Circadian re-set repairs long-COVID in a prodromal Parkinson’s parallel: a case series

Abstract

Background

In this case series, results from daily visual exposure to intense polychromatic light of 2000 to 4000 LUX is presented. Bright light treatment is a standard procedure for treating seasonal affective disorder and prodromal Parkinson’s disease with high success. With the post-encephalitic symptoms of long-COVID closely approximating those of prodromal Parkinson’s disease, we treated insomnia and sleep-related parameters in these patients, including total sleep, number of awakenings, tendency to fall back to sleep, and fatigue, to determine whether mending sleep could improve quality of life.

Case presentation

We present three female and two male Caucasian patients aged 42–70 years with long-COVID that persisted from 12 weeks to 139 weeks after contracting coronavirus disease.

Conclusion

A light presentation protocol was adapted for long-COVID that not only restored sleep in all patients, but also unexpectedly repaired the depression, anxiety, and cognitive changes (brain fog) as well. A robust pattern of recovery commencing 4–5 days after treatment and was maintained for weeks to months without relapse. These preliminary findings represent a novel, minimally invasive approach for managing the most debilitating symptoms of long-COVID, making it an ideal candidate for the drug hypersensitive, post-encephalitic brain. That a compromised circadian mechanism seen in Parkinson’s disease may also underlie post-encephalitic long-COVID implicates a compromised role of the circadian system in these disorders.

Background

The prodromal symptoms [1] of insomnia, fatigue, depression, anxiety, and cognitive change observed decades prior to the development of Parkinson’s disease (PD) [2–6] are thought to be mediated by circadian disruption [7, 8]. Repair of circadian function is achieved by bright light presentation at critical times during the light/dark cycle [9–11]. The symptoms of long-COVID are strikingly similar to those of idiopathic PD [12, 13, compare 10, 14] in post-encephalitic brain and may signal the onset of viral Parkinsonism [15–17].

Recent epidemiological work [18] is consistent with findings from longitudinal studies demonstrating remission of prodromal symptoms of PD [19] after light treatment (LT) resynchronizes the circadian cycle [20], with the slow, incremental rate of dopamine (DA) degeneration paralleling that of homeostatic circadian regulation [21]. A similar incremental rate of symptom development in post-encephalitic brain is reported to have occurred after the 1919 pandemic [22]. In the presence of this background [10, 19, 23], we implemented LT to treat patients with long-COVID on a case-by-case basis to determine whether this would help in repairing their sleep disturbance.

Case report 1

A 58-year-old Caucasian female patient presented with fatigue, depression, anxiety, and irritability. She reported an inability to fall asleep, stay asleep, getting back to sleep, and getting adequate sleep to remain functional for the duration of waking hours (Fig. 1C). The intractability of her sleep state increased anxiety, producing a circular exacerbation of insomnia. Periodic internal tremor, joint pain, and loss of appetite (Fig. 1B) were also present. Her history revealed hereditary elliptocytosis and a minor surgical procedure. She was not on any long-term medications and reported allergies to morphine and codeine. A diagnosis of COVID was made 20 weeks prior to presentation with five acute symptoms lasting for 1–15 weeks prior to commencing LT (Fig. 1A). Figure 1B illustrates the effect of LT on sleep and psychiatric parameters measured at the different timepoints for the duration of LT. Note that sleep deteriorated for the first 3 days after LT was applied in the evening. The time of LT administration was then shifted to 7:30 A.M., whereby all sleep and psychiatric parameters improved slightly to moderately within the first 1.5 weeks. Improvement continued after the third week with all parameters showing moderate to marked improvement by that time.

Fig. 1.

