Scalp Application of LED Therapy to Improve Thinking and Memory in Veterans With Gulf War Illness

The purpose of this study is to learn if an experimental treatment can help thinking ability, and memory in Veterans with Gulf War Veterans Illnesses (GWVI). The experimental treatment uses light-emitting diodes (LEDs), that are applied outside the skull, to the head using a helmet that is lined with near-infrared diodes. LEDs are also placed in the nostrils (one red diode; and one near-infrared diode), near-infrared photons to the olfactory bulbs located on the orbito-frontal cortex. There are connections between the olfactory bulbs and the hippocampus. A treatment takes about 30 minutes. The participants receive a series of LED treatments which take place as outpatient visits at the VA Boston Healthcare System, Jamaica Plain Campus. The FDA considers the helmet LED device used here, to be a non-significant risk device. The diodes in the device placed in the nose are low-risk devices, within the FDA Category of General Wellness. In addition, a single, 90 mW near-infrared (NIR) LED was placed on each ear. The LEDs do not produce heat.

Rationale: Impaired cognition is one of the 3 major symptom areas of GWVI. Mitochondrial dysfunction is reported in Gulf War Illness (GWI) veterans, associated with neurotoxicant exposures during deployment - e.g., organophosphate pesticides (OP); and pretreatment nerve agents, pyridostigmine bromide (PB) pills. Improved mitochondrial function with increased production of adenosine tri-phosphate (ATP) has been reported in hypoxic/compromised cells treated with red/NIR photons. Recent studies in humans have shown an increase in regional cerebral blood flow subjacent to where the near-infrared LEDs were placed on the scalp (Schiffer et al., 2009; Nawashiro et al., 2012; Chao L.L., 2019). General Explanation of the Light-Emitting Diode Treatment Procedure: LED treatments take place as outpatient visits at the VA Boston Healthcare System (VABHS) Jamaica Plain Campus (JP Campus), 150 South Huntington Ave., Boston, MA, 02130 or a second site, VA Medical Center, San Francisco, San Francisco, CA. The transcranial LED treatment procedure is painless, non-invasive and no heat is generated. Administration of treatment: Three devices were placed at the same time on the participant. All LED devices were approved for use by the VABHS safety committee. Device 1. NIR, LED lined helmet (PhotoMedex, Horsham, PA or Thor Photomedicine, Inc., Hampstead, MD). If a participant's head circumference was larger than 24" circumference a larger LED helmet (Thor Photomedicine) was used to accommodate head size. Device 2. Two intranasal LEDs (red, 633 nm and near-infrared, 810nm, single diodes, Vielight, Inc., Toronto). One Intranasal LED was placed in each nostril, held in place by a plastic clip. Placement of red and NIR Intranasals was alternated by side (left/right) at each session. Intranasal LEDs have an automatic timer, and treatment was administered for 25 minutes, during the ongoing helmet treatment. Device 3. Two cluster heads (MedX Health, Toronto) were used simultaneously on the L and R ears for 4 mins, turned on towards the end of treatment. Each participant wore a clear plastic liner beneath the LED Helmet to protect the LEDs and for hygienic reasons. No liquids or gel are used to hold the LEDs in place on the head. This clear plastic liner was assigned to each participant, and only used by that participant. It was kept in a locked filing cabinet in the treatment room, and discarded after the participant's completion of the study. Sham and Real LED devices were identical in look and feel, except no photons were emitted in the Sham devices. For the purpose of blinding, the participant and the person administering treatments wore goggles (LS-DIO, Phillips Safety Products, Inc.) that blocked 600-900nm wavelengths including red 633 nm wavelength, emitted from the red Intranasal diode. The participant will not feel anything when the LED lights are on; it is a painless, noninvasive treatment procedure. There are small fans built into the LED helmet (to assist in cooling). The participant will always hear the fans, during each LED treatment, whether the LEDs are on, or they are not on. Sequence and Timing of LED treatment, each session: Part 1: The LEDs in the center of the helmet are on for about 14 minutes; then turned off. Part 2: The LEDs in the left and right side of the helmet are on for about 14 minutes; then off. During Parts 1 and 2, the intranasal devices are on for about 25 minutes. Part 3: Two LED cluster heads were used simultaneously on the L and R ears for 4 mins, turned on towards the end of treatment. Approximate treatment time is about 30 minutes, per treatment visit. The participant was treated in a soft recliner chair. Each visit lasts about 35 to 45 minutes, allowing time for record keeping. There were 15 visits in an LED treatment series. Each LED treatment series lasts for 7.5 weeks. Each LED treatment visit is scheduled twice per week, with at least 48 hours between each visit. If the participant needs to miss an appointment, that appointment will