Effects of the Attention Training Technique on Brain Activity in Healthy University Students Assessed by EEG Source Imaging

Abstract

This study aimed to investigate the neurocognitive effects of the Attention Training Technique (ATT) on brain activity in healthy participants. The participants included 20 university students who were asked to practice ATT as a homework assignment for 20 days. The intracerebral source localization of their electroencephalogram during rest and the ATT task, which comprised selective attention, attention switching, and divided attention conditions, was evaluated by standardized low-resolution brain electromagnetic tomography. Brain activity during rest was subtracted from that during the ATT task, and that was compared before and after the homework assignment. The results for the divided attention condition indicated significantly decreased alpha 1 frequency band power in the left orbital frontal cortex (OFC) and alpha 2 power in the right inferior temporal cortex. Further, decreased alpha 1 power in the left OFC correlated with reduced subjective difficulty during the divided attention condition. One possibility is that the brain activity changed as the effect of ATT practice, although this study cannot confirm causality. Further studies are required which include a control group that would complete similar training without the ATT task.

Keywords

Attention training technique divided attention EEG source localization metacognitive therapy

Introduction

Attention Training Technique (ATT) is a component of metacognitive therapy (Wells, 1990) based on Self-Regulatory Executive Function (S-REF; Wells & Matthews, 1994, 1996) model. According to S-REF theory, the maintenance of emotional disorders is related to a particular style of thinking called cognitive attentional syndrome (CAS). The CAS consists of repetitive thinking in the form of worry and rumination, threat-oriented attention, and maladaptive self-regulatory behaviors (Wells et al., 2009). These problems diminish voluntary attentional control, which causes decreased control over negative thoughts and emotions (Normann et al., 2014). ATT is designed to train attentional control through the use of auditory attentional exercises and to promote the external processing of non-self-relevant auditory stimuli by interrupting the excessive self-focused attention associated with CAS (Wells, 2009).

ATT involves three phases for improving attentional control functions: selective attention, attention switching, and divided attention, and it seems to modify neurophysiological measures of attentional control. Knowles and Wells (2018) demonstrated the neurophysiological effects of a single dose of ATT. Healthy volunteers were randomized to either active (follow ATT instructions) or control (ignore ATT instructions) conditions. Resting state electroencephalogram (EEG) was recorded for 3 min with eyes open and 3 min with eyes closed before and after exposure to 12 min ATT training, and the power of the theta, alpha, and beta-bands were analyzed from the frontal, midline, and posterior electrodes. Participants undergoing active ATT showed a significantly elevated change in the frontoparietal alpha and beta-band activity. Furthermore, anterior and midline regions in both alpha and beta were significantly correlated in the active ATT condition, indicating greater degrees of neuronal synchrony. In contrast, limited theta-band activity was observed in both the active and control conditions. These results suggest that a single dose of ATT increases alpha and beta-oscillations in frontoparietal networks associated with top-down attentional or executive control.

Furthermore, the effects of repeated ATT practice on neurophysiological changes have been reported. According to the ATT procedure (Papageorgiou & Wells, 2000; Wells et al., 1997), a crucial component of ATT is repeated practice as daily homework. Siegle et al. (2007) studied the effects of Cognitive Control Training (CCT), including repeated practice of ATT, and measured brain activity using functional magnetic resonance imaging (fMRI). CCT consists of multiple cognitive training types, including ATT and cognitive training based on the Paced Auditory Serial Attention Task (PASAT; Gronwall, 1977), which involves continuously adding serially presented digits to improve working memory. Patients with major depression who received CCT intervention with treatment as usual, which included medication management and intensive group treatment with dialectical behavior therapy (DBT; Linehan et al., 1993), along with a control group that received their usual treatment only, had their brain activity measured before and after the intervention. Compared with the control group, after 2 weeks of CCT intervention the CCT group showed increased dorsolateral prefrontal cortex activity, related to attentional or executive control, during the digit sorting task and decreased amygdala activity, related to negative emotional processing, during the personal relevance rating task. From these results, ATT intervention may correlate with changes in brain activity, especially in areas with increased attentional or executive control and decreased negative emotional processing. However, this was observed in patients who underwent CCT, which is composed of ATT and PASAT training in addition to treatment for major depression by medication and DBT. Although this study showed the effects of repeated doses of a treatment including the ATT, the neurocognitive effect of ATT as a stand-alone intervention on healthy people has not yet been shown.

In the present study, we analyzed multichannel EEG recordings from healthy university students during the ATT task before and after 20 days of ATT practice. We used standardized low-resolution brain electromagnetic tomography (sLORETA; Pascual-Marqui, 2002) to localize brain areas active during the ATT task to determine the neurocognitive effects of ATT-alone practice. In addition, we included several subjective measures of daily attention control abilities and psychological state (e.g., the degree of difficulty) during the ATT task. By measuring subjective variables before and after the ATT practice, we were able to determine the effects of ATT on subjective measures and relationship between neurophysiological and subjective measures of attention after ATT practice.

Method

Participants

We recruited 26 students at Waseda University to participate in this experiment. The inclusion criteria were being aged 18–40 years and no personal history of psychiatric diseases. The exclusion criteria included neurological illness, traumatic brain injury, auditory or visual disorder, alcohol consumption within 12 hours, medication use within 24 hours, and a sleepless or exhausted state. We also excluded participants with depressive symptoms, as individuals with unrelated problems affecting neurocognitive function may skew results. The Japanese Beck Depression Inventory-Second Edition (BDI-II; Kojima et al., 2002) was used to exclude participants with depressive symptoms. This study was approved by the Ethics Review Committee on Research with Human Participants of Waseda University (Protocol no. 2013–278), and all participants provided written informed consent.

