Hemodynamic response to familiar faces in children with ADHD

Seventeen ADHD boys and 16 TD boys were recruited. All diagnoses were guided by the DSM-5 [2] and were made by two pediatric neurologists who worked with neurodevelopmental disorders for over 5 years. Exclusion criteria included learning, language, neurological or developmental disabilities. Current study protocols were approved by the Ethical Committee of the Dokkyo Medical University Saitama Medical Center (hosp-k 24,016). We obtained written informed consent from the participants and their parents. The experiments were conducted according to the Declaration of Helsinki.

After excluding subjects whose fNIRS data failed the motion criteria (frequent sharp changes in the time series of the raw fNIRS data), 23 subjects remained: 9 boys with ADHD (mean age, 9 years 9 months; SD, 1 year 6 months; range, 8–13 years), and 14 TD boys (mean age, 9 years 4 months; SD, 1 year 6 months; range, 7–12 years) (Table 1). All participants were right-handed. Six out of nine boys with ADHD received methylphenidate, and one ADHD boy received atomoxetine. All of them participated in the examination for 4–6 h following medication. ADHD and TD subjects were not different according to age (see Table 1). The ADHD-Rating Scale scores were 34.2 (range, 11–52; SD, 13.9) in the ADHD boys and 9.8 (range, 2–18; SD, 4.5) in the TD boys. The Wechsler Intelligence Scale of Children Third Edition (WISC-III) full IQ scores of the TD boys (mean, 111.1; SD, 10.0; range, 90–129) were significantly higher (t = 3.68, p = 0.048) than those of the ADHD participants (mean, 94.1; SD, 16.2; range, 78–125).

Each child was examined while sitting on a chair and facing a computer screen approximately 50 cm away. The boys watched the stimuli passively while their hemodynamic response was measured, and their behavior was recorded on video during the test. The sequence of stimulus presentations was composed of baseline and test periods. The test periods consisted of two stimulus conditions: familiar face and unfamiliar face conditions. In those conditions, the participants looked at familiar face or unfamiliar face color photographs, respectively (Fig. 1). The order of the conditions was counterbalanced across the participants. The familiar face photograph was the participant’s mother’s face. The unfamiliar face photographs were the faces of age-matched unknown females. To exclude the external features of the faces, we removed the hair from the face images so that only the internal features were visible. The sizes of the photographs were approximately 13 × 10 deg for the faces. The duration of the test period was fixed at 10 s. Each face image was shown 10 times in one test period. The duration of each face image was 800 ms, and the 200 ms inter-stimulus interval was filled by the presentation of a fixation point (a blinking black dot, 3.5 × 3.5 deg). Each test period was preceded by a baseline period that was at least 20 s long. During the baseline period, the screen of the monitor was a uniform blank white screen. The blank screen was presented for 800 ms, and a 200 ms inter-blank screen interval was filled with the same fixation point (a blinking black dot), the same as in the test period. The results obtained from viewing the blank screen were used as the baseline. To draw and retain the attention of the boys, both face stimuli and the blank screen were accompanied by a beeping sound, as in previous studies [26].

We used a Hitachi ETG-4000 system (Hitachi Medical, Chiba, Japan), which can record 24 channels simultaneously with 12 channels over the right temporal area and 12 over the left. This instrument generates two wavelengths of NIR (695 nm and 830 nm) and measures the time courses of oxyhemoglobin (oxy-Hb) and deoxyhemoglobin (deoxy-Hb) levels and their sum (total hemoglobin: total-Hb) with a 0.1 s time resolution. The fNIRS probes contained nine optical fibers (3 × 3 arrays) composed of five emitters and four detectors. The distance between the emitters and the detectors was set at 3 cm. Each pair of adjacent emitting and detecting fibers defined a single measurement channel.

The placement of the probes was set on the bilateral temporal cortices centered at T5 and T6 according to the International 10–20 system [28] (Fig. 2). This location of the probes was the same as that reported by Otsuka et al. [24]. When the probes were positioned, the experimenter checked to see if the fibers were touching the boys’ scalps correctly. The Hitachi ETG-4000 system automatically detects whether the contact adequately measures emerging photons for each channel.

The analysis of fNIRS data in the present study was similar to that reported by Nakato et al. [26]. Before data analysis, we monitored the video recording of the boys’ behavior to evaluate good trials for analysis. We removed a trial from the analysis if either the accumulative looking time within the trial did not reach 5 s or if movement artifacts were detected by the analysis of sharp changes in the time-series of the raw fNIRS data.

The raw data on oxy-Hb, deoxy-Hb, and total-Hb concentrations from each channel were digitally band-pass-filtered at 0.02–1.0 Hz to remove noise caused by heartbeat pulsations or any longitudinal signal drifts. The mean concentration of each channel within a subject was then calculated by averaging data across the trials in a time-series with a 0.1 s resolution.