be re-scheduled. However, if the participant only made 1 appointment, over a two-week period, he/she was withdrawn from the study. The length of time for participation in this study is 2-4 months including pre- testing, and final follow-up testing after the last LED treatment visit. This study is sponsored by the Department of Veterans Affairs, Clinical Science Research and Development, Office of Research & Development (ORD). There are no potential conflicts of interest associated with this research. Referral and Screening of Potential Participants: Participants were recruited through a Fort Devens, MA, cohort of Gulf War Veterans (Proctor et al., 1998), and through the VA Informatics and Computing Infrastructure (VINCI)/Corporate Data Warehouse (CDW) database, with approval from VINCI. The San Francisco VA Medical Center (SF VAMC) was a second site on the study (35 Veterans were run in Boston; 12 were run in San Francisco on the same protocol.) Those recruited from the VINCI/CDW database, resided within a 25 mile radius of the Boston VA Healthcare System (VABHS) or 25 miles of the SF VA Medical Center (VAMC). The Institutional Review Board at the VA BHS and the SF VAMC (University of California, San Francisco) approved the study. In accordance with the Declaration of Helsinki, Informed Consent and HIPAA authorization was obtained from all participants. All study visits take place at the VA BHS, JP Campus or San Francisco VA Medical Center. After referral to this study, the potential participant was contacted by telephone, and a description of the full study protocol is explained, including time required, and reimbursement for time and effort. After the initial contact by telephone, an initial visit to the VABHS, JP campus was scheduled. At this time, the entire study was explained and questions answered. After the informed consent form (ICF) has been signed, an appointment is then scheduled for Neuropsychological (NP) Screening testing. These tests measure the ability to think and remember recent information. If results from the Neuropsychological Screening tests show that the potential participant is eligible for entry into this study, an additional appointment is scheduled where Additional Neuropsychological Testing is performed, and Additional Health Information is obtained. NP Assessments before and after LED Intervention Series NP testing was completed within 1 week before (T1), and at 1 week (T2) and 1 month (T3) after the intervention series. For those who received Sham First, there was an optional, second Real Series. NP testing was also completed within 1 week (T4) and 1 month (T5) after the second optional, intervention series. Each NP Testing session was completed in 1-1/2 hours. Primary Outcome Measures assessed 3 neurocognitive domains impaired in GW veterans: 1) Attention/Executive Function: Digit Span Forward and Backward (WAIS-IV; Wechsler, 2008), Delis-Kaplan Executive Function (D-KEF) Trails and Color-Word Interference (Stroop) (Delis, Kaplan & Kramer, 2001); 2) Learning and Memory: California Verbal Learning Test-II (CVLT-II; Delis, Kramer, Kaplan, & Ober, 2000); and 3) Attention/Visual Spatial: Conner's Continuous Performance Test II (Administered on laptop computer; CPT) (Letz & Baker, 1988; Rosvold et al., 1956); Rey Osterrieth Complex Figure Test (ROCFT; Knight & Kaplan, 2004; Osterrieth, 1944; Rey, 1941). An alternate version of the CVLT was administered at every other testing session to avoid practice effects. Secondary Measures were assessed for: 1) Pain: Visual Analog Scale (VAS) Pain Scale (0-10) (Farrar et al., 2001); Short Form McGill Pain Questionnaire (Melzack, 1984) and the West Haven-Yale Multi-dimensional Pain Inventory (WHYMPI, Kerns et al., 1985). ; 2) Fatigue: Multi-Dimensional Fatigue Inventory (Smets et al., 1995); 3) Sleep: Pittsburgh Sleep Quality Index (PSQI) (Buysee et al., 1989), Karolinska Sleepiness Scale (KSS); 4) Mood: Beck Depression inventory (BDI; Beck, 2006); and 5) General Physical Health: Short Form-36V Plus (Ware et al., 2000); and the Health Symptom Checklist (HSC) adapted from Bartone et al., (1989); PTSD Checklist PCL-C (Weathers et al., 1994). There were a total of 59 tests/subtests. Statistical Analyses and Power Statement: Tests and subtests within each cognitive domain were analyzed pre- and post- LED intervention. Attention/Executive Function: Digit Span Subtests (WAIS-IV; Wechsler, 2008); Trail-Making Test (Delis, Kaplan, Kramer, 2001); and Color Word Interference Test (Stroop) (Delis, Kaplan, Kramer, 2001) Learning and Memory: California Verbal Learning Test-II (CVLT-II; Delis, Kramer, Kaplan, & Ober, 2000) Psychomotor/Visual Spatial; Continuous Performance Test (Administered on computer; CPT-R) (Letz & Baker, 1988; Rosvold et al., 1956); Rey Osterrieth Complex Figure Test (ROCFT) (Knight & Kaplan, 2004) Power Analysis: Power was computed under the following assumptions for ANOVA: a) alpha = .05 (2-tail); b) to establish clinical relevance, Cohen's large effect size (.25) was used. Power = .86 to detect a significant between-group difference in change from Baseline to end of Treatment for an interim analysis.