Procedure

After obtaining written informed consent, we explained the methods and effects of ATT on attentional control, self-focus, and depression to the participants. They then completed a medical check sheet to collect data on present and past history of neurological and psychiatric diseases, the BDI-II, and the Voluntary Attention Control Scale (VACS; Imai et al., 2015). The BDI-II is a subjective rating scale measuring depressive severity. It is comprised of 21 items, which were evaluated by using a 4-point Likert scale (0–3) for each item (e.g., “Sadness”). The criterion validity and high internal consistency (α = 0.87) of this scale were confirmed for Japanese respondents (Kojima et al., 2002). Participants with a BDI-II score above 13 points, which indicates a mild grade of depressive severity (Hiroe et al., 2005) were excluded. The VACS is a subjective rating scale that measures voluntary attention control which are trained by ATT. It was comprised of 18 items, which were evaluated using a 6-point Likert scale (1–6). The VACS has three subscales assessing subjective attentional control abilities that correspond to each component of the ATT: selective attention (e.g., “I can focus on the one thing that has to be done, even when the surroundings are noisy”), attention switching (e.g., “I can alternate between two tasks between that require focus”), and divided attention (e.g., “I can divide my attention and focus on many things at the same time”). The criterion validity, high internal consistency (selective attention: α = 0.86; attention switching: α = 0.89; divided attention: α = 0.92; total: α = 0.93), and high test-retest reliability of this scale were confirmed for Japanese respondents (Imai et al., 2015). This scale was used to measure the voluntary aspects of attention control through a pre- and post-test.

In a pre-test before ATT practice, EEG data were recorded from all participants during an EEG-recording task. This task comprised distinct components: an initial resting component; a mental arithmetic component; three ATT components: selective attention, attention switching, and divided attention; and a final resting component. After performing the EEG-recording task, we assessed the degree of difficulty (how difficult is each ATT component?), boredom (how boring is each ATT component?), and effort (how much effort do you need for each ATT component?) during each ATT component using 10-cm sliding visual analog scales, the score on which was converted to 0–100 points. Participants were interviewed regarding their thoughts during each ATT component. Finally, the homework to practice ATT was described.

Participants were instructed to practice ATT for 15 minutes comprising of 5 minutes for each component per day for 20 days prior to the post-test. ATT practice was guided by a CD that included stereophonic natural sounds (i.e., piano, a ticking clock, footsteps, a bell ringing, rain sounds, and birds singing) and prompts about how to perform each of the three ATT components. After their daily homework, participants were asked to write the date, place of ATT practice and the contents of their thoughts while performing ATT. Additionally, participants were queried about their progress and critical events, such as being unable to continue the homework during the implementation period, through the completion of online forms every week after starting the homework. None of the participants reported any critical events (e.g., accidents or hospitalization).

In the post-test, we confirmed how many times participants performed the homework assignment and their thoughts about the ATT practice. After completing the VACS questionnaire for the second time, the participants performed the same EEG-recording task used in the pre-test. Next, they completed the visual analog scales to measure difficulty, boredom, and effort regarding each ATT component and were interviewed about their thoughts during each ATT component.

Each participant was paid 2,500 Japanese yen (approximately 23 USD) after finishing the pre-test and 2,000 Japanese yen (approximately 19 USD) after finishing the post-test as compensation for their time participating in the study.

EEG-recording task

In the EEG-recording task, participants performed a task comprising the following components: (1) an initial resting component; (2) a mental arithmetic task; (3) ATT components including selective attention, attention switching, and divided attention tasks; and (4) a final resting component. They repeated this task twice with a 10 min break without EEG-recording between them (Figure 1). Participants were asked to focus on a visual fixation point presented on a PC monitor during the task.

Figure 1.

Task and experimental design.

(1) Initial resting component (3.5 min): This eyes-open resting component was recorded at the beginning of the session. EEG data from the last 3 minutes of the resting task were analyzed.

(2) Mental arithmetic component (3 min): Participants were asked to continue to subtract 7 from 2014. Data acquired from this task were not analyzed.

(3) ATT: In the three ATT components, participants listened to five competing sounds on headphones with their eyes open: rain, a bell, a car, a clock, and singing birds. While performing the ATT, patients were instructed to focus on the auditory stimuli and not to suppress or avoid any thoughts or feelings (Papageorgiou & Wells, 2000). The sounds were provided by the Metacognitive Therapy Institute (http://www.mct-institute.com).

(3.1) Selective attention (3 min): Participants were asked to direct their attention to a particular sound among the five competing sounds as instructed (e.g., “birds singing”). Participants were directed to another sound through instructions at 1-min intervals.

(3.2) Attention switching (3 min): Participants were asked to direct their attention to a particular sound among the five competing sounds following the same instructions as in the selective attention task, but the instructions were repeated more quickly at 10 s intervals.

(3.3) Divided attention (3 min): Participants were asked to listen to all sounds simultaneously. No further instructions were provided.

(4) Final resting component (3 min): This eyes-open resting component was similar to the initial resting component recorded at the beginning of the session. Data acquired from this task were not analyzed.