A The principle acute (black bars) and extended long-COVID (solid grey bars) symptoms occurring after case 1 contracted COVID. Diagnosis was confirmed within 3 days after the acute symptoms first appeared and these are listed in the right-hand column. The duration of symptom expression is represented on the abscissa with the units of time expressed as weeks. The solid bars represent severe symptoms while diagonal bars represent times when symptoms were slight to moderate. The open bars represent the time over which LT was applied. The time at which the acute symptoms first emerged is indicated by the black arrow at the top-left of the diagram. B Symptomatic expression of case 1 during the acute and chronic phases of COVID. The status of each symptom displayed after contracting the virus and prior to attending the clinic are expressed in the left-hand column and is labelled acute (pre). When the patient commenced treatment with evening LT (E) at 20 weeks extending over a 5-day period, her sleep deteriorated. The time of LT administration was then changed to the morning (M) which was continued to the end of the observation period at 76 weeks. Sleep and related parameters are expressed in the top row of boxes and measured at the times listed at 20, 22, 23, 27, and 76 weeks post-COVID (p/c). The total amount of sleep per night is expressed in hours as are the number of awakenings, while the tendency to fall back to sleep after awakening at night is rated as rarely, sometimes, frequently or always. The psychiatric symptoms are listed in the second row for the five periods of observation over the 76-week period. Fatigue and the psychiatric symptoms observed during the acute period were evaluated and then rated as being slight (SLT), moderate (MOD), or marked (MKD) in severity. The ratings during the treatment period were based on the change in symptom severity compared with that of the acute period and were designated as unchanged (=), slightly improved (+), moderately improved (++), or markedly improved (+++), while (−) indicated that the symptom worsened. B. fog/cog, brain fog/cognition; Irrit., irritability; Int. trem., internal tremor; nd, not detected; pre, before LT commenced; pc, post-COVID; p/t, post-treatment. (1) indicates that knee pain existed prior to contracting COVID. C The sleep diary entries for case 1 for 12 days prior to and 19 days after commencing LT. The dark squares represent the time spent sleeping while the unfilled blocks represent the time awake. The left two columns represent the day in the program and the actual date, from left to right, respectively. The numbers across the top of the chart represent the hours of the day (black numbers on white) versus the night (white numbers on black). Evening LT commenced 5 days before the switch to morning light application. The switch to morning light was made because sleep deteriorated with application of LT in the evening. Note that the longer that LT was applied, the more consolidated sleep became

Case report 2

A 70-year-old retired Caucasian male patient presented with bradyphrenia and reported that sleep (Fig. 2C) during the past 12 weeks was marked by increased awakenings (Fig. 2B,C). He expressed concern about the fatigue interfering with his ability to undertake everyday tasks, as he became tired easily. Psychic depression and brain fog were also present, with suspected PD including bradykineasia, muscular rigidity, tremor of the left hand, shortened gait, and reduced arm swing. Anosmia and impaired balance were also present and he stated “I don’t feel like myself…”. His history revealed benign prostatic hypertrophy and psoriasis, for which he was prescribed Duodart® and topical Daivobet®, respectively. Diagnosis of COVID was made 13 weeks earlier with a mild cough and sore throat being the main acute symptoms, lasting 1–2 weeks, respectively (Fig. 2A). All other symptoms developed during the acute stages of COVID but all symptoms responded well to the application of LT, with the exception of Reynaud’s disease and anosmia, both of which the patient “suspected” were present before contracting COVID. All other long-COVID and PD symptoms improved with LT, with depression showing the most dramatic improvement.

Fig. 2.