Gulf War Veterans Illness, Veterans Health, Executive Function, Memory, Attention, Cognition, Laser Therapy, Low-Level, Phototherapy, Light Therapy, Depression, Light-Emitting Diodes

This was a blinded, randomized, sham-controlled partial crossover study, with 2 Groups. Group 1: Sham LED Series First followed by optional Real LED Series Second. Group 2: One Real LED Series, only. Neuropsychological (NP) tests were completed at Entry (T1), within 1 week post- (T2) and at 1 month post- the 15th treatment (T3) (First series). NP testing was administered after the optional Second Real Series within 1 week post- (T4) and at 1 month post- the 15th treatment (T5). Primary Outcome Measures included NP tests in 3 domains: 1) Attention/Executive Function; 2) Learning and Memory; 3) Psychomotor/Visual Spatial. The Secondary Outcome Measures included sleep, pain, fatigue, and mood.

The participant and the person administering treatments wore goggles (LS-DIO, Phillips Safety Products, Inc.) that blocked 600-900nm wavelengths including red 633 nm wavelength, emitted from the red Intranasal diode. All participants were randomly assigned by computer to Group 1 (Sham First) or Group 2 (Real First) in blocks of 10.

Participants in this group receive a series of 15 real LED treatments with the helmet and intranasal devices: Transcranial NIR, 830nm LED Helmet and two intranasal nose clips (633 nm, and 810 nm, Vielight, Inc., Toronto), 28-minute treatment, 2 days/ week, 7.5 weeks, at least 48 hours between treatments. Additionally, two NIR 870nm LED cluster heads (MedX Health, Toronto) were placed over the L and R ears during the last 4 minutes of the treatment session. LEDs are FDA-cleared, non-significant risk. Both the participant and the person performing the treatment wore goggles that block red wavelength (from the red intranasal). Real or Sham LED devices look and feel identical.

Participants in this group first receive a series of 15 sham LED treatments with the helmet and intranasal devices containing sham LEDs (no photons were emitted): Transcranial LED Helmet and two intranasal nose clips (sham LEDs, Vielight, Inc., Toronto), 28-minute treatment, 2 days/ week, 7.5 weeks, at least 48 hours between treatments. Additionally, two sham LED cluster heads (MedX Health, Toronto) were placed over the L and R ears during the last 4 minutes of the treatment session. These participants were offered an optional Second Real Series of identical 15 real LED treatments. Both the participant and the person performing the treatment wore goggles that block red wavelength (from the red intranasal). Real or Sham LED devices look and feel identical.