EEG recording and preprocessing

We analyzed pre- and post-test EEG data for the initial resting component and all three ATT components (selective attention, attention switching, and divided attention). Prior to analysis, we combined two sets of 3-min data of each component repeatedly acquired in the EEG-recording task. EEG data were recorded using a 64-channel Hydrocel Geodesic Sensor Net (Electrical Geodesics, Inc., Eugene, OR) and 19 electrodes (F3, Fz, F4, F7, F8, C3, C4, P3, Pz, P4, T3, T4, TP9, TP10, T5, T6, 01, O2, and Oz) according to the international 10–10 system at a 250 Hz sampling rate referenced to the Cz electrode. Impedance was maintained under 50 kΩ as per the recommendations of Electrical Geodesics, Inc. We filtered the data using a 0.3–70 Hz band-pass filter and a 50 Hz notch filter. Eyeblink artifacts were corrected using a regression method and EEG data in each channel were interpolated relative to the environmental channels. Two-second epochs were recomputed to the average reference; epochs including artifacts such as eye movement were removed. Finally, the remaining epochs (mean ± SD) were used in each component (initial resting component: pre-test, 137.55 ± 36.84 epochs; post-test, 134.20 ± 35.90 epochs; selective attention: pre-test, 131.05 ± 34.47 epochs; post-test, 129.25 ± 36.18 epochs; attention switching: pre-test, 118.00 ± 38.64 epochs; post-test 119.80 ± 40.44 epochs, divided attention: pre-test, 111.80 ± 37.80 epochs; and post-test, 122.25 ± 40.27 epochs). These processes were completed using Waveform Tools, Net Station Version 4.2 (Electrical Geodesics, Inc., Eugene, OR).

Frequency band-wise sLORETA analysis

To convert EEG data into functional tomographic images of the brain, we used sLORETA (Pascual-Marqui, 2002; available at http://www.uzh.ch/keyinst/loreta.htm). The sLORETA solution space corresponds to cortical gray matter sampled at a resolution of 5 mm, resulting in a total of 6,239 voxels. Analysis was performed separately for each of the eight statistically independent frequency bands (Kubicki et al., 1979; Niedermeyer & da Silva, 2005): delta (1.5–6 Hz), theta (6.5–8 Hz), alpha-1 (8.5–10 Hz), alpha-2 (10.5–12 Hz), beta-1 (12.5–18 Hz), beta-2 (18.5–21 Hz), beta-3 (21.5–30 Hz), and gamma (35–44 Hz). At each voxel, the sLORETA value represents the power, i.e., the squared magnitude of the computed intracerebral current density. For each participants and band, sLORETA values were normalized to a total power of 1 and log-transformed before statistical analyses (Pizzagalli et al., 2004). To assess pre- to post-test differences in brain electrical activity during each component, we used t statistics on log-transformed sLORETA images. Correction for multiple testing was performed using Statistical non-Parametric Mapping (Holmes et al., 1996). The sLORETA voxels are attributable to Brodmann areas (BA) based on their Montreal Neurological Institute (MNI) coordinates. The procedure for analyzing EEG data is fully described by Faber et al. (2012), who performed a similar sLORETA analysis with 19 electrodes.

Confirmation of ATT homework completion and ATT during EEG recording

To confirm whether participants performed their ATT homework assignments as instructed, we reviewed the log-book of their homework. To be included in the analysis, participants had to complete the homework assignments for more than 10 days. Furthermore, to confirm whether participants performed ATT as instructed during EEG recording, we gathered the participants’ thoughts on each ATT component. If participants showed a misunderstanding of each ATT component during EEG recording or created records deviating from the instruction about the homework assignments, we concluded that they did not engage as instructed and excluded their data.

Changes in subjective variables after ATT practice

To determine the effects of ATT, we used the Wilcoxon signed-rank test to compare pre- and post-test scores by analytical software (IBM SPSS software Statistics 23). We estimated the effect size based on VACS scores overall and for each subscale (selective attention, attention switching, and divided attention), as well as for visual analog scales measure of difficulty, boredom, and effort in each ATT component.

Changes in brain electrical activity after ATT practice

First, we compared pre- and post-test electrical activity during the initial resting component to obtain the effect of ATT practice on resting state. Next, in order to control for variability in brain electrical activity caused by the time course between the pre- and post-tests, the power value of the initial resting component was subtracted from that of each ATT component for each participant. The subtracted value in each ATT component and frequency band was then compared using the pre- and post-tests by paired t-test to indicate significant changes in brain activities. sLORETA tomography images with t values indicating significant differences were obtained in each ATT component and frequency band. Furthermore, we compared pre- and post-test sLORETA values of ROIs, which showed significant changes in the above-mentioned analyses, during each ATT component and the initial resting component to justify the significant changes in subtracted values. We utilized the Wilcoxon signed-rank test and calculated the effect size by analytical software (IBM SPSS software Statistics 23). Note that we only used sLORETA values in MNI coordinates that showed the maximal t value with a significant pre- to post-test difference in the analyses using the subtracted values.

Correlations between changes in brain electrical activity and changes in subjective variables

To confirm that it was ATT practice that had a significant effect on brain electrical activity, we performed Spearman’s correlation analyses by analytical software (IBM SPSS software Statistics 23). We examined the relationships between change in sLORETA value (post-test subtracted power value minus pre-test subtracted power value) in regions of interest (ROIs) that had significant differences in each ATT component and changes in subjective variables (post-test score minus pre-test score), namely VACS, each VACS subscale, and visual analog scale measures of difficulty, boredom, and effort. We only used the subtracted power value in the MNI coordinates that showed the maximal t value in each brain region in each frequency band with a significant pre- to post-test difference. Note that we matched ATT conditions between the subtracted power value of brain electrical activity and the VACS subscales and visual analog scale measures of difficulty, boredom, and effort.

Results

Subject characteristics

Of the 26 participants, two participants with BDI-II scores >13 and two others that did not complete the protocol were excluded. Following the post-test, two other participants who failed to complete the homework assignment at least 10 times were excluded from the analysis. One of these participants did not engage in the homework as instructed and wrote about the mental arithmetic task that was part of the EEG-recording task but not the homework assignment. After reviewing the participants’ recordings about the homework, we found no other evidence of incorrect performance. Thus, we analyzed data from 20 participants (mean age, 19.80 years; SD, 1.54; range, 18–22 years; 8 males) who correctly performed the homework for a mean of 18.50 days (SD, 1.93; range, 14–20).

