Near and Far Transfer Effects of Computerized Progressive Attention Training (CPAT) Versus Mindfulness Based Stress Reduction (MBSR) Practice Among Adults With ADHD

Abstract

Objective:

The present study evaluated the near (attention) and far (reading, ADHD symptoms, learning, and quality of life) transfer effects of a Computerized Progressive Attention Training (CPAT) versus Mindfulness Based Stress Reduction (MBSR) practice among adults with ADHD compared to a passive group.

Method:

Fifty-four adults participated in a non-fully randomized controlled trial. Participants in the intervention groups completed eight 2-hr weekly training sessions. Outcomes were assessed before, immediately after, and 4 months post-intervention, using objective tools: attention tests, eye-tracker, and subjective questionnaires.

Results:

Both interventions showed near-transfer to various attention functions. The CPAT produced far-transfer effects to reading, ADHD symptoms, and learning while the MBSR improved the self-perceived quality of life. At follow-up, all improvements except for ADHD symptoms were preserved in the CPAT group. The MBSR group showed mixed preservations.

Conclusion:

Both interventions have beneficial effects, however only the CPAT group exhibited improvements compared to the passive group.

Introduction

Attention Deficit Hyperactivity Disorder (ADHD) is a prevalent chronic neurobiological disorder characterized by behavioral symptoms of inattention and/or hyperactivity/impulsivity. ADHD in adults is often associated with poor outcomes in academic, occupational, social, and emotional functioning (Able et al., 2007; Arnold et al., 2020; Barkley et al., 2006; Biederman et al., 2006, 2012).

In particular, and of interest to the current study, academic achievements are closely related to the ability to read a text effectively. Individuals with ADHD have reported forgetting what they have just read at the top of the page by the time they get to the end (Barkley 2006). One probable cause of this phenomenon is elevated levels of mind wandering (Lanier et al., 2021). Quite often, students with ADHD are forced to read texts repeatedly, a few times to comprehend what they just read.

Medication

Currently, ADHD is primarily treated with medication, which indeed ameliorate many of the core ADHD symptoms (Faraone et al., 2006). However, about third of the people (Banaschewski et al., 2004; Rabiner et al., 2010) with ADHD do not respond to medication or have adverse responses that prevent them from medication treatment (Aagaard & Hansen, 2011; Banaschewski et al., 2004; Charach & Fernandez, 2013; Rabiner et al., 2010; Wilens et al., 2002). Moreover, most individuals treated with medication continue to evidence some functional impairments (e.g., Swanson et al., 2011) and it seems that medication has less impact on everyday functional outcomes (Pelham et al., 2017). As a result, approximately 20% of individuals with ADHD cease stimulant use within the first year of taking them (Toomey et al., 2012). Thus, there is a strong need for the development and investigation of effective non-pharmacological interventions.

Cognitive Training

One of the non-pharmacological approaches to treat ADHD is the direct process cognitive training approach, which addresses core cognitive processes. Notably, several reviews and meta-analysis of cognitive training for ADHD concluded that there is not enough evidence to support the efficacy of cognitive training in ADHD (Cortese et al., 2015; Rapport et al., 2013; Sonuga-Barke et al., 2013). In particular, most of the reviewed studies failed to show transfer of the training outcomes to other cognitive domains and/or everyday functioning when blinded raters assessed the efficacy of the interventions. Although, S. J. Beck et al. (2010) and Bigorra et al. (2016) (in Veloso et al., 2020), reported significant results of assessments by blinded raters showing that blind raters can detect changes in everyday situations following cognitive training interventions.

While most of the existing literature of cognitive training for ADHD focuses on working memory training, multi-component training that targets a broader range of neuropsychological deficits may be more successful given the complex and neuropsychological heterogeneous nature of ADHD (Cortese et al., 2015). Moreover, it has been asserted that cognitive training for ADHD should target core deficits that are well documented (Rapport et al., 2013). A recent study by Kolodny et al. (2017), addressed the above suggestions using the Computerized Progressive Attention Training (CPAT) program that was developed by Shalev et al. (2007). The CPAT is a theory driven program, which focuses on training four components of attention and cognitive control. In Kolodny et al.’s study (2017), 30 higher education students with ADHD were randomly assigned either to the CPAT or to the active control (computer games) groups and completed sixteen 1-hr training sessions across 8 weeks. Participants in the CPAT group exhibited significantly greater improvements in selective-spatial attention and in executive attention (assessed using different tasks from the trained ones) compared to participants in the active control group, and these gains were maintained at follow-up 3 months after the program ended. These results provided strong evidence for near transfer effects of the CPAT in adults with ADHD. However, no change in either group was found in ADHD symptoms. The current study aimed to fill the gap and elaborate the outcome measures to assess whether the near transfer effects of the CPAT program could be extended to everyday functioning.

Mindfulness-Based Stress Reduction

Mindfulness training is described operationally as “the awareness that emerges through paying attention on purpose, in the present moment, and non-judgmentally to the unfolding of experience moment by moment” (Kabat-Zinn, 2003, p. 145). Mindfulness has been used to increase awareness and to teach the individual to identify and respond to the physiological and psychological processes that are involved in maladaptive behaviors and emotions (Bishop et al., 2004). Among individuals with ADHD, studies on meditation-based training suggest that meditation may enhance certain attentional capabilities (Baijal & Gupta, 2008) and may have possible benefits in reducing symptoms of ADHD (Cairncross & Miller, 2020). The severity of ADHD symptoms is the most studied outcome of mindfulness interventions in ADHD participants. Results from two recent meta-analyses (Cairncross & Miller, 2020; Oliva et al., 2021) show that mindfulness significantly reduced attention related and total ADHD symptoms (medium effect sizes) in adults diagnosed with ADHD. However, subgroup analysis and meta-regression were consistent in supporting a significantly higher efficacy of mindfulness versus passive controls than versus active controls (Oliva et al., 2021). Beyond ADHD symptoms, neuropsychological outcomes were also studied in the adult population with and without ADHD. Practicing mindfulness improved sustained attention in a non-ADHD sample (Semple, 2010) and in a heterogeneous sample (Tarrasch et al., 2016), attentional conflict in an ADHD sample (Zylowska et al., 2008) and in a non-ADHD sample (Moore & Malinowski, 2009), response inhibition in an ADHD sample (Bueno et al., 2015) and in a heterogeneous sample (Tarrasch et al., 2016), and selective attention in a non-ADHD sample (Jensen et al., 2012). Also, improved perceived quality of life was documented in an ADHD sample (Bueno et al., 2015). A study that was not conducted specifically on students with ADHD but relevant to the current study showed that practicing mindfulness reduced mind wandering during text reading (Mrazek et al., 2013). Another relevant review that included 101 randomized controlled trials with heterogeneous samples (psychological problems, general population, and somatic conditions) assessed the effect of MBSR on quality of life. It was found that MBSR has a medium effect size on quality of life at post-intervention when compared to an inactive (waiting list or treatment as usual) control, and the same effect as other active interventions when compared to other active treatments (de Vibe et al., 2017).

