Sex-Specific Electrocortical Interactions in a Color Recognition Task in Men and Women with Opioid Use Disorder

Abstract

Background: Opioid use disorder (OUD) and associated overdose deaths have reached epidemic proportions worldwide over the past two decades, with death rates for men consistently reported at twice the rate for women. We have recently reported sex-specific differences in electrocortical activity in persons with OUD in a visual object recognition task. The mesolimbic dopamine (DA) system is implicated in OUD but also plays a critical role in some disorders of visual attention and a modulatory role in the processing of visual stimuli in the blue cone pathway of the retina. We hypothesized that electrocortical responses to color stimuli would be affected differentially in men and women with OUD. Methods: Using a controlled, cross-sectional, age-matched (18-56 years) design, we evaluated color processing in male and female subjects recruited from a community-based, high-intensity residential substance abuse and detoxification treatment program. We evaluated electroencephalogram (EEG) event-related potentials (ERPs) and reaction time (RT), in male and female participants with OUD (n = 38) vs. sex- and age-matched non-OUD control participants (n = 37) in a simple color recognition Go/No-Go task, as well as perceptual and behavioral responses in physiological and neuropsychological tests. Results: N200, P300, and late potential (LP) Relevant stimulus-induced ERPs were evoked by the task and were well-differentiated from Irrelevant distractor stimuli. P300 amplitudes were significantly greater and N200 and LP latencies were significantly shorter in male vs. female non-OUD controls in this task. There were significant sex differences in N200, P300, and LP amplitudes and latencies between male vs. female non-OUD subjects and OUD differences with blue color as the Relevant stimulus. In the Binocular Rivalry Test, there were shorter dwell times for perceiving a blue stimulus in male OUD subjects and there were significant sex and OUD differences in neuropsychological tests including Finger Tapping, Trails A/B, and Symbol Digit Modalities Test. Conclusions: These findings suggest that there are significant sex-related physiological, perceptual, and cognitive differences in color processing that may result from deficits in DA production in the retina that mirror deficits in mesolimbic DA transmission correlating with OUD, suggesting that blue color processing has the potential to be an effective biomarker for brain DA and for diagnosis and monitoring of treatment efficacy in substance use disorders.

Keywords: color vision processing; dopamine (DA); electroencephalogram (EEG); event-related potentials (ERP); opioid use disorder (OUD); sex-related differences; substance use disorder (SUD); visual attention; visual evoked potential (VEP).

Figure 1

Sex differences in event-related potentials in a simple Go/No-Go color recognition task in non-OUD control participants. Red, Green, and Blue stimuli were randomly presented at 2–4 s intervals. Participants responded to the Relevant stimulus, in this case Blue. (A,B) These graphs show grand-averaged VEPs recorded in male and female non-OUD participants at electrode Pz in response to Red, Green, and Blue stimuli. The vertical scales are normalized to facilitate comparisons between gender. The dashed lines indicate the time of presentation of the stimulus. The components of the VEP included the N50, P100, N100, P200, N200, and P300. Reaction time (RT) is shown with an arrow for the averaged Go response to the Relevant stimulus. Note that this simple color recognition task produced P300s differentiated to Relevant vs. Irrelevant stimuli. The 128 sensor topomaps (insets) represent grand averaged potentials in male and female non-OUD participants at 349 ms (P300) after the presentation of the color stimuli. The color map scales are also normalized to facilitate comparisons between gender for non-OUD participants in this figure. The top of each oval is the front of the head and the bottom of each oval represents the back of the head, as if looking down on the head from above. Violet represents extreme negative potentials and red represents extreme positive potentials. (C,D) These graphs show grand-averaged VEPs recorded in male and female non-OUD participants at electrode Fz in response to Red, Green, and Blue stimuli, with Blue as the Relevant stimulus. Note that this task induced a late positive (LP) potential associated with the Relevant stimulus that was differentiated from Irrelevant stimuli. The 128 sensor topomaps (insets) represent grand averaged potentials in male and female non-OUD participants at 699 ms (LP) after the presentation of the color stimuli. Average RT is shown for the Relevant stimulus. Note that females have a less prominent LP than males. (E,F) These box plot graphs summarize N200, P300, and LP ERP amplitude and latency measurements for all male (n = 24) and female (n = 22) non-OUD participants by gender and ERP component. Male non-OUD subjects were characterized by significantly larger P300 and LP amplitudes and shorter N200 and LP latencies than female non-OUD subjects. Asterisks *, *** indicate significance levels p < 0.05 and p < 0.001, respectively.