Changes in subjective variables after ATT practice

We compared pre- and post-test VACS scores (total score and subscale scores) and visual analog scale measures of difficulty, boredom, and effort using the Wilcoxon signed-rank test and calculated the effect size. Table 1 shows median, Z value, p value, and effect size. After practicing ATT, the total VACS score was found to have significantly increased; the score for each subscale also significantly increased. The score for difficulty only significantly decreased in the divided attention component. The score for boredom only significantly decreased in the selective attention component.

Table 1.

Pre- and post-test median of scores, Z value, p value, and effect size (r) of changes in subjective variables.

	Median (pre)	Median (post)	Z	p	r
VACS	59.00	68.50	−2.75^b**	.006	.62
Selective attention	20.00	23.50	−2.06^b*	.039	.46
Attention switching	20.00	24.00	−2.32^b*	.020	.52
Divided attention	18.50	21.50	−2.47^b*	.014	.55
Difficulty (selective attention)	41.50	22.00	−1.61^a	.107	.36
Difficulty (attention switching)	44.50	29.00	−1.25^a	.211	.28
Difficulty (divided attention)	67.00	48.00	−2.40^a*	.017	.54
Boredom (selective attention)	62.00	50.00	−2.73^a**	.006	.61
Boredom (attention switching)	32.00	28.00	−.43^a	.668	.10
Boredom (divided attention)	67.00	56.50	−1.33^a	.185	.30
Effort (selective attention)	48.00	42.00	−1.70^a	.159	.38
Effort (attention switching)	50.00	40.00	−1.72^a	.089	.39
Effort (divided attention)	75.50	62.00	−1.66^a	.086	.37

Z values were calculated by positive rank (^a) and negative rank (^b). **p < 0.01 and *p < 0.05.

Changes in brain electrical activity due to ATT practice

There were no significant differences between pre- and post-test electrical activity during the initial resting component in any of the eight frequency bands. Next, a comparison of pre- to post-test changes in the subtracted value of the power of each ATT component minus that of the initial resting component revealed the only significant difference in divided attention. Compared to pre-test measures, the weaker post-test current density appeared as a large cluster in the alpha-1 and alpha-2 frequencies (Figure 2). Alpha-1 activity was significantly weaker on the left side of the inferior frontal gyrus, orbital gyrus, rectal gyrus, middle frontal gyrus, medial frontal gyrus, superior frontal gyrus, subcallosal gyrus, and anterior cingulate cortex (see Figure 2(a)). Alpha-2 activity was significantly weaker on the right side of the fusiform gyrus, inferior temporal gyrus, and middle temporal gyrus (see Figure 2(b)).

Figure 2.

Comparison of brain activity during the divided attention component before and after ATT practice. Brain regions with significantly weaker (a) alpha-1 activity and (b) alpha-2 activity by subtraction (power value of the divided attention component minus the power value of the initial resting component) from pre- to post-test during the divided attention component (p < 0.05, corrected).

Table 2 shows the location and magnitude of the brain activity of which subtracted alpha-1 or alpha-2 power value (divided attention minus initial resting) was significantly different between pre- and post-test during the divided attention component. We also compared pre- and post-test sLORETA values of the ROIs showed in Table 2 during each of the divided attention and initial resting components using the Wilcoxon signed-rank test and calculated the effect size. Alpha-1 ROIs included the left side of the inferior frontal gyrus (X = −15, Y = 35, Z = −25), orbital gyrus (X = −20, Y = 35, Z = −25), rectal gyrus (X = −10, Y = 30, Z = −20), middle frontal gyrus (X = −25, Y = 40, Z = −20), medial frontal gyrus (X = −10, Y = 30, Z = −15), superior frontal gyrus (X = −20, Y = 40, Z = −20), subcallosal gyrus (X = −10, Y = 25, Z = −15), and anterior cingulate cortex (X = −5, Y = 25, Z = −10). Alpha-2 ROIs included the right side of the fusiform gyrus (X = 50, Y = −40, Z = −15), inferior temporal gyrus (X = 50, Y = −40, Z = −20), and middle temporal gyrus (X = 55, Y = −40, Z = −15). Table 3 shows the pre- and post-test median scores, median of change in brain activity, Z value, p value and effect size during the divided attention task and initial resting task for each ROI. After practicing the ATT, the sLORETA value in alpha-2 activity on the fusiform gyrus, inferior temporal gyrus, and middle temporal gyrus significantly decreased during the divided attention component. There were no significant pre- to post-test differences in alpha-1 activity during the divided attention component or in alpha-1 and alpha-2 activity during the initial resting component.

Table 2.

The location and magnitude of the brain activity for subtracted alpha-1 or alpha-2 power value (divided attention minus initial resting) was significantly different between pre- and post-test during the divided attention component.

Frequency	MNI			t	p	BA	Structure	Number of voxels
Frequency	X	Y	Z	t	p	BA	Structure	Number of voxels
Alpha-1	−15	35	−25	−4.47	.007	11	Left inferior frontal gyrus	39
	−20	35	−25	−4.45	.007	47	Left orbital gyrus	7
	−10	30	−20	−4.39	.008	11	Left rectal gyrus	14
	−25	40	−20	−4.35	.008	11	Left middle frontal gyrus	20
	−10	30	−15	−4.29	.008	25	Left medial frontal gyrus	9
	−20	40	−20	−4.27	.011	11	Left superior frontal gyrus	4
	−10	25	−15	−4.23	.011	11	Left subcallosal gyrus	3
	−5	25	−10	−4.04	.021	32	Left anterior cingulate	5
Alpha-2	50	−40	−15	−3.82	.033	37	Right fusiform gyrus	22
	50	−40	−20	−3.81	.033	37	Right inferior temporal gyrus	10
	55	−40	−15	−3.74	.037	20	Right middle temporal gyrus	10

Table 3.