Thus, our goal in the current study was to investigate the effectiveness of two intervention programs on attention performance and the transfer to academic and everyday functioning among higher-education students with ADHD. The present study was comprised of three groups: (1) An experimental group who practiced Computerized Progressive Attention Training (CPAT; Shalev et al., 2007), (2) An active control group who practiced Mindfulness Based Stress Reduction (MBSR; Kabat-Zinn, 1990) with adaptation for ADHD, and (3) A passive control group. Outcomes in the academic domain included both objective eye-movements measures reflecting reading efficiency and subjective self-reported learning performance measures. In the context of everyday functioning outcomes, self-reported ADHD symptoms and self-perceived quality of life were recorded. The inclusion of objective measures of reading efficiency allowed us to overcome the potential bias of using subjective self-reported outcomes of the interventions that were investigated in the study.

Based on previous studies, we expected that both intervention groups would show improvements in performance in neuropsychological attention tasks compared to their initial performance before the interventions. Based on Shalev et al. (2007) we hypothesized that reading efficiency and perceived learning performance will improve whereas ADHD symptoms will decrease for the CPAT intervention group. We also expected to get reduced ADHD symptoms for the MBSR group based on previous studies. In addition, based on Mrazek et al. (2013) and on de Vibe et al. (2017) we expected improvements in reading efficiency and in self-perceived quality of life. Yet, as CPAT is derived from a theory referring specifically to ADHD, and targets well documented core deficits in ADHD, we hypothesized that CPAT will be more effective compared to MBSR. We expected no improvements for the passive control group. We also expected that the positive outcomes would be preserved at follow-up testing that took place 4 months after the end of the interventions.

Method

Participants

Ninety-two adults previously diagnosed (self-reported) with ADHD by a psychiatrist or neurologist were recruited to participate in the study through advertisement within university and college campuses. To exclude participants with clinical levels of either anxiety or depression, participants had to score below 60 on the State-Trait Anxiety Inventory (STAI; Spielberger, 1983); and below 20 on the Beck Depression Inventory (BDI; A. T. Beck et al., 1961). Remaining candidates underwent a psychiatric interview-based evaluation conducted by author ST—a certified psychiatrist with a specialization in ADHD. The evaluation included medical and psychiatric history and mental status examination according to DSM-5 (APA, 2013) criteria. Participants were excluded if they did not meet criteria for ADHD; if they were suffering from a psychopathology other than ADHD such as schizophrenia, clinical depression or anxiety, autism spectrum disorder, or addiction to substances; or if they were using non-ADHD medications effecting neural function. This procedure resulted in a group of 54 participants (See Diagram 1) who were divided into three experimental groups: (1) The CPAT group. (2) The MBSR group. (3) The passive control group. According to G*Power analysis, to obtain statistical power of (1−β) = .80 to detect medium-sized effects (f = 0.25) in ANOVAs with repeated measures, within-between interaction of three groups and two assessments 42 participants were needed (analysis conducted via G*Power 3, Faul et al., 2007). Participants who could not commit to consistently arrive on campus to attend the intervention programs were assigned to the passive control group. Remaining participants were assigned to one of the two intervention groups through a semi-randomized lottery controlling for their baseline attentional performance (details on the procedure can be found in Section 1 in the Supplemental Materials). Five participants (three from the CPAT group and two from the MBSR group) withdrew from the intervention groups before the interventions started. Two of them were added to the passive control group (the other three left the experiment altogether), yielding a final sample of 16 participants in the CPAT group (eight males; mean age = 29.9; SD = 5), 17 participants in the MBSR group (seven males; mean age = 27.3 SD = 5.1), and 18 participants in the passive control group (seven males; mean age = 26.1; SD = 3.9; F_(2,48) = 2.87, p = .06; χ²_Gender = 0.63, p = .89). Two participants (one from the CPAT group and one from the passive control group) failed to complete the post-test session, yielding a sample of 15 and 17 participants in the CPAT and in the passive control group, respectively. Four other participants failed to complete the follow-up session thus leaving 13 participants in the CPAT group and 15 in the MBSR group.

All the participants were higher education students at the time of the study or in the past. The present experimental protocol was approved by the Tel-Aviv University’s Ethic Committee [university name removed for blind review]”. All participants filled a written informed consent form describing the study and were aware of the existence of two intervention groups in the study. They were told that both intervention protocols are based on previous studies investigating non-pharmacological treatments for ADHD.

Assessment Tasks

The assessment tasks included objective measures (attention functioning measured by neuropsychological tasks, reading patterns based on eye-movement measures) and self-report subjective measures (ADHD symptoms, learning performance, and quality of life).

Attentional Assessment

The attentional tasks were used to assess different aspects of attention functioning. Each attention task starts with a short practice and then it lasts approximately 12 min.

The Conjunctive Continuous Performance Task (CCPT) was used to assess sustained attention. Participants were presented with a long series of stimuli and were requested to respond to a single pre-specified target, defined by a conjunction of color and shape (a red square), while withholding responses to all other stimuli. This task is described in detail in Shalev et al. (2011). For the CCPT, the dependent measures were Reaction Time (RT) to the target and the Standard Deviation (SD) of the RT to the target that represents performance consistency.

The Conjunctive Visual Search task (CVST) was administered to evaluate selective-spatial attention. Participants were instructed to search for a target stimulus, defined by a conjunction of color and shape (a blue square), appearing among distractors. The set-size varied between 4, 8, 16, and 32 items creating different levels of attentional load. Detailed description of this task can be found in Segal et al. (2015) and Shalev et al. (2016). For the CVST task, the dependent measure was inverse efficiency score (IES), computed as mean RT divided by accuracy rates in the different set-size (8, 16, or 32). Inverse efficiency scores are used to incorporate accuracy and reaction times within a single measure (Kolodny et al., 2017; for the detailed calculation see Lukov et al., 2014 or Segal et al., 2015).

A Go/No-Go task was administered to assess executive attention—response inhibition. The task is similar to the CCPT. While in the CCPT task, the target frequency is low (30% of the trials), in the Go/No-Go task the frequency is high (70% of the trials). See the detailed description in Kolodny et al. (2022). For the Go/No-Go, the dependent measure was the commission errors rate.

In all tasks, mean reaction time was calculated in each condition for correct responses only, after excluding extremely short (<150 ms) or long (>3,000 ms) responses and responses deviating more than two SDs (for the CVST) or three SDs (for the CCPT and the Go/No-go) from the participants’ mean RT in the respective condition.

