditions, namely sham, anodal dlPFC/cathodal vmPFC, anodal vmPFC/cathodal dlPFC, anodal dlPFC/ EC return electrode, and anodal vmPFC/EC return electrode. These intervention conditions provided an opportunity to study the relevance of the dorsolateral, and ventromedial prefrontal cortices for time perception in combination, and separately. The impact of the specific stimulation conditions differed in the two tasks. For time discrimination, all real stimulation conditions improved accuracy of time discrimination, indicating the relevance of both areas for successful conduction of this task. Accordinlgy, earlier neuroimaging studies describe a role of the vmPFC^17,18, and the dlPFC^18–22 in time perception. With respect to the dual pathway hypothesis of ADHD both, executive and temporal dysfunctions are central behavioral symptoms of ADHD^4,57. Time perception as an executive function involves both, the dlPFC and the vmPFC⁵⁸. In detail, two main components of the time discrimination task might be attributed to the dlPFC and vmPFC in the light of known functions of these areas. The vmPFC might be involved in implicit perception of stimulus duration, whereas the dlPFC might be involved in decision making about the equality or difference of stimulus presentation duration.

In the time estimation/reproduction task, independent of task demand and stimulus duration, only two montages improved time perception, and both montages involved tDCS-induced activity enhancement of the vmPFC. With respect to the task demand and stimulus duration, the effects of stimulation were limited to the reproduction of long duration stimuli. Here, in addition to the efffects of anodal vmPFC stimulation with an extracranial reference electrode or the cathode positioned over the dlPFC, anodal dlPFC stimulation with an extracranial reference electrode improved accuracy of long duration stimulus reproduction. This pattern of results shows an involvement of both prefrontal cortices in time reproduction, however it might be assumed that the vmPFC is more critical for this function, since for this area performance was improved by upregulation of this area alone, and combined with downregulation of the dlPFC. In contrast, only selective upregulation of the dlPFC, but not dlPFC upregulation combined with downregulation of the vmPFC had a performance-improving effect, which suggests that dowregulation of the vmPFC counteracted any performance improvement accomplished by dlPFC upregulation. One could thus speculate that in this situation, the performance-improving effect of anodal dlPFC stimulation is accomplished via its impact on the vmPFC, which would be abolished by vmPFC downregulation.

With respect to task demands, the time estimation/reproduction task is performed without time pressure or a forceed choice situation. In this task, participants should estimate the duration of the stimulus, enter a respective numerical estimate, and reproduce the duration of the stimulus. Therefore, compared with the time pressure and forced comparative estimation-based choice in the time discrimination task, the executive demand of that task is lower, which might explain the less critical contribution of the dlPFC suggested by our results. In line with the present study, an earlier tDCS study revealed reduced delay discounting during anodal vmPFC/cathodal dlPFC stimulation in children with ADHD⁴⁶, resulting in a more conservative decsion making, which might be explained by better future-/consequence-related representations.

However, the effects of tDCS are not limited to isolated modulation of the targeted cortical areas but may also involve broader network-level interactions. The vmPFC and dlPFC are key nodes within the frontostriatal and corticocerebellar networks, which are critical for temporal processing and executive functions⁵⁹. The anodal stimulation of the right vmPFC and left dlPFC likely influences interconnected regions, including the basal ganglia, cerebellum, and other cortical areas, via network-level effects⁶⁰. These regions are known to contribute to temporal processimg^61,62. This network-based perspective underscores the importance of considering tDCS effects beyond the immediate stimulation sites, as they likely engage and modulate distributed neural circuits involved in task performance. Future studies employing advanced computational modeling and functional connectivity analyses could further elucidate the specific network dynamics underlying these effects.