Based on mean concentrations in the time-series, we calculated the Z-scores for oxy-, deoxy-, and total-Hb in the familiar face and the unfamiliar face conditions for each channel within a subject [7, 20, 24‐27]. The Z-score at each time-point indicates the deviation of hemodynamic response from the baseline during the presentation of faces. The mean concentration of the Z-score at 3 s before the trial was used as a baseline. The Z-scores were calculated separately for oxy-Hb, deoxy-Hb, and total-Hb in the familiar face and unfamiliar face conditions.

The Z-scores obtained from the 12 channels within the left and right temporal areas were averaged across trials for every subject in order to increase the signal-to-noise ratio. Although raw NIRS data consist of relative values that cannot be averaged directly across subjects or channels, normalized data including Z-scores can be averaged regardless of the unit. This calculation of the Z-score is a reliable analysis for changes in Hb concentration in the brain regions under the sensors since the analysis is independent of the differential path length factor (DPF). On the basis of these methodological advantages for fNIRS data, we applied the same analysis of the Z-score in the present study as in previous fNIRS studies with children [7, 20, 24‐27].

Consistent with our previous studies using fNIRS [7, 20, 24‐27], we found a response peaking a few seconds after the stimulus onset. The hemodynamic response of oxy-Hb in the familiar face condition increased gradually from approximately 3 s after stimulus onset and reached a plateau approximately 10 s after stimulus onset in ADHD boys (see Fig. 3). In contrast, the hemodynamic response of oxy-Hb in the familiar face condition increased gradually from approximately 10 s after stimulus onset in TD boys (see Fig. 3). Therefore, we performed statistical analyses against the mean Z-scores for the 3–10 s (the early phase) and 10–17 s (the late phase) durations of the test period. A two-tailed one-sample t-test against a chance level of 0 (baseline) was conducted for the mean Z-score during the early phase and late phase in both temporal areas.

Data from the 23 boys who looked at faces over more than three trials in both the familiar face and unfamiliar face conditions were obtained. Available trials were averaged across trials for each condition (familiar face, mean 5.06, SD 0.89; unfamiliar face, mean 4.97, SD 0.95). The mean number of removed trials for the familiar face condition was 0.33 (SD 0.74) while that for the unfamiliar face condition was 0.30 (SD 0.64). There was no significant difference between the numbers of available trials under each condition.

On the basis of the wavelengths of the Hitachi ETG-4000 system (695 nm and 830 nm), estimations of oxy-Hb concentration have been reported to better reflect changes in regional cerebral blood flow than deoxy- or total-Hb concentrations [29].

The time-course of the average change in oxy-, deoxy-, and total-Hb concentrations in the boys during the presentation of the familiar face and unfamiliar face conditions is shown in Fig. 3. To examine the possibility that the activity for both the familiar and unfamiliar face stimuli were greater or smaller than the baseline, we conducted a two-tailed one-sample t-test with the mean Z-scores for oxy- deoxy-, and total-Hb concentrations for the 3–10 s (the early phase) and 10–17 s (the late phase) durations of the test period in each familiar and unfamiliar face condition (Fig. 4).

In the familiar face condition, the Z-scores for oxy-Hb concentrations in the ADHD boys were significantly increased in both temporal cortices in the early phase: right temporal cortex; t (8) = 3.80, p = 0.005; left temporal cortex; t (8) = 2.79, p = 0.023. They also showed a significant increase in both temporal cortices in the late phase: right temporal cortex; t (8) = 5.27, p = 0.0007; left temporal cortex; t (8) = 2.53, p = 0.035. However, in the unfamiliar face condition, there was no significant change in the concentration of oxy-Hb in the ADHD boys in either the early or late phase. TD boys showed no significant change in the early phase; however, in the late phase, they showed significant increases in the familiar face condition in the right temporal cortex: t (13) = 3.12, p = 0.008. Three-way ANOVA (2 × 2 × 2: group x condition x hemisphere) showed the only main effect in the condition (group p = 0.27; condition p = 0.011; hemisphere p = 0.087).

The Z-scores for deoxy-Hb concentrations of the ADHD boys showed no significant change in the familiar face condition. However, in the unfamiliar face condition, the Z-scores of the ADHD boys were significantly decreased in both the right and left temporal cortices in the early phase: right temporal cortex; t (8) = 2.47, p = 0.039; left temporal cortex; t (8) = 4.97, p = 0.001. In the late phase, they showed significant decreases in the left temporal cortex: t (8) = 2.42, p = 0.041. TD boys showed no significant change in the concentration of deoxy-Hb in either condition in either the early or late phase.

The Z-scores for the total-Hb concentrations in the ADHD boys showed no significant change in either condition in the early phase. However, in the late phase, they showed significant increases in the right temporal cortex (t [8] = 2.67, p = 0.021), but only in the familiar face condition. TD boys showed no significant change in the concentration of the total-Hb in either condition in both the early and late phases.