Reading Patterns Assessment

Reading patterns were evaluated with an ecological academic task for higher education students. Participants were asked to read a segment of a scientific paper and answer reading comprehension questions.

Apparatus

An EyeLink 1000 eye tracker (SR research, Ottawa, Ontario, Canada) monitored the gaze location of the participant’s left eye during reading. The eye tracker had a spatial resolution of 30 arcmin and a 1,000-Hz sampling rate. Participants viewed the text on a 768 × 1,024 computer monitor located 65 cm from their eyes. Chin and forehead rests were used to minimize head movements. The eye tracker was calibrated only once, at the beginning of the session, to enable the participants to read without intrusions that may artificially reduce inattentive reading. For administering the experiment and data collection, we used SR Research’s Experiment Builder and SR Research’s Data Viewer.

Materials

Two texts were taken from the same scientific article: “All about children and numbers: How the standardized test and achievement policy ruins our schools” (Goldshmidt, 2011; in Hebrew) to minimize differences in style and difficulty level. Half of the participants read one text in the pre-test and the other half read the other text. At post-test the texts were swapped so that all participants read the text they did not read in the pre-test. In the follow-up session, participants read the same text that they read in pre-test. Each text consisted of about 1,330 words and was divided to 18 passages of five to eight lines each. Each passage was displayed on a computer monitor upon clicking on the “Enter” key by the participant. Participants were instructed to read the text attentively in order to comprehend it since they will be asked questions about the text afterward. When completing the reading they were asked to answer five simple comprehension questions.

Eye Movement Measures

To assess reading patterns, we calculated the following measures from the data obtained by the eye tracker:

Mean dwell time per word: We computed an overall proportional measure by dividing the sum of all fixation durations (including those occurring as a result of inter-word regressions) by the number of all fixated words. We did not remove fixations that were shorter than 80 ms or longer than 1,000 ms as is usually done in eye tracking experiments due to the possibility of erratic nature of eye movements during unattended reading (Reichle et al., 2010).

Proportion of words that were fixated more than twice: These are the words that the reader returned to read more than twice by moving the eyes backward (regression saccades). We assumed that the reader returned to read these words repeatedly because the first reading was not attentive enough for comprehending. We computed the ratio between the number of words that were re-fixated more than twice and the number of words in the article.

ADHD Symptoms

To evaluate ADHD behavioral symptoms, participants filled the Hebrew version of the Adult ADHD Self-Report Scale (ASRS; Adler et al., 2006) in each testing session (pre-test, post-test, and follow-up). The questionnaire is based on the DSM-IV list of symptoms. Higher total score represents high level of symptoms.

Self-Report of Learning Performance

The learning performance questionnaire (Arbel & Shalev-Mevorach, 2008) consists of 50 items related to learning. From organizing learning materials, through reading and understanding texts, studying for exams, to preparing assignments and behavior during exams. Participants were asked to evaluate themselves and mark how well the statement in each item describes them. High scores reflect self-report of poor learning performance.

Quality of Life

Participants filled the Adult ADHD Quality of Life Questionnaire (AAQoL; Brod et al., 2006). The questionnaire consists of 29 items assessing four areas: productivity, daily activities, mental health, and relationships. Participants were asked to evaluate the degree or the frequency their ADHD has affected their life over the past 2 weeks. High scores reflect low quality of life.

Interventions

CPAT Training Group

For the attention intervention, we used the training software CPAT (Shalev et al., 2007). The software is based on compatibility and expansion of well-studied attention tasks that are known as activating the different attention functions. The software consists of four types of computerized exercises, each aimed at different function of attention: sustained-, selective-spatial-, orienting-, and executive-attention. Each exercise has a range of levels of difficulty. The participant advancements between levels is individually adjusted according to improvement of RT and accuracy relative to one’s own base line for each exercise separately. A full description of the training tasks could be found in Kolodny et al. (2017). The attentional cognitive intervention was held in a format of a workshop with eight weekly sessions of 2 hr each and a final summary session. The sessions were conducted by specially trained graduate students in the Department of Special Education and School Counseling. Participants trained in groups of approximately six participants. Each graduate student supervised two to three trainees. The supervision included the confirmation of instruction comprehension, encouragement during the training sessions and sending reminders before each training session. During each training session participants completed between 15 and 25 exercises. The final summary session lasted approximately 3 hr during which the participants shared their experiences, difficulties and achievements, as well as received recommendations for effective coping with academic and other daily tasks in the future.

MBSR Training Group

The MBSR workshop is a structured group program developed by Kabat-Zinn (1990). The workshop consisted of eight weekly sessions, 2 hr each, during which the participants learned the following methods: The force of breathing, body scan, body postures, balance and flexibility, awareness in daily life, and coping with pain. The content of the sessions was designed to provide participants with a collection of techniques for practicing and understanding how to practice Mindfulness. The workshop was conducted by an experienced instructor. Each session started and ended with an open conversation encouraging the participants to express the emotions that arise following the practice, the difficulties in the practice and the way they can apply the techniques in daily life, thus becoming more aware of their feelings and thoughts, more focused on their tasks and behaviors. During each session, the participants practiced short (5–20 min) Mindfulness exercises. Then, the participants shared their experience, and received personal feedback from the instructor. Afterward, participants learned and practiced new Mindfulness exercises. Participants were encouraged to practice daily at home. A graduate student was appointed for every two to three participants to be in touch with them between sessions and encourage them to practice, but the extent and frequency of the actual practice was not monitored. At the end of the eight sessions, a ninth retreat session of about 5 hr was held, during which the techniques learned during the workshop were practiced. The workshop was adapted to ADHD in the following ways: (a) The duration of the mindfulness exercises at the beginning was short (5 min) and gradually increased over the course of the program to 20 min which was shorter than in other MBSR programs (typically 45 min is recommended); (b) mindful awareness in daily living was emphasized; (c) ADHD related difficulties in daily life and in practice were discussed and it was explained how mindful awareness and mindfulness practice can overcome or attenuate these difficulties.

Intervention Compliance

All the CPAT group participants except one who participated in five out of eight sessions, completed all training sessions. In the MBSR workshop, seven participants missed out one session and one participant participated in five out of eight sessions.

Procedure

Participants using psychostimulants abstained from medication on the day they had an assessment session. The pre-test tasks were administered over two sessions in the lab. Pre-test measurements included three computerized attention tasks (CCPT assessing sustained attention, CVST assessing selective-spatial attention, and Go/No-Go assessing response inhibition), recording of reading patterns using eye-tracker, self-reported ADHD symptoms, learning performance, and quality of life. Post-test assessments were held over two sessions within 2-weeks following the final training session except one participant who completed the post-test 4 weeks after the intervention ended. The post-test assessments were the same as the pre-test assessments. Follow-up assessments were conducted in the intervention groups 4 months after the end of the interventions. The follow-up tests were administered in a single session and therefore included only two computerized attention tasks (CCPT and CVST), self-reported ADHD symptoms, recording of reading patterns using eye-tracker, and quality of life. Participants also received in the follow-up session a written report regarding their attention functioning before and following the intervention along with recommendations derived from their attentional profiles. The results of the passive control group were used to examine the effect of the re-test of the various assessment tasks. Participants in this group performed the pre-test and post-test similar to the participants of the two intervention groups (while maintaining a similar period of time between the two assessments) but did not perform the follow-up assessment. Participants in the passive group were paid for their participation in the assessment sessions.