Pre- and post-test median sLORETA value, and median of change in sLORETA value, Z value, p value, and effect size (r) for the changes between pre- and post-tests during the divided attention task and initial resting task for each region of interest.

Frequency (MNI: X, Y, Z)	Divided attention						Initial resting
Frequency (MNI: X, Y, Z)	Median (pre)	Median (post)	Median (change)	Z	p	r	Median (pre)	Median (post)	Median (change)	Z	p	r
Alpha-1 (−15,35,−25)	29.00	23.40	−1.20	−.97^b	0.33	0.22	18.55	19.70	0.90	−.92^a	0.36	0.21
Alpha-1 (−20,35,−25)	31.05	25.55	−1.00	−.75^b	0.46	0.17	22.90	21.55	2.10	−.90^a	0.37	0.20
Alpha-1 (−10,30,−20)	26.50	22.20	−1.10	−.89^b	0.38	0.20	17.10	17.80	1.55	−.86^a	0.39	0.19
Alpha-1 (−25,40,−20)	29.50	23.05	−1.55	−.90^b	0.37	0.20	17.20	19.15	2.85	−1.38^a	0.17	0.31
Alpha-1 (−10,30,−15)	23.55	20.40	−1.70	−.82^b	0.41	0.18	16.00	16.35	0.19	−.60^a	0.55	0.13
Alpha-1 (−20,40,−20)	27.05	22.35	−1.20	−.41^b	0.68	0.09	16.20	17.65	2.94	−1.16^a	0.25	0.26
Alpha-1 (−10,25,−15)	25.15	21.70	−0.25	−.71^b	0.48	0.16	16.40	17.00	1.44	−.93^a	0.35	0.21
Alpha-1 (−5,25,−10)	20.90	18.90	−1.90	−.82^b	0.41	0.18	13.95	15.15	1.61	−.84^a	0.40	0.19
Alpha-2 (50,−40,−15)	41.45	26.85	−13.40	−2.13^b*	0.03	0.48	26.85	20.65	−0.55	−.11^b	0.91	0.03
Alpha-2 (50,−40,−20)	43.70	29.85	−13.75	−2.02^b*	0.04	0.45	30.35	21.65	−0.50	−.64^b	0.53	0.14
Alpha-2 (55,−40,−15)	41.95	28.25	−11.50	−2.13^b*	0.03	0.48	30.35	21.25	−1.25	−.36^b	0.72	0.08

Z values were calculated by positive rank (^a) and negative rank (^b). *p < 0.05.

Correlations between changes in brain electrical activity and changes in subjective variables

Correlations between changes in brain electrical activity in the described alpha-1 and alpha-2 ROIs and changes in subjective variables were assessed using Spearman’s correlation analysis. There were significant positive correlations between changes in difficulty in the divided attention component and the subtracted power value of alpha-1 activity in the middle frontal gyrus (Spearman’s rho (r_s) = 0.48, p = 0.03), medial frontal gyrus (r_s = 0.45, p = 0.048), rectal gyrus (r_s = 0.49, p = 0.03), subcallosal gyrus (r_s = 0.48, p = 0.03), superior frontal gyrus (r_s = 0.50, p = 0.02), and anterior cingulate cortex (r_s = 0.52, p = 0.02). The other changes in subjective variables were not significantly associated with subtracted power values of electrical activity in these brain regions.

Discussion

In this study, we investigated whether 20 days of ATT practice altered brain function and the relationship between brain electrical activity and subjective variables. The results demonstrated that only divided attention component of ATT showed significantly different activities in the alpha frequency. Alpha-1 activity was significantly weaker around the left orbital frontal gyrus, and alpha-2 activity was significantly weaker around the right inferior temporal gyrus. In addition, decreased alpha-1 frequency in some regions of the left orbital frontal gyrus was associated with decreased difficulty in the divided attention component of ATT. As this study did not include a control group, the present results might be attributed to either ATT practice or also to the time course. With these caveats, we hope that the present results are worth discussing with the understanding that differences may be attributed to ATT practice.

Changes in subjective variables after ATT practice

Total VACS score and scores for each VACS subscale significantly increased after ATT practice, which suggests that subjective attentional control may be increased through ATT practice. No significant changes were observed in difficulty scores in the selective attention and attention switching components, which may indicate that pre-test difficulty scores were already low in these components and that participants easily performed these tasks as instructed. The decreased difficulty score in the divided attention component only may be an effect of the repeated performance of ATT. The boredom score significantly decreased only in the selective attention component. Boredom is the aversive state that occurs when we are not able to successfully engage attention with internal or external information required for participating in a satisfying activity, and is related to high distraction (Eastwood et al., 2012). ATT rationale and instructions emphasize that ATT is not intended as a distraction strategy for reducing awareness of internal events, but a practice focusing attention flexibly in a particular way (Wells, 2007). The decreased boredom score in the selective attention component may relate to decreased distraction and improved attentional focus thorough continuous ATT practice. The effort score did not significantly change for any components, although there was a moderate reduction in the effect size in all components. These results may indicate that participants became more skillful or felt more comfortable performing the ATT task.

Changes in brain electrical activity after ATT practice and their correlations with changes in subjective variables

In this study, only the divided attention component showed decrease in alpha activity. Because this result was obtained by subtracted brain activity (divided attention minus initial resting), either the change from pre- to post-test in divided attention component or that in resting component could contribute to this finding. Alpha-2 frequency activity significantly decreased pre- to post-test during the divided attention component, but not the initial resting component. The alpha-1 frequency band displayed no significant changes during both the divided attention and initial resting components. According to the median pre- to post-test change in alpha-1 brain activity, activity during the divided attention component tended to decrease while the initial resting activity tended to increase compared to activity measured post-test. Both of these changes may relate to the significant subtracted brain activity in alpha-1, and we could argue that alpha-1 activity in the divided attention component decreased relative to alpha-1 activity observed during the initial resting component.