Figure 2

Effects of wavelength on event-related potentials in male and female non-OUD control participants in the color recognition task. Participants responded to the Relevant stimulus for each of the wavelengths Red, Green, and Blue in separate Go/No-Go experiments. The vertical scales and color maps are NOT normalized in this figure in order to compare Relevant responses for Red, Green, and Blue within sex. (A,B) Topomaps show grand-averaged P300s (Pz) and LPs (Fz) on the head sensor net in males and females. There was no obvious effect of wavelength on P300s or LPs in non-OUD participants. (C,D) Synoptic plots showing a heat map of negative (blue) and positive (red) voltage for all 128 electrodes on the head corresponding to the Relevant stimulus for Red, Green, and Blue. It is evident that there were slight differences in the amplitudes of P300 and LP by wavelength. Color plots were normalized within sex for wavelength. (E–G) Summary of sex differences for the effects of wavelength on ERP amplitudes. There were significant differences between male vs. female P300 amplitudes for Blue as the Relevant stimulus, but not for Red or Green. There was also a significant effect of wavelength on LPs for Red as the Relevant stimulus. Asterisks * and ** indicate significance levels p < 0.05 and 0.001 respectively between males and females, and hashtags ### indicate significance level p < 0.001 between colors.

Figure 3

Event-related potentials in a simple visual recognition task in OUD subjects by gender. (A,B) These graphs show grand-averaged waveforms obtained at Pz in both males and females. The vertical scales are NOT normalized in this figure in order to compare Relevant responses for control vs. OUD within gender. Compared to controls, OUD subjects were characterized by smaller N200 and P300s at Pz on the grand-averaged waveform (Relevant stimulus only shown) in both males and females. (C,D) These graphs show grand-averaged waveforms obtained at Fz in both males and females. The vertical scales are NOT normalized in this figure in order to compare Relevant responses for control vs. OUD within gender. Similarly to Pz, OUD subjects were characterized by reduced LP amplitudes at Fz. (E,F) These graphs show all the ERP amplitude replicates in male and female controls and OUD for the P300 at Pz and LP at Fz. These are individual amplitude measurements within-subject. Note that male OUD subjects showed smaller P300 amplitudes. (G,H) These graphs show all the ERP latency replicates in male and female controls and OUD for the P300 at Pz and LP at Fz. These are individual amplitude measurements within-subject. Note that while P300 latencies did not seem to be affected in OUD subjects, LP latencies were significantly longer in males but shorter in females. Asterisks *, *** indicate significance levels p < 0.05 and p < 0.001 within sex and hashtags #, ### indicate significance levels between sex, respectively.

Figure 4

Sex and OUD differences in blue color processing in binocular rivalry. The image above the graph shows the superimposed number “1” in red and “2” in blue. When viewed through Red/Blue anaglyph glasses perception alternates between the 2 stimuli at intervals termed binocular rivalry. Subjects were instructed to press the corresponding key on the computer keyboard when they perceived “1” vs. “2”. The graph below shows the ratio of the Blue vs. Red perception elapsed time for all replicates. Male OUD subjects were characterized by less time in the Blue perception compared to Red. Asterisk * indicates significance level p < 0.05.

Figure 5

Sex and OUD differences in neuropsychological tests. (A) This graph shows differences between males vs. females in non-OUD and OUD subjects for all replicates in the Finger Tapping Test. There was marked significance between males vs. females, but no OUD differences. (B) This graph shows differences between males vs. females in non-OUD and OUD subjects for all replicates in the Trails A/B Test. There were significant differences between non-OUD and OUD in this test for both males and females. (C) This graph shows differences between males vs. females in non-OUD and OUD subjects for all replicates for Symbol Digit Score. There were significant differences between non-OUD and OUD subjects in this test for both males and females. (D). This graph shows differences between males vs. females in non-OUD and OUD subjects for all replicates for Symbol Digit Error. There were no sex or OUD differences. Asterisks *, **, *** indicate significance levels p < 0.05, p < 0.01, and p < 0.001 by OUD, hashtags # indicates number of taps, and hashtags ### represent significant differences in errors made in comparisons between sex.