Missing Data

Due to technical issues the following data of participants from the passive control group are missing: pre-test of the sustained attention task (the CCPT) of three participants, pre-test of the selective-spatial attention task (the CVST) of one participant, and pre-test of the response inhibition task (the Go/No-go task) of two participants. In addition, the post-test data of the sustained attention task (the CCPT) and data of the selective-spatial attention task (the CVST) of one participant from the CPAT group is missing.

Data Analysis and Design

To evaluate the effects of the interventions on attention functioning two sets of mixed repeated-measures ANOVAs were conducted on the dependent measures of the attention tasks. One set of ANOVAs compared the attention functioning of all the participants before and after the interventions and included the following independent variables: Time (pre-test, post-test) as repeated within subject factor and Group (CPAT, MBSR, passive control) as a between-subjects factor. Another set of ANOVAs compared the attention functioning of participants in the two intervention groups before the intervention and in the follow-up assessment. These analyses included the following independent variables: Time (pre-test, follow-up) as repeated within subject factor and Group (CPAT, MBSR) as a between-subjects factor.

To assess the effects of the interventions on everyday functioning (i.e., far-transfer effects), mixed repeated-measures ANOVAs were performed for the following dependent variables: reading eye-movement measures, ADHD symptoms, learning performance, and quality of life questionnaires scores with Time as a within subject factor and Group as a between-subject factor. Follow-up tests were conducted when a significant interaction between Time and Group was obtained. Due to the small sample sizes, follow-up tests were also administered when a marginally significant interaction between Time and Group was recorded. Bonferroni-adjusted pairwise comparisons (using a .05 significance level) were used for multiple comparisons correction. In addition, when a significant main effect of Time was obtained without interaction between Time and Group paired t-tests were conducted in each group.

To explore the effects of the interventions in the individual differences level and to correlate between differences in attention functioning, ADHD symptoms, and reading eye-movement measures, change scores were computed for each participant in each attention task, reading measures, and ADHD questionnaire. Change scores were computed as the differences in performance between the post-test and pre-test corresponding measurements. Bivariate Spearman correlations between the change scores of attention tasks, ADHD symptoms, and reading measures were performed. In addition, to calculate the number of responders, that is, patients who improved ≥30% in ADHD symptoms (e.g., Wilens et al., 1999), the percent of ADHD symptom reduction was calculated as follows: for each participant the change score was divided by the pre-test score (The pre-test score was calculated as the total score of the questionnaire in the pre-test phase minus 18 points which is the base-line score of the Hebrew ASRS questionnaire—thus, one cannot have a score lower than 18).

To examine the effects of the interventions in the individual level we created an Overall Improvement Index that reflected the number of improved outcome measures. To do so, we calculated z-scores for each change score based on the corresponding mean and standard deviation of the passive control group change score. Z-change score of one or above was considered as an improvement. We then accumulated the number of improvements across all outcome measures per each participant, which we call the Overall Improvement Index (OII). We differentiated between four levels of OII: none (0 improvements), small (improvement in 1–2 outcome measures), medium (improvement in 3 outcome measures), and large (improvement in at least 4 outcome measures).

Effect sizes were estimated by Partial eta squared (η_p²) for ANOVAs, and by Cohen’s d for t tests. When calculated for paired t-tests, Cohen’s d was calculated with the correlation between the pre- and post-test considered as suggested by Morris and DeShon (2002). Data were analyzed using IBM SPSS Statistics Version 27.

Results

Post-Test Versus Pre-Test Effects

Attention Functioning: Near-Transfer Effects of Objective Measures

Descriptive statistics and ANOVAs statistics of the assessment tasks and questionnaires are detailed in Table 1.

Table 1.

Attention Performance, Reading Patterns Measures, ASRS ADHD Symptoms Questionnaire Scores, Learning Performance, and Quality of Life Questionnaire Scores by Intervention Groups and Testing Session (Pre-Test and Post-Test).

	Passive control		CPAT		MBSR		Statistics
	Mean	STD	Mean	STD	Mean	STD	F	Sig.	ηp 2
Sustained attention	(n = 16)		(n = 14)		(n = 17)
RT (ms)
Pre-test	497.495	60.367	460.547	56.129	485.757	95.184
Post-test	502.972	55.151	418.739	42.098	461.688	96.122	9.007	.004	.170
Interaction group × Time							4.117	.023	.158
SD of RT (ms)
Pre-test	111.085	51.338	87.265	18.804	103.304	57.847
Post-test	100.665	47.941	64.311	19.003	92.973	59.329	7.392	.009	.144
Interaction group × Time							0.580	.564	.026
Selective-spatial attention	(n = 18)		(n = 15)		(n = 17)
Pre-test	11.709	7.390	12.775	13.727	13.593	7.322
Post-test	10.615	8.324	8.105	5.095	10.001	7.506	8.424	.006	.152
Interaction group × Time							0.985	.381	.040
Response inhibition (% commissions)	(n = 17)		(n = 14)		(n = 17)
Pre-test	0.071	0.052	0.066	0.050	0.069	0.081
Post-test	0.059	0.058	0.030	0.015	0.053	0.047	7.020	.011	.135
Interaction group × Time							0.821	.446	.035
ADHD symptoms	(n = 17)		(n = 15)		(n = 17)
Pre-test	64.940	10.963	62.00	9.849	63.350	10.136
Post-test	62.590	11.292	52.670	11.349	58.350	10.271	27.302	<.001	.372
Interaction group × Time							3.548	.037	.134
Dwell time per word (ms)	(n = 17)		(n = 14)		(n = 17)
Pre-test	515.587	87.844	552.281	153.927	536.214	255.960
Post-test	474.644	122.740	451.32	107.418	455.364	168.149	13.310	<.001	.228
Interaction group × Time							0.747	.480	.032
% re-fixated words	(n = 17)		(n = 14)		(n = 17)
Pre-test	0.129	0.046	0.187	0.109	0.149	0.145
Post-test	0.120	0.072	0.101	0.060	0.086	0.065	16.118	<.001	.264
Interaction group × Time							3.067	.056	.120
Learning performance	(n = 15)		(n = 14)		(n = 15)
Pre-test	130.600	19.913	133.860	24.391	133.870	21.344
Post-test	132.470	19.694	116.640	22.058	124.600	23.591	9.130	.004	.182
Interaction group × Time							4.124	.023	.167
Quality of life	(n = 15)		(n = 14)		(n = 17)
Pre-test	81.130	18.458	80.640	16.378	85.590	17.000
Post-test	81.330	19.100	74.070	15.464	74.710	17.496	6.998	.011	.140
Interaction group × Time							2.283	.141	.096