The significant alpha-1 cluster was located around the orbital frontal cortex (OFC), and the significant cluster in alpha-2 was located around the inferior temporal cortex (ITC), middle temporal cortex, and fusiform. Previous reports have shown that increased alpha signals are correlated with decreased blood oxygen level-dependent signals in multiple regions of the occipital, superior temporal, inferior frontal, and cingulate cortex during simultaneous EEG and fMRI (Goldman et al., 2002). These findings suggest that the weaker alpha-1 and alpha-2 activities in this study reflect increased brain activity in the corresponding regions.

Decreases in alpha-1 activity around the OFC were correlated with a decrease in the difficulty score during the divided attention component. This significant activation around the OFC could result from the ease with which the divided attention component was performed in the post-test. As the OFC is involved in various cognitive functions, there are multiple possibilities for the role of OFC activity during the divided attention component. The OFC receives input from the five classic sensory modalities: gustatory, olfactory, somatosensory, auditory, and visual sensory systems. It is uniquely placed to integrate sensory information and modulate behavior (Kringelbach, 2005). In the divided attention component of ATT, participants were asked to direct their attention to all sounds. Increased activity in the OFC suggests that participants were able to simultaneously listen to multiple sounds due to improvements in the processing of auditory input and integration.

The OFC is an important part of the networks involved in emotional processing. Kringelbach (2005) reported that the OFC has direct reciprocal connections with brain structures related to emotional processing, including the amygdala, cingulate cortex, insula/operculum, hypothalamus, hippocampus, striatum, and periaqueductal gray, and direct connections with dorsolateral prefrontal cortex. Functional neuroimaging studies have shown dysfunction of the OFC is associated with emotional disorder. The OFC shows decreased blood flow (Ring et al., 1994) and metabolism (Bremner et al., 2002) in patients with depression. Similarly, decreased OFC volume is associated with high anxiety traits in panic disorders (Roppongi et al., 2010). In addition, repeated ATT practice has been shown to improve psychological disorders related to emotional processing dysfunction such as major depression (Papageorgiou & Wells, 2000; Siegle et al., 2007), social phobia (Wells et al., 1997), and panic disorder (Wells, 1990). Therefore, in individuals with emotional processing dysfunction, OFC activation by assigning repeated divided attention tasks might be helpful.

Additionally, the OFC has been implicated in regulating affective responses via manipulation of the contextual evaluation of sensory events (Rolls & Grabenhorst, 2008), as well as processing reward value during cognitive modulation (Petrovic & Ingvar, 2002). According to the participants’ reflections on the components of the ATT prior to ATT practice, some attempted to hear all of the sounds in the auditory stimulus by focusing their attention on each individual sound, one at a time, successively. On the other hand, according to the participants’ reflections after ATT practice, some participants reported that they were able to determine the location of all sounds simultaneously by constructing a visual space image that they could manipulate. Wells (2009) explained that the ATT utilizes an inner metacognitive map for the spatial allocation of attention. The sounds used in the ATT are located in the surrounding space and some sounds are “potential sounds,” which appear and disappear during the session. For example, when the sounds are not heard from the far-right direction, participants are still asked to “focus on any potential sounds that might be heard in that direction.” Thus, attention is allocated “here and there” in the surrounding space, no matter whether or not the sounds in a specific location are detectable. Because the participants reported a strategy in which they constructed a visual space image after ATT practice, the metacognitive map for spatial allocation of attention may gradually develop by repeating ATT practice. In this new context, the strategy of divided attention differed, and participants might be able to reconstruct the metacognitive map for the location of all sounds in the visual space image; this strategy might be associated with OFC activation.

Consistent with the above-mentioned discussion, the decrease of alpha-2 activity after ATT practice in the ITC may involve the integration of visual information to form a mental image of an object (Girkin & Miller, 2001). When asked their thoughts about the pre-test EEG-recording task, participants said: “It was easy for me to direct my attention to bigger sounds such as a bell and a clock” and “I couldn’t simultaneously hear all of the sounds.” Post-test thoughts about the EEG-recording task included: “I felt that I made visual space images during the divided attention component, such that the birds singing was located on the top side and the bell was located on the right side” and “I felt the sounds as visual images.” Therefore, while it was easier for participants to focus their attention on louder sounds and suppress others without integration to internal images prior to ATT practice, it may become possible for them to integrate all sounds into visual images after ATT practice.

The present study demonstrated that a repeated dose of ATT increased brain activity in the OFC and ITC. Furthermore, changes of activity around the OFC were correlated with decreases in the difficulty score of the divided attention component. Knowles and Wells (2018) demonstrated that a single dose of ATT increased brain activity in frontoparietal networks, which are associated with top-down attentional or executive control. While the brain areas associated with top-down attentional or executive control may be actively used at the beginning of treatment, those areas could become less energy-consuming as participants learn how to practice ATT. Furthermore, other brain areas, such as the OFC and ITC, may be activated alternately and may contribute to changing the functions attributed to the long-term effects of ATT, including integration of sensory stimuli and emotional control, and manipulation of the contextual evaluation of sensory stimuli.