The relevance of time perception domains

In the present study, three domains of time perception, namely time discrimination, time estimation, and time reproduction, were evaluated in two tasks. The estimation and reproduction domains were measured as two components of one task, which provides the possibility to compare performance in these components. The results showed a better performance of time reproduction compared to time estimation, independent of stimulation condition. Furthermore, for the stimulation conditions anodal vmPFC/cathodal dlPFC, anodal vmPFC/extracranial reference electrode, and anodal dlPFC/extracranial reference electrode, performance was affected only for time reproduction. As suggestd above, the vmPFC might be more critical than the dlPFC with respect to performance of this specific task. This might be explained by the retrospective demands of time estimation compared to the prospective demand of time reproduction. The vmPFC is promintently involved in the representation of future events and prospective processing^63,64. Furthermore, time estimation in this task required a numerical analysis, which involves the dlPFC, whereas time reproduction can be performed initutively, without numerical processing, and thus might involve primarily the vmPFC.

The relevance of stimulus properties

In the current study, two stimulus properties were assessed in the two tasks: stimulus modality and duration in the discrimination task and stimulus duration in the estimation/reproduction task. The results of the time discrimination task revealed null effects of stimulus duration and modality. Although a trend of better performance was found for long duration, as compared to short duation discrimination, this trend was simillar in both modlaities. In accordance, two metaanlyses described similar patterns of time perception impairment for visaul and auditory modalities in individuals with ADHD^5,65. In the estimation/reproduction task, the positive impact of anodal vmPFC/cathodal dlPFC, anodal vmPFC/extracranial reference electrode, and anodal dlPFC/extracranial reference electrode stimulation conditions on performance was limited to the long stimulus duration. Earlier studies described intact time perception in individuals with ADHD for short duration stimuli^23,66. We conclude that an intact short-time perception in children with ADHD might have.caused a ceiling effect which left no room for improvement via tDCS in the current study.

Conclusions, limitations and future direction

Some limitations of the present study should be taken into account. This was an exploratory study with a relatively small sample size, and a single session intervention approach. Repetitive stimulation sessions with follow-up are proposed for future studies to explore the suitability of this intervention for clinical application. Furthermore, for evaluation of clinical applicability, the transferability of the intervention effect to clinical ADHD symptoms as an outcome measure should be taken into consideration. The single blinding approach of the present study is another limitation. Furthermore, we discussed our finding in the light of cognitive models of ADHD and respective neural underpinings. Furthermore, although session order and task order were controlled through the crossover design, the cumulative impact of repetition could still have contributed to changes in performance that were not solely attributable to the tDCS intervention. The neuroanatomy of ADHD and the neural underpinings of time perception are not limited to the two targeted areas explored in the present study. For a more comprehensive picture, also other areas should be explored. Furthermore, combination of brain stimulation with neuroimaging methods might provide more precise information, especially with respect to the speculation about the contribution of the dlPFC to time reproduction. We explored two tasks, one for time discrimination and the other for time estimation/production. Future studies could consider to include also other domains of time perception, including timing and temporal foresight.

Furthermore, a notable limitation of our study is the potential spread of current beyond the intended target regions (left dlPFC and right vmPFC) due to the use of a contralateral shoulder return electrode. This montage may induce significant current flow through contralateral posterior cortical areas, subcortical structures such as the basal ganglia, and the cerebellum, which are involved in temporal processing. Future studies should consider alternative electrode montages or computational modeling to better isolate and verify the target areas.

The main results of this study suggest that tDCS improves impaired time perception in ADHD. Speficically time reproduction was improved by excitability-enhancing tDCS over the vmPFC and partially also dlPFC, while discrimination performance was improved similarly by both, dlPFC and vmPFC stimulation. These findings might be explained in the light of different demands of the applied time perception tasks. The time discrimination task includes both, executive and attentional components to a similar degree, whereas time reproduction has less executive demands. In the present study, we showed that prefrontal tDCS over specific subregions improves time perception in ADHD. If this cognitive improvement is therapeutically valuable should be explored in future multi-session studies, which induce more stable effects, and by adding readouts of clinical symptoms.

Author contributions

V.N: Conceptualization, analysis, writing, and supervision ; FMK: Data gathering; MAN: Review.

Funding

This study did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.