To test the effects of the interventions on sustained attention, repeated measures ANOVAs on RT and SD of RT in the CCPT were conducted (see Figure 1, panels A and B). The ANOVA on RT revealed a significant interaction between Time and Group on RT, F_{(2, 44)} = 4.12, p = .023, η_p² = .16 and a significant main effect of Time, F_{(1, 44)} = 9.01, p = .004, η_p² = .17, suggesting faster RTs at post-test, after the interventions, compared to pre-test. To interpret the interaction, we compared the change scores of RTs by group. A one way ANOVA showed a significant difference between the groups, F_{(2, 44)} = 4.12, p < .023, η_p² = .16. A Bonferroni-adjusted pairwise comparisons revealed that participants in the CPAT group improved significantly (M = 41.8, SD = 37.58) more than participants in the passive control group (M = −5.48, SD = 59.24; p = .022, d = 0.95), but there was no significant difference between the MBSR group (M = 24.07, SD = 36.62) and the passive control (p = .21, d = 0.6) and CPAT (p = .87, d = 0.48) groups. ANOVA on SD of RT yielded a significant main effect of Time, F_{(1, 44)} = 7.39, p = .009, η_p² = .14, but no interaction of Time × Group F_{(2, 44)} = 0.58, p = .56, η_p² = .03. Nonetheless, paired samples t-tests within each group between the pre-test SD of RT and the post-test SD of RT found a significant difference with large effect size only in the CPAT group (CPAT: t₍₁₃₎ = 4.54, p = .001, d = 1.22; MBSR group: t₍₁₆₎ = 0.95, p = .36, d = 0.23; passive control: t₍₁₅₎ = 1.09, p = .29, d = 0.26), suggesting a significant reduction in SD of RT compared to the initial performance in the CPAT group only.

Diagram 1.

A consort flow.

To assess the effects of the interventions on selective-spatial attention, a repeated measures ANOVA was conducted on the inverse efficiency score (IES) of the CVST revealing a main effect of Time (F_{(1, 47)} = 8.42, p = .006, η_p² = .15) with no Time × Group interaction (F_{(2, 47)} = 0.99, p = .38, η_p² = .04) (see Figure 1, panel C) . Paired samples t-tests within each group between the pre-test score and the post-test score yielded a significant difference with a moderate effect size in the MBSR group, a marginally significant effect in the CPAT group and no effect in the passive control group (MBSR group: t₍₁₆₎ = 2.17, p = .046, d = 0.53; CPAT group: t₍₁₄₎ = 1.8, p = .09, d = 0.56; passive group: t₍₁₇₎ = 0.82, p = .49, d = 0.21). These results suggest that participants in the MBSR group improved their selective-spatial attention compared to their initial performance. Notice that, the effect sizes in both intervention groups were moderate.

Figure 1.

Performance in the attentional tasks and ADHD symptoms by group and time of assessment. Error bars denote standard errors. (A) RT of sustained attention (ms). (B) SD of RT of sustained attention. (C) Inverse efficiency score of spatial selective attention. (D) Commission rate of executive attention—response inhibition (%). (E) ASRS score-ADHD symptoms.

Finally, to assess the effects of the interventions on executive attention—response inhibition, a repeated measures ANOVA was conducted on the commission error rates in the Go/No-go task (see Figure 1, panel D). A main effect of Time was found (F_{(1, 46)} = 7.02, p = .011, η_p² = .14) with no Time × Group interaction (F_{(2, 45)} = 0.82, p = .45, η_p² = .04). Further paired samples t-tests within each group between the pre-test and the post-test of the commission error rate yielded a significant difference with moderate effect size in the CPAT group (t₍₁₃₎ = 2.78, p = .015, d = 0.59). While no significant differences were obtained in the other two groups (MBSR group: t₍₁₆₎ = 0.92, p = .37, d = 0.19; passive group: t₍₁₆₎ = 1.25, p = .23, d = 0.33), suggesting that the CPAT group reduced their commission error rate compared to their initial performance.

ADHD Symptoms: Far Transfer Effects of Subjective Measures

To assess the effects of the interventions on ADHD symptoms a repeated measures ANOVA with Time and Group as independent variables was conducted on the ASRS questionnaire total score. A significant interaction between Time and Group was found (F_{(2, 46)} = 3.55, p = .037, η_p² = .13), as well as a significant main effect of Time (F_{(1, 46)} = 27.3, p < .001, η_p² = .37; Figure 1, panel E). These effects indicate diminished ADHD symptoms after the interventions. To interpret the interaction, we compared the change scores in ADHD symptoms severity of each group. Bonferroni-adjusted pairwise comparisons suggested that participants in the CPAT group reduced the severity of ADHD symptoms significantly (M = 9.33, SD = 9.69) compare to participants in the passive control group (M = 2.35, SD = 4.12; p = .033, d = 0.94), pinpointing a large effect size. Yet, no significant difference was found between the MBSR group and the passive control group (M = 5.00, SD = 7.75; p = .92, d = 0.43). Note however, that the above insignificant difference has a low-medium effect size.

Does Improvement in Attention Functioning Predict Improvement in ADHD Symptoms?

To examine whether the changes in attention functioning associate with the change in self-reported ADHD symptoms we computed the correlations between the change scores of the attention measures and the change scores of the ASRS questionnaire among participants in the two treatment groupsSignificant correlations were found between the change score in sustained attention as measured by SD of RT and the change score in ADHD symptoms as measured by the ASRS total score (r = .51, p = .003, n = 31) as well as between the change score in response inhibition and the change score in ADHD symptoms (r = .49, p = .005, n = 31).

Linear regression with the significant variables of the correlation found that 30.7% of the variance in the ADHD symptoms change score is explained by the change scores in sustained attention and response inhibition (R² = .307; F_{(2, 27)} = 5.98, p = .007). Yet, according to the regression model, only the change score in sustained attention significantly predicts the improvement in ADHD symptoms (B = 0.114, p = .011).