PhotoMedex Helmet: 12 Pods, 4.5 x 4.8cm (21.6 cm2); 20 diodes; 830nm; Power Output (PO) 665.3 mW; Power Density (PD)/LED 30.8 mW/cm2; CW; Time/session 28m 10s*; Dose: Energy Density 26 J/cm2. MedX Device: 1 Cluster Head/ear, 2 inch diameter (22.48 cm2); 870nm; PO 90 mW; PD/LED 4 mW/cm2; CW; Time/session 4m; Energy Density 4 J/cm2; 2 Vielight Intranasal Devices: Red, 633nm diode; PO 6.5 mW; PD/LED 7.6 mW/cm2; CW; Time/session 25m; Energy Density 11.4 J/cm2 and NIR, 810nm diode; PO 6.5 mW; PD/LED 7.6 mW/cm2; Pulse frequency 10Hz; Duty Cycle 50%; Time/session 25m; Energy Density 11.4 J/cm2 Thor Photomedicine Helmet: 10 Pods, 6cm diameter (28.3 cm2); 35 diodes; Wavelength 850nm; PO 1050 mW; PD/LED 50 mW/cm2; CW; Time/session 23m, 24s; Energy Density 26 J/cm2 2 Placement sets: 1) midline, 6 cluster heads; 2) L&R side of head 12 cluster heads (14m, 5s, each set); 1) midline, 5 cluster heads; 2) L&R side of head 10 cluster heads (11m, 42s, each set) CW, Continuous Wave

Repeat a series of numbers, exactly as spoken, in the same order. Starts with 2 digits and continues to 7 digits in a row. Participant is given 2 attempts at each level. If 0/2 attempts are correct for a given level, testing is discontinued. Scoring: 2 points both trials per set; 1 point 1 trial; 0 points failed both trials Forwards; measures Attention/Executive Function with a min score of 0 and a max score of 14; higher score = better outcome

Repeat a series of numbers, in reverse order. Starts with 2 digits and continues to 7 digits in a row. Participant is given 2 attempts at each level. If 0/2 attempts are correct for a given level, the test is discontinued. Scoring: 2 points both trials correct for each set; 1point 1 trial correct; 0 points no trials Backwards; measures Attention/Executive Function with a min score of 0 and a max score of 14; higher score = better outcome

Total (Forwards+Backwards); measures Attention/Executive Function; working memory with a min score of 0 and a max score of 28; higher score = better outcome

Delis-Kaplan Executive Function (DKEFS) Trails 2; Number Sequencing; measures Attention/Executive Function Reaction Time with lower = better outcome Connect the numbers in consecutive order from 1 to 16, as quickly as possible.

Delis-Kaplan Executive Function (DEKFS) Trails 4; Number-Letter Switching; measures Attention/Executive Function Reaction Time lower is better Connect numbers and letters, alternating between numbers and letters. (1-A-2-B...) as quickly as possible. Ends at P.

Color-Word Interference Test Trial 3; measures Attention/Executive Function; inhibition Reaction Time lower = better outcome Name the color ink each word is printed in for a series of words, as quickly as possible.

Color-Word Interference Test Trial 4 (Inhibition/Switching); measures Attention/Executive Function Reaction Time lower = better outcome Participants must sometimes name the color ink in which the word (color) is printed or, if there is a box around the word, they must read the word.

CVLT-II Total Trials 1-5, measures Learning/Memory over 5 consecutive trials; with a min score of 0 and a max score of 80; higher score = better outcome Participant must immediately recall words from a 16 word list A read aloud by the examiner. The same list is presented for each of the 5 trials. Total across 5 trials is scored.

California Verbal Learning Test -II, Short Delay Free Recall, measures Memory; with min score of 0 and a max score of 16; higher score = better outcome A second list B is presented for one trial of recall before asking participants to recall items from the first list A. Number of items recalled from list A is scored.

California Verbal Learning Test -II, Short Delay Cued Recall, measures Learning/Memory; with a min score of 0 and a max score of 16; higher score = better outcome Participants are now asked to recall items from list A again, but are given cues by category. (4 categories, 4 items each). Total is score across all 4 categories.

California Verbal Learning Test- II, Long Delay (20 min) Free Recall, measures Memory with a min score of 0 and a max score of 16; higher score = better outcome After a 20 minute delay (other tasks run during this delay), participant is asked to recall items from List A. Total is number of words recalled.

California Verbal Learning Test- II, Long Delay (20 min) Cued Recall measures Memory; with a min score of 0 and a max score of 16; higher score = better outcome After 20 minute delay participant is given cues (4 categories) and asked to recall the words within each category. Total number of items recalled across all categories.

CPT; False Alarms (Administered on laptop computer) measures Selective and Sustained Attention percent errors; lower = better outcome Participant presses space bar when an X appears on screen. False alarms is number of errors made (responding to non-targets), reported as percentage.