Study limitations

There are several limitations to our study. Given the small sample size, it is difficult to make generalizations from our results. Therefore, caution must be exercised in generalizing these findings. Moreover, this study consisted of only a single-arm, and a control group that did not perform ATT training was not included. The decrease in alpha-1 and alpha-2 activity might be attributed to ATT training, although it could also be attributed to time course. The ATT group should be compared to a control group that completed training without the ATT task with a similar cognitive load for the same number of days in future studies. In addition, performance of the ATT task during the EEG recording and homework assignments were insufficiently controlled. An objective measure should have been used to confirm how closely participants performed the ATT task as instructed. In the future, we could measure a biological marker (e.g., Kawashima & Kumano, 2017) or behavioral marker (e.g., Tomita et al., 2017) which relates to the attentional control function to confirm improvement through the homework assignment.

Conclusion

This study demonstrated that brain electrical activity during the divided attention component of ATT changed after 20 days of ATT practice, which may be due to either ATT practice or related to the time course selected. For measuring EEG, brain activity during rest was subtracted from that during the ATT task, and which was compared before and after the ATT homework assignment. Alpha-2 activity around the ITC and alpha-1 activity around the OFC during the divided attention component of ATT decreased after the homework, and that decreased alpha-1 power in the left OFC correlated with reduced subjective difficulty during the divided attention component. One possibility is that brain activity changed as an effect of ATT practice, although we could not confirm causality in this study. Further studies are required which include a control group that would complete training without the ATT task, but instead a different task with a similar cognitive load for the same number of days.

Footnotes

Acknowledgments

We appreciate Enago () for the English language review.

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Ibuka Funding.

ORCID iDs

Kaori Usui

Issaku Kawashima

Nozomi Tomita

Toru Takahashi

Hiroaki Kumano

Author Biographies

Kaori Usui has got an MA at the Graduate School of Human Sciences, Waseda University, Japan. She is a clinical psychologist working in a hospital.

Issaku Kawashima is a researcher at ATR Brain Information Communication Research Laboratory Group, Japan.

Nozomi Tomita is a senior lecturer of clinical psychology at the Faculty of Human Sciences, Waseda University, Japan.

Toru Takahashi is a research associate of cognitive neuroscience at the Faculty of Human Sciences, Waseda University, Japan.

Hiroaki Kumano is a professor of clinical psychology at the Faculty of Human Sciences, Waseda University, Japan. He is also a psychosomatic physician specialized in anxiety disorders, mood disorders and psychosomatic disorders. His research focuses on attention and awareness in mindfulness meditation.

References

Bremner

J. D.

Vythilingam

Vermetten

Nazeer

Adil

Khan

Staib

L. H.

Charney

D. S.

(2002). Reduced volume of orbitofrontal cortex in major depression. Biological Psychiatry, 51(4), 273–279. https://doi.org/10.1016/S0006-3223(01)01336-1

Eastwood

J. D.

Frischen

Fenske

M. J.

Smilek

(2012). The unengaged mind: Defining boredom in terms of attention. Perspectives on Psychological Science: A Journal of the Association for Psychological Science, 7(5), 482–495. https://doi.org/10.1177/1745691612456044

Faber

P. L.

Lehmann

Tei

Tsujiuchi

Kumano

Pascual-Marqui

R. D.

Kochi

(2012). EEG source imaging during two qigong meditations. Cognitive Processing, 13(3), 255–265. https://doi.org/10.1007/s10339-012-0441-4

Girkin

C. A.

Miller

N. R.

(2001). Central disorders of vision in humans. Survey of Ophthalmology, 45(5), 379–405. https://doi.org/10.1016/S0039-6257(00)00208-3

Goldman

R. I.

Stern

J. M.

Engel

J. J.

Cohen

M. S.

(2002). Simultaneous EEG and fMRI of the alpha rhythm. Neuroreport, 13(18), 2487–2492. https://doi.org/10.1097/01.wnr.0000047685.08940.d0

Gronwall

D. M. A.

(1977). Paced auditory serial-addition task: A measure of recovery from concussion. Perceptual and Motor Skills, 44(2), 367–373. https://doi.org/10.2466/pms.1977.44.2.367

Hiroe

Kojima

Yamamoto

Nojima

Kinoshita

Hashimoto

Watanabe

Maeda

Furukawa

T. A.

(2005). Gradations of clinical severity and sensitivity to change assessed with the Beck Depression Inventory-II in Japanese patients with depression. Psychiatry Research, 135(3), 229–235. https://doi.org/10.1016/j.psychres.2004.03.014

Holmes

A. P.

Blair

R. C.

Watson

J. D. G.

Ford

(1996). Nonparametric analysis of statistic images from functional mapping experiments. Journal of Cerebral Blood Flow and Metabolism: Official Journal of the International Society of Cerebral Blood Flow and Metabolism, 16(1), 7–22. https://doi.org/10.1097/00004647-199601000-00002

Imai

Kumano

Imai

Nedate

(2015). Subjective aspects of voluntary attention control and its relation to depression and anxiety. The Japanese Journal of Cognitive Psychology, 8(1), 85–95.

10.

Kawashima

Kumano

(2017). Prediction of mind-wandering with electroencephalogram and non-linear regression modeling. Frontiers in Human Neuroscience, 11, 365. https://doi.org/10.3389/fnhum.2017.00365

11.

Knowles

M. M.

Wells

(2018). Single dose of the attention training technique increases resting alpha and beta-oscillations in frontoparietal brain networks: A randomized controlled comparison. Frontiers in Psychology, 9, 1768. https://doi.org/10.3389/fpsyg.2018.01768

12.

Kojima

Furukawa

T. A.

Takahashi

Kawai

Nagaya

Tokudome

(2002). Crosscultural validation of the Beck Depression Inventory-II in Japan. Psychiatry Research, 110(3), 291–299. https://doi.org/10.1016/S0165-1781(02)00106-3

13.