Reading Patterns: Far Transfer Effects of Objective Measures

Other far transfer effect of the interventions was assessed using an ecological task of reading a scientific paper, a substantial skill higher education students are using throughout their studies and beyond. Reading patterns during the above task were estimated by two eye-movement measures that reflect reading effectiveness: The mean dwell time per word (Figure 2, panel A) and the proportion of words that were fixated more than twice (Figure 2, panel B) (see Method). A repeated measures ANOVA on the mean dwell time per word yielded a main effect of Time (F_{(1, 45)} = 13.31, p = .001, η_p² = .23), suggesting that the dwell time per word was shortened in the post-test relative to the pre-test. There was no interaction of Time × Group (F_{(2, 45)} = 0.75, p = .48, η_p² = .03). Nonetheless, paired samples t-tests within each group between the pre-test and the post-test values yielded a significant difference with large effect size in the CPAT group (t₍₁₃₎ = 3.72, p = .003, d = 0.93). In addition, a marginally significant difference with small effect size was obtained in the MBSR group (t₍₁₆₎ = 1.75, p = .09, d = 0.39) and a non-significant difference with a small-medium effect size was recorded in the passive control group (t₍₁₆₎ = 1.62, p = .13, d = 0.49). A marginally significant interaction between Time and Group was obtained in the repeated measures ANOVA on the proportion of words that were fixated more than twice (F_{(2, 45)} = 3.07, p = .056, η_p² = .12) along with a significant main effect of Time indicating less re-fixating words after the interventions (F_{(1, 45)} = 16.12, p < .001, η_p² = .26). To find out if there was a difference in the efficiency of reading as indicated by less repeatedly read words between each of the intervention groups and the passive control group, we used the change score which represents the difference between pre-test and post-test of the proportion of words that were fixated more than twice for each participant. The change score was submitted to univariate ANOVA as the dependent variable for each intervention group and the passive control group separately. A significant difference between the CPAT gains compared to the passive group (F_{(1, 29)} = 6.95, p = .013, η_p² = .19) was found. The MBSR group and the passive group comparison was only marginally significant (F_{(1, 32)} = 3.64, p = .07, η_p² = .10), yet the intervention groups (CPAT and MBSR) did not differ in their gains (F_{(1, 29)} = 0.04, p = .84, η_p² = .01).

Figure 2.

Reading patterns by groups and time of assessment. Error bars denote standard errors. (A) Reading efficiency—dwell time per word. (B) Reading efficiency—percentage words that were fixated more than twice.

Does Change in Attention Functioning Predicts Change in Reading Efficiency?

Spearman correlations between the change scores of the attention measures and the change scores of the two reading eye-movement measures yielded marginally significant correlations between the improvement in response inhibition and the reduction in dwell time per word and the reduction in re-fixating words (r = 32, p = .08, n = 30, r = .35, p = .06, n = 30, respectively).

Self-Report of Learning Performance: Far Transfer Effect of Subjective Measure

To estimate the transfer effect of the interventions to perceived learning performance, a repeated measures ANOVA on the total score in the self-report learning performance questionnaire was conducted with Time and Group as independent variables yielding a significant interaction between Time and Group (F_{(2, 41)} = 4.12, p = .023, η_p² = .17), along with a significant main effect of Time (F_{(1, 41)} = 9.13, p = .004, η_p² = .18). To interpret the above interaction, we compared the change scores in the learning performance questionnaire of each group. Bonferroni-adjusted pairwise comparisons suggested that participants in the CPAT group improved significantly more than participants in the passive control group (CPAT group: M = 17.21, SD = 24.9; passive group: M = −1.87, SD = 12.84; p = .02, d = 0.93). But there was no significant difference between the MBSR group and the passive control group (MBSR group: M = 9.27, SD = 4.65; p = .29, d = 0.82). Note however, that this insignificant difference actually yielded a large effect size, suggesting that both intervention groups reported significant improvements in learning performance which was absent in the passive control group. The improvement in self-reported learning performance did not correlate with the gain in attentional performance.

Self-Report of Quality of Life: Far Transfer Effect of Subjective Measure

To estimate the transfer effect of interventions to the participants’ quality of life, a repeated measures ANOVA with the total score in the quality of life questionnaire as the dependent variable and Time and Group as independent variables yielded a significant main effect of Time (F_{(1, 43)} = 6.99, p = .011, η_p² = .14). The interaction between Time and Group did not reach significance (F_{(2, 43)} = 2.28, p = .114, η_p² = .096). However, since there was a trend of improvement in both intervention groups compared to the passive control group where the score remained almost the same in the two time points, we further conducted two separate repeated measures ANOVAs to compare each intervention group with the passive control group. The comparison with the MBSR group yielded a significant Time by Group interaction (F_{(1, 30)} = 5.25, p = .029, η_p² = .15), in addition to a significant main effect of Time (F_{(1, 30)} = 0.88, p = .035, η_p² = .14). There was neither significant main effect (F_{(1, 27)} = 1.74, p = .19, η_p² = .06) nor significant interaction (F_{(1, 27)} = 1.96, p = .17, η_p² = .07) in the comparison with the CPAT group. Taken together, these results suggest that the improvement in quality of life was unique to the MBSR group.

Follow-Up Assessment

To assess whether the outcomes of the interventions are long lasting, participants in the intervention groups completed a third assessment approximately 4 months after the end of the interventions—a follow-up session. Preserved improvements were found in attention functioning, symptoms of ADHD, reading patterns, and quality of life. The detailed results of these analyses are reported in Table 2 and in section 2 of the Supplemental Materials. The CPAT group showed improvement in all measures of sustained attention and the MBSR group showed improvement in selective spatial attention and in the RT component of the sustained attention task, compared to their initial performance before the intervention. Both groups showed improvement at follow-up in reading measures compared to their initial performance before the intervention. The MBSR group showed significant improvement, while the CPAT group showed marginally significant improvement of ADHD symptoms compared to their initial measurements before the intervention.

Table 2.

Attention Performance, Reading Patterns Measures, ASRS ADHD Symptoms Questionnaire Scores, and Quality of Life Questionnaire Scores by Intervention Groups and Testing Session (Pre-Test and Follow-Up).

	CPAT (n = 13)		MBSR (n = 15)		Statistics
	Mean	STD	Mean	STD	F	Sig.	ηp 2
Sustained attention RT (ms)
Pre-test	457.796	57.897	464.261	50.299
Follow-up	423.634	39.467	438.567	41.890	15.940	<.001	.380
Interaction group × Time					0.319	.577	.012
Sustained attention-SD RT
Pre-test	86.785	20.573	87.154	34.218
Follow-up	50.095	18.593	70.219	23.399	17.546	<.001	.403
Interaction group × Time					1.165	.290	.043
Selective-spatial attention
Pre-test	13.33	14.74	12.75	6.67
Follow-up	8.34	4.58	8.98	5.83	6.20	.02	.19
Interaction group × Time					0.12	.73	.01
Dwell time per word (ms)
Pre-test	548.680	151.529	560.529	263.236
Follow-up	438.132	114.201	447.617	163.486	24.583	<.001	.496
Interaction group × Time					0.003	.959	.000
% words that were fixated more than twice
Pre-test	0.187	0.113	0.159	0.152
Follow-up	0.088	0.060	0.080	0.081	28.206	<.001	.530
Interaction group × Time					0.347	.561	.014
ADHD symptoms
Pre-test	62.150	9.433	61.600	8.895
Follow-up	56.850	8.464	55.330	9.514	11.904	.002	.314
Interaction group × Time					0.082	.777	.003
Quality of life
Pre-test	78.150	14.023	83.930	17.618
Follow-up	73.690	15.950	78.070	14.668	3.120	.090	.111
Interaction group × Time					0.057	.813	.002