Conner's Continuous Performance Test II; Mean Correct Detections, Reaction Time, (Administered on laptop computer) measures Selective and Sustained Attention; shorter = better outcome Participant must press the space bar when they see an X. Reaction Time for correct responses only.

Conner's Continuous Performance Test II (CPT); D' (sensitivity index) (Administered on laptop computer) measures Selective and Sustained Attention; ability to detect the signal (x) within noise; values range -2 to 2; higher value better Value d' is a measure of the difference between the signal (X) and noise (non-target letters) distributions, assessing the participant's discriminative power.

ROCFT; Immediate Recall measures Visuospatial memory with min 0 and max 36; higher score better outcome Participant is presented with a figure. They are asked to copy the figure. Then they are immediately asked to recall and draw the figure. Figure is scored on 18 units; up to 2 points per unit Scoring based on accuracy and placement of figure parts: A score of 0, 0.5 recognizable and incorrectly placed, 1 accurate or correctly placed or 2 both accurate and correctly placed; 0 is given if part is omitted or unrecognizable.

Rey Osterrieth Complex Figure Test (ROCFT); Delayed (20-min) Recall measures Visuospatial memory with min 0 and max 36; higher = better outcome Participant is asked to draw the figure from memory after a 20 minute delay (other tasks completed during this time). Figure has 18 units; Scoring is based on accuracy and placement: 0 if unrecognizable or omitted, .5 recognizable, 1 either accurate or correctly placed or 2 both accurate and correctly placed, per unit.

BDI; 21 item Questionnaire; Mood. With min 0 and max 63; 0-13 minimal; 14-19 mild, 20-29 moderate, 30-63 severe; lower scores = better outcome Score per item: 1-3

Post Traumatic Stress Disorder 17-item, symptom questionnaire; higher scores indicates greater endorsement of symptoms with min 17 and max 85; Reliable decrease 5-10; Clinically meaningful decrease 10-20 points (Monson et al., 2008); Lower scores = better outcome Score ranges from 1-5 for each item.

Visual Analog Pain Rating Scale (VAS); Current Pain Rating scale of 1-10; lower score = better outcome *must be lower than 7 at entry into study for study inclusion Participant rates current level of pain (includes body, muscular, headache etc) by marking a line on a continuous scale from 1-10.

Short Form McGill Pain Questionnaire; Pain Rating (for the last 30 days) with min 0 and max 45; lower = better outcome Participant rates 15 types of pain (None = 0, mild = 1, moderate = 2, severe = 3). Totaled for all types of pain.

Global Sleep Score with min score of 0 and max score of 27; lower score = better outcome 9 item questionnaire- make up 7 component scores (subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medications, and daytime dysfunction) with a min of 0 and max of 3 per component. Global score is total of the 7 component scores.

Measure of Sleepiness; Participant indicates current level of sleepiness on a scale of 0 extremely alert to 10 extremely sleepy; lower score = better outcome

36 Question General Health Survey/Questionnaire for Veterans; Physical Composite Score (PCS) measures endorsement of physical health symptoms related to physical problems which affect 7-areas (physical functioning, role physical, bodily pain, general health, vitality, social functioning, role emotional, mental health); Z-Score; lower = better outcome.

36 Question General Health Survey/Questionnaire for Veterans; Mental Composite Score (MCS) measures endorsement of mental health symptoms which affect 7-areas (physical functioning, role physical, bodily pain, general health, vitality, social functioning, role emotional, mental health); Z-Score; lower score is better (based on scoring of the Veterans RAND 36 item Health Survey (SV-36) formerly Veterans SF-36.

Measures General Fatigue Levels; lower = better; General Fatigue Subscale = sum of items numbers 1, 5, 12 and 16 (max = 5, min =0 for each item) Max possible sum is 20.

Physical Score Subscale Measures Physical Fatigue Levels; lower = better outcome; sum of items numbers 2, 8, 14 and 20 (max = 5, min =0 for each item)

Reduced Activity Subscale measures fatigue levels during activity; lower = better outcome; sum of items numbers 3, 6, 10 and 17 (max = 5, min =0 for each item) Max possible sum is 20.

Reduced Motor Activity subscale measures if fatigue affects motivation levels; lower score = better outcome; sum of items numbers 4, 9, 15, and 18 (max = 5, min =0 for each item) Max possible sum is 20.