Kringelbach

M. L.

(2005). The human orbitofrontal cortex: Linking reward to hedonic experience. Nature Reviews. Neuroscience, 6(9), 691–702. https://doi.org/10.1038/nrn1747

14.

Kubicki

Herrmann

W. M.

Fichte

Freund

(1979). Reflections on the topics: EEG frequency bands and regulation of vigilance. Pharmakopsychiatrie, Neuro-Psychopharmakologie, 12(2), 237–245. https://doi.org/10.1055/s-0028-1094615

15.

Linehan

M. M.

Heard

H. L.

Armstrong

H. E.

(1993). Naturalistic follow-up of a behavioral treatment for chronically parasuicidal borderline patients. Archives of General Psychiatry, 50(12), 971–974. https://doi.org/10.1001/archpsyc.1993.01820240055007

16.

Niedermeyer

da Silva

F. L.

(2005). Electroencephalography: Basic principles, clinical applications, and related fields (5th ed.). Lippincott Williams and Wilkins. https://trove.nla.gov.au/version/45745233

17.

Normann

van Emmerik

A. A.

Morina

(2014). The efficacy of metacognitive therapy for anxiety and depression: A meta-analytic review. Depression and Anxiety, 31(5), 402–411. https://doi.org/10.1002/da.22273.

18.

Papageorgiou

Wells

(2000). Treatment of recurrent major depression with attention training. Cognitive and Behavioral Practice, 7(4), 407–413. https://doi.org/10.1016/S1077-7229(00)80051-6

19.

Pascual-Marqui

R. D.

(2002). Standardized low-resolution brain electromagnetic tomography (sLORETA): Technical details. Methods & Findings in Experimental & Clinical Pharmacology, 24, 5–12.

20.

Petrovic

Ingvar

(2002). Imaging cognitive modulation of pain processing. Pain, 95(1–2), 1–5. https://doi.org/10.1016/S0304-3959(01)00467-5

21.

Pizzagalli

D. A.

Oakes

T. R.

Fox

A. S.

Chung

M. K.

Larson

C. L.

Abercrombie

H. C.

Schaefer

S. M.

Benca

R. M.

Davidson

R. J.

(2004). Functional but not structural subgenual prefrontal cortex abnormalities in melancholia. Molecular Psychiatry, 9(4), 393–405. https://doi.org/10.1038/sj.mp.4001469

22.

Ring

H. A.

Bench

C. J.

Trimble

M. R.

Brooks

D. J.

Frackowiak

R. S.

Dolan

R. J.

(1994). Depression in Parkinson's disease: A positron emission study. The British Journal of Psychiatry: The Journal of Mental Science, 165(3), 333–339. https://doi.org/10.1192/bjp.165.3.333

23.

Rolls

E. T.

Grabenhorst

(2008). The orbitofrontal cortex and beyond: From affect to decision-making. Progress in Neurobiology, 86(3), 216–244. https://doi.org/10.1016/j.pneurobio.2008.09.001

24.

Roppongi

Nakamura

Asami

Hayano

Otsuka

Uehara

Fujiwara

Saeki

Hayasaka

Yoshida

Shimizu

Inoue

Hirayasu

(2010). Posterior orbitofrontal sulcogyral pattern associated with orbitofrontal cortex volume reduction and anxiety trait in panic disorder. Psychiatry and Clinical Neurosciences, 64(3), 318–326. https://doi.org/10.1111/j.1440-1819.2010.02085.x

25.

Siegle

G. J.

Ghinassi

Thase

M. E.

(2007). Neurobehavioral therapies in the 21st century: Summary of an emerging field and an extended example of cognitive control training for depression. Cognitive Therapy and Research, 31(2), 235–262. https://doi.org/10.1007/s10608-006-9118-6

26.

Tomita

Imai

Kanayama

Kawashima

Kumano

(2017). Use of multichannel near infrared spectroscopy to study relationships between brain regions and neurocognitive tasks of selective/divided attention and 2-back working memory. Perceptual and Motor Skills, 124(3), 703–720. https://doi.org/10.1177/0031512517700054

27.

Wells

(1990). Panic disorder in association with relaxation induced anxiety: An attentional training approach to treatment. Behavior Therapy, 21(3), 273–280. https://doi.org/10.1016/S0005-7894(05)80330-2

28.

Wells

(2007). The attention training technique: Theory, effects, and a metacognitive hypothesis on auditory hallucinations. Cognitive and Behavioral Practice, 14(2), 134–138. https://doi.org/10.1016/j.cbpra.2006.01.010

29.

Wells

(2009). Metacognitive therapy for anxiety and depression. The Guilford Press. https://www.guilford.com/books/Metacognitive-Therapy-for-Anxiety-and-Depression/Adrian-Wells/9781609184964

30.

Wells

Fisher

Myers

Wheatley

Patel

Brewin

C. R.

(2009). Metacognitive therapy in recurrent and persistent depression: A multiple-baseline study of a new treatment. Cognitive Therapy and Research, 33(3), 291–300. https://doi.org/10.1007/s10608-007-9178-2

31.

Wells

Matthews

(1994). Attention and emotion: A clinical perspective. Lawrence Erlbaum Associates, Inc. https://doi.org/10.1002/cpp.5640020208

32.

Wells

Matthews

(1996). Modelling cognition in emotional disorder: The S-REF model. Behaviour Research and Therapy, 34(11–12), 881–888. https://doi.org/10.1016/S0005-7967(96)00050-2

33.

Wells

White

Carter

(1997). Attention training: Effects on anxiety and beliefs in panic and social phobia. Clinical Psychology & Psychotherapy, 4(4), 226–232. https://doi.org/10.1002/(SICI)1099-0879(199712)4:4<226::AID-CPP129>3.0.CO;2-M