Analyses of Individual Level Effects

Previous studies revealed significant individual variance in regard to cognitive intervention efficiency (Jaeggi et al., 2011, 2014). Thus, to complement the group analyses presented above we evaluated the outcomes of the interventions per each individual. Figure 3 presents the individual change scores in each outcome measure by group. Specifically, in the CPAT group 33.33% of participants were responders (i.e., had 30% or more of ADHD symptom reduction) while only 11.76% of participants in the MBSR group were responders (i.e., had 30% or more symptom reduction) and none of the participants in the passive control group was classified as a responder. for comparing frequencies of responders and non-responders between the intervention groups and the passive control group yielded significant difference between the CPAT and the passive control group (Fisher’s Exact Test two-sided; p = .018). No significant difference was found between the MBSR and the passive control groups (Fisher’s Exact Test two-sided; p = .485), nor between the CPAT and MBSR groups (Fisher’s Exact Test two-sided; p = .209). Figure 4 shows the proportion of the different levels of Overall Improvement Index (OII) by group. As can be seen, more participants in the CPAT group gained four and above improvements than in the MBSR and the passive control group. for comparing the OII distribution between the intervention groups and the passive control group yielded significant difference between the CPAT and the passive control group ( No significant difference was found between the MBSR and the passive groups ( nor between the CPAT and MBSR groups (.

Figure 3.

Gain scores of individual participants. Gain scores were calculated as differences in performance of attention tasks, ADHD symptoms evaluation, reading patterns, learning performance, and quality of life assessment between pre-test and post-test. Each bar represents one participant; blue bars represent participants in the CPAT group; green bars represent participants in the MBSR group; orange bars represent participants in the passive group. The dashed line demonstrates the cutoff of 30% reduction in ADHD symptoms.

Figure 4.

The proportion of the different levels of Overall Improvement Index (OII) by group. The blue bar represents proportion of participants with no improvement, the orange bar represents proportion of participants with improvements in one to two outcome measures, the gray bar represents proportion of participants with improvements in three outcome measures and the dark yellow bar represents proportion of participants with improvement in at least four outcome measures.

Discussion

The present study has investigated the effects of CPAT, a theory driven computerized attention training program implementing the direct process approach, on higher education students with ADHD. Cognitive training based on the direct process approach was studied during the last decade as an alternative or as a potentially additional component to the common pharmacological treatment for ADHD. One of the major challenges of cognitive training research is to demonstrate far transfer of the cognitive training outcomes to daily functioning. Based on Shalev et al. (2007), our main hypothesis was that cognitive training will improve the performance of the trained attention functions and most importantly, the improvement will be transferred to daily activities. Thus, we included objective measures to assess reading efficiency, as well as subjective measures of ADHD symptoms, and self-report of learning performance and quality of life. All of these outcome measures are considered far transfer. To check if the cognitive training can produce unique outcome effects we included in this study a group that practiced MBSR as another training program. To isolate the effect of re-test of the assessment tasks, we included a passive control group that did not undergo any intervention in the study. It should be noted that in the present study the CPAT sessions were administered as group sessions and this is an innovative approach (see Kolodny et al., 2017).

Near Transfer Effects: Attention Functioning

Our results demonstrate improvements in sustained attention and executive attention—response inhibition that were unique to the CPAT group. Whereas improvement in selective-spatial attention was obtained only for the MBSR group. Jensen et al. (2012) also found that the MBSR group improved significantly in selective attention measured by the d2 paper and pencil letter cancelation task compared to both active and incentivized control groups (but see the review of Lao et al. (2016) that found mixed evidence for improvement in selective attention). Note that sustained attention is perhaps the most promising candidate to be addressed by ADHD interventions since deficient sustained attention is the most prevalent deficit in ADHD (Avisar & Shalev, 2011) and it is related to ADHD symptoms (Dankner et al., 2017; Rapport et al., 2013).

Far Transfer Effects

ADHD Symptoms

Right after the completion of the training, ADHD symptoms severity diminished significantly compared to the passive control group only for the CPAT group. Notably, this result is different from Kolodny et al. (2017) that trained adult students with ADHD using CPAT and did not find improvement in ADHD symptoms. One possible explanation may be the difference between the improvements in attention functions obtained in these studies. In Kolodny et al. (2017) improvements in selective-spatial attention and executive attention-conflict resolution were obtained. Whereas, in the current study the sustained attention and executive attention—response inhibition were improved which may have far transfer to core ADHD symptoms. This potential explanation is supported by the significant correlation that was found in the current study between the improvements in sustained attention and response inhibition and the attenuation of ADHD symptoms. These correlations suggest that cognitive training of specific cognitive functions can be transferred to various behaviors that are not directly trained. This finding supports the theoretical approach that claims that ADHD is rooted in one or more deficits in cognitive mechanisms (Chacko et al., 2014; Tsal et al., 2005). Another difference between the present study and Kolodny et al.’s (2017) study that may contribute to the improvement in ADHD symptoms is that for the first time CPAT sessions were administered in a group format where participants could be motivated by their peers and therefore improve the quality of their training and/or the way they perceive their ADHD-related behaviors. Finally, in the present study training sessions were especially long (2 hr instead of two 1-hr session) which may have increased the demand for sustained attention.

Reading Efficiency

Another novel feature of the present intervention study is the use of eye tracker while reading text to get objective measures without interruptions, minimizing changes in reading patterns that might result from presenting probes that traditional measurements involve. The CPAT group exhibited significant shortened dwell time per word after the intervention compared to their initial performance. Based on Reichle et al. (2010) that showed that dwell times are shorter in normal reading compared to inattentive reading, the above finding suggests that participants in the CPAT group were more attentive while reading after the intervention. The CPAT participants also exhibited significant reduction in re-reading of words after the intervention. This result may be a further indication of more attentive reading in which there is less need for re-reading after the intervention. The MBSR group showed a marginally significant shortening of dwell time per word and a marginally significant reduction of re-reading after the intervention. This finding seems consistent with Mrazek et al. (2013) that reported on less wandering thoughts while reading after practicing Mindfulness assuming that these reading patterns indeed reflect reduction in wandering thoughts. These results are of great importance because in the context of this study this is considered a far transfer of cognitive training as both interventions did not include components of reading during the interventions. Furthermore, all participants were higher education students with ADHD and as such they have practiced reading comprehension quite intensively as part of their academic studies; hence the indications for reduced mindless reading as a result of the interventions should be particularly appreciated.