Mental Fatigue subscale measures mental fatigue levels; lower = better outcome; sum of items numbers 7, 11, 13 and 19 (max = 5, min =0 for each item) Max possible sum is 20.

Inclusion Criteria: Participants were recruited from the participants in a Department of Defense (DoD) study of a longitudinal cohort of Gulf War Veterans who returned from their deployment in 1991 through Ft. Devens, MA. This cohort has been followed at multiple time points since the end of war (Proctor et al., 1998), and through the VA Informatics and Computing Infrastructure (VINCI)/Corporate Data Warehouse (CDW) database, with approval from NDS. The San Francisco VA Medical Center (SF VAMC) was a second site on the study (35 Veterans were run in Boston; 12 were run in San Francisco on the same protocol.) Those recruited from the VINCI/CDW database, resided within a 25 mile radius of the Boston VA Healthcare System (VABHS) or 25 miles of the SF VA Medical Center (VAMC). The Institutional Review Board at the VA BHS and the SF VAMC (University of California, San Francisco) approved the study. In accordance with the Declaration of Helsinki, Informed Consent and HIPAA authorization was obtained from all participants. Participants answered 'Yes' to the following questions: 1) Difficulty concentrating; and/or 2) Difficulty remembering recent information. Must be a Veteran deployed in 1990-1991 Gulf War, in the Kuwait Theatre Meets criteria for GWVI as defined by "Symptom Questions used to identify Gulf War Illness by Kansas Case Definition, and Chronic Multisymptom Illness by Fukuda Case Definition" (Steele, 2000; Fukuda et al., 1998). Participants must have the presence of 1 or more chronic symptoms (lasting >6 months) from at least 2 of 3 symptom categories from Fukuda et al., (1998): 1) musculoskeletal (muscle pain, or joint pain, stiffness); 2) mood-cognition 3) fatigue. Ages 38 - 65 years Must be physically able to travel to the VA Boston Healthcare System, Jamaica Plain or San Francisco VA Medical Center, for Neuropsychological testing and transcranial LED treatments Must meet screening criteria from the Eligibility Screening: The following Neuropsychological (NP) tests were administered at Screening: Trail Making Test A & B (Reynolds, 2002); Controlled Oral Word Association Test (COWAT (FAS); Spreen & Benton, 1977; Benton and Hamsher, 1989); California Verbal Learning Test - II (Delis et al., 2000); Color-Word Interference Test (Stroop; Delis, Kaplan, Kramer, 2001). Additional screening tests included: Short Form McGill Pain Questionnaire (Melzac, 1984); Overall VAS current pain rating (0-10); and the PTSD Checklist- Civilian (PCL-C, Weathers et al., 1994). Participants were required to score at least 2 SD below the standardized norm (age, education, gender) on at least 1 NP screening test or 1 SD below the standardized norm on at least 2 NP screening tests. The Word Reading Subtest from the Wide Range Achievement Test-4 (Wilkinson and Robertson, 2006) was used to estimate premorbid level of cognitive functioning. The SD for each participant on each NP screening test was adjusted by his/her estimated premorbid cognitive level. The Test of Memory Malingering (TOMM, 1996) was administered. Participants who failed Trial 2, or Trial 1 and 2 were excluded from the study. If a participant failed Trial 1, but did not fail Trial 2, he/she was not excluded if he/she showed evidence of poor learning on other NP screening tests such as the CVLT (Schroeder et al., 2013, Arch Clin Neuropsych) Participants were required to have a level of pain 7/10 or less on the VAS and less than 38/50 on the McGill pain questionnaires at screening, as pain has been shown to influence cognition (Moriarty et al., 2011, Review). Exclusion Criteria: Presence of a neurodegenerative disease such as ALS, Parkinson's, Dementia Presence of a life-threatening disease such as cancer Presence of a severe mental disorder such as schizophrenia, or bipolar depression (not associated with PTSD) Current substance abuse or active treatment within last 6 months

Martin PI, Chao L, Krengel MH, Ho MD, Yee M, Lew R, Knight J, Hamblin MR, Naeser MA. Transcranial Photobiomodulation to Improve Cognition in Gulf War Illness. Front Neurol. 2021 Jan 21;11:574386. doi: 10.3389/fneur.2020.574386. eCollection 2020.