Learning Performance

Another outcome that the CPAT yielded was significant improvement of self-perceived learning performance. This finding also strengthens the assumption that academic difficulties experienced by people with ADHD stem at least to a certain extent from impaired attention functioning.

Quality of Life

In line with our prediction and in accordance with Bueno et al. (2015), participants in the MBSR group significantly improved their self-perceived quality of life. The improvement was significant compared to the passive control group in accordance with the review of de Vibe et al. (2017). Notably, improved self-perceived quality of life was unique to the MBSR intervention. However, this effect was not maintained at follow-up.

Do the Near and Far Transfer Effects Are Sustainable? Follow-Up Effects

Four months after the end of the interventions, both near transfer (sustained attention) and far transfer (reading eye-movement measures, ADHD symptoms) effects tested at follow-up were preserved in the CPAT group, though ADHD symptoms were only marginally reduced compare to the pre-intervention level. Interestingly, participants in the MBSR group showed mixed preservations including the emergence of significant improvement in reading eye-movement measures that was not present right after the end of the intervention. This can be attributed to a continuation of improvement in attention functioning as they have reached marginally significant gains at post-test session. It should be noted that it can also be attributed to the fact that the same text as in the pre-test session was used in the follow-up session and that might have affected their performance. In contrast, the improvement in the perceived quality of life was not preserved 4 months after the end of the intervention.

Individual Level Effects

The group analyses showed near and most importantly far transfer effects for both active groups; nevertheless taken together the outcome measures and transfer effects seemed superior for the CPAT group. Reinforcement to that comes from analyzing the change scores of the outcome measures individually. We found that most of the CPAT participants gained positive change scores in attention measures, ADHD symptoms, and perceived learning performance. Moreover, more participants in the CPAT group gained improvements in at least four outcome measures than in the MBSR and the passive control groups. In addition, third of the participants in the CPAT group showed a positive response on the ADHD symptoms outcome measure (≥30% reduction) which is considered a clinically significant improvement (Hepark et al., 2019; Mitchell et al., 2017; Zylowska et al., 2008). This provides further support that the CPAT is an effective intervention for higher education individuals with ADHD. This should be further investigated with more heterogeneous populations and larger samples. Notably, participants in the present study were higher education students and it might be that demonstrating benefits of cognitive training in adults who are already high functioning and using compensatory mechanisms to overcome their attention deficits is more difficult.

To sum up, one of the most important issues regarding cognitive training is whether the outcomes generalize and can be transferred to other domains. In the current study the CPAT produced far transfer effects to several domains: reading, ADHD symptoms, and perceived learning difficulties. The far transfer effects that were obtained included objective measures of eye-movements as well as subjective measures of self-reported questionnaires. Of particular importance are the links that were found between gains in objective measures of attention and the reduction in ADHD symptoms as well as the decrease in eye-movements patterns indicative of mindless reading that were found when pooling data of participants from both intervention groups. Moreover, the correlations between improvement in sustained attention and reduction of ADHD symptoms’ severity support the models that assume neurocognitive deficits in ADHD (Chacko et al., 2014; Sonuga-Barke, 2005; Tsal et al., 2005).

Limitations

The main limitation of the current study is the small sample size. However, although the small sample size weakens the ability to draw definitive conclusions, we obtained significant large effects in some of the measures. Other limitations to be considered are: (1) using identical texts in the pre-intervention and follow-up assessments that might have contributed to the obtained improvements in reading at follow-up. (2) The partial blindness of the testers. At pre-test the participants were not divided into intervention groups. As a result, testers were of course blind for the group membership. However, this was not the case with the Passive control group whose participants were recruited separately. At post-test and follow-up, all testers except the first author were still blind to group membership. Although this is a limitation, it should be clarified that all assessments were done in a separate room without the tester’s presence. (3) The study is not a fully randomized controlled trial. Participants were assigned to the intervention groups by a lottery between pairs with similar attention profile. This procedure was chosen to assure, as much as possible, similarity in the attention profiles between the groups. Also, as mentioned, assignment to the Passive control group was done separately. (4) There was no control for the extent and frequency of the actual practice of mindfulness between the intervention sessions. This information would have been valuable for exploring individual effects in the MBSR group and for understanding the impact of mindfulness training duration on attention, reading, ADHD symptoms, learning, and quality of life. Nevertheless, future studies could investigate if and how practicing between sessions give the MBSR participants advantage over session-based only practice of cognitive training. (5) Both the MBSR and the CPAT programs did not include a psychoeducation component. (6) All participants in this study were higher education students. Therefore, the results are specific to this group and not to other ADHD patient groups. Considering the promising results we obtained with the CPAT training for high functioning ADHD individuals, future research should be large scaled and with broader populations. In addition, future studies should investigate the impact of CPAT when combined with pharmacological treatment and/or with other evidenced-based interventions such as MBSR. Finally, since this study is the first one (and the only one) that applied group CPAT sessions in adults it will be informative to investigate in future large scale studies whether certain individuals benefit more from individual/group training sessions.

We conclude that both interventions that were investigated in the present study—CPAT and MBSR, have beneficial effects on higher education students with ADHD, however it seems that the CPAT has an advantage over the MBSR because only CPAT participants exhibited significant near and far transfer effects compared to the passive group participants. Nevertheless, in light of the beneficial individual effects that the MBSR provided for some participants, MBSR can be considered as a potential intervention especially for individuals that the relaxing and pleasant way of the MBSR intervention better suit them.

Author Biographies

Pnina Stern (PhD) has been a PhD student of Education at Tel-Aviv University during the conduction of this research and a postdoctoral research associate at the Faculty of Education, Bar-Ilan University.

Tamar Kolodny (PhD) has been a PhD student of Cognitive Science at the Hebrew University during the conduction of this research, and is currently a postdoctoral research associate at the Psychology department at the University of Washington.

Shlomit Tsafrir (MD), is the head of the psychiatric services in the Jerusalem district at Clalit Health Services. She is a senior psychiatrist as well as Child & Adolescent Psychiatrist, and the secretary of the Israeli child & adolescent psychiatric association.

Galit Cohen (M.A. Education). has completed her MA studies in Learning Disabilities at Tel-Aviv University. She has worked as a special education teacher in an elementary school since 2014. She specializes in teaching children with attention and/or learning difficulties, Autism and mental disorders.

Lilach Shalev (PhD). is currently Head of the Department of Educational counselling and Special Education and Head of the Attention Lab. She is a professor in Cognitive Neuropsychology at Tel-Aviv University. She has developed and investigated the effects of cognitive training among individuals with attention difficulties