A The principle acute (black bars) and extended long-COVID (solid grey bars) symptoms occurring after case 2 contracted COVID. Diagnosis was confirmed within 3 days after the acute symptoms first appeared and these are listed in the right-hand column. The legend for this diagram is the same as that expressed in Fig. 1A. The asterisks represent the only Parkinsonian symptoms that were present at the time of disease onset. B Symptomatic expression by case 2 during the acute and long phases of COVID. The status of each symptom displayed after contracting the virus and prior to attending the clinic is expressed in the left-hand column and labelled acute to 13 weeks (pre). When the patient commenced treatment with evening LT (E) at 14 weeks post-COVID (pc), it was then continuously administered for 53 weeks. Parameters are expressed in the top rows of boxes and measured at the times listed as 14, 16, and 55 weeks p/c. The psychiatric symptoms are listed in the second row for the three periods of observation over the 55-week period. For Parkinsonian features, Brady., bradykinaesia; Rigid., rigidity; Trem., tremor; A. swing, arm swing; F. L. Self, feeling like oneself; Olfac., olfaction; nd, not detected; pre, before LT commenced; pt, post treatment. The legend for sleep parameters, psychiatric measures, and the fatigue, as well as the code for change in symptom severity, are the same as those expressed in Fig. 1. (1) indicates that the patient experienced back pain. C The sleep diary entries for case 2 for 2 days prior to and 13 days after commencing LT. The dark squares represent the time spent sleeping while the open squares represent times awake. Evening LT (grey diagonal banded arrow) commenced on the second day after entering the program. Note that the longer that LT was maintained, the more consolidated sleep became. The left two columns represent the day in the program and the actual date, from left to right, respectively. The numbers across the top of the chart represent the hours of the day (black numbers on white) versus the night (white numbers on black). The asterisk represent a day when the light was not used and the patient forgot to record the time but indicated that the sleep on this day was the same as the preceding and subsequent day, thus the mean of the two days was used

Case report 3

A 42-year-old Caucasian female patient, visiting the clinic from a foreign domicile, presented with insomnia and broken sleep (Fig. 3c). She was getting 6 hours of sleep per night with two awakenings and could not fall back to sleep readily. Sustained fatigue was present during the day, with marked anxiety in response to her inability to work more than a few hours per week. She was also depressed, exhibiting cognitive impairment (brain fog) and other symptoms (Fig. 3B). Past history revealed a tibial fracture, otitis media, and panic attacks, while she was prescribed 1.53 mg twice a day intranasal estrogen and 100 mg progesterone DPO for premenopausal insomnia at the time of admission. A diagnosis of COVID was made 139 weeks earlier, with a second infection 76 weeks later and a relapse at 118 weeks. The acute symptoms included sore throat, respiratory inflammation, headache, and tinnitus, all of which were present in varying intensities at the time from the first infection up until the commencement of LT (Fig. 3A). The duration of long-COVID symptoms were also present in severities that varied in accordance with reinfection and relapse. Leg pain did not appear until after the relapse and was present when LT commenced. As shown in Fig. 3C, the sleep architecture of this patient gradually changed during the course of LT. Note that just prior to this, the patient was getting a total of 5–6 hours of sleep per night with awakenings lasting 2–3.5 hours. Within the first 3 weeks of treatment, the duration of time awake ranged from 0.5 to 1 hour in most instances. By the end of the period of observation at 5 weeks after commencing treatment, there was one awakening per night lasting no longer than 0.5 hours. From this, one could surmise that LT dramatically improved sleep architecture, demonstrating a gradual process typically taking weeks to months. This patient showed moderate improvement in all symptoms by the first week and marked improvement in fatigue, depression, cognition, and anxiety by the second week post-treatment. This improvement was maintained for the duration of the study up to 24 weeks.

Fig. 3.

A The principle acute (black bars) and extended long-COVID (solid grey bars) depict symptoms occurring after case 3 contracted COVID. Diagnosis was confirmed 2 days to 6 weeks after the acute symptoms first appeared and these are listed in the right-hand column. The time at which the acute symptoms first emerged after the first and second infections is indicated by the black arrow at the top of the diagram as is the time of onset of the relapse. The legend for this diagram is the same as that expressed in Fig. 1A. B Symptomatic expression of case 3 during the chronic phase of COVID. The status of each symptom displayed after contracting the virus and prior to attending the clinic are expressed in the left-hand column and labelled acute; 0–139 weeks post-COVID (p/c). The patient commenced treatment with evening morning LT at 140 weeks (M) and continued for 2, 5, 11, 19, and 24 weeks p/t until the end of the study at 164 weeks post-treatment (p/t). Sleep and related parameters are expressed in the top rows of boxes and measured at the times listed as weeks (p/c). Sleep and psychiatric symptoms are listed in the first and second rows for the five periods of observation over the 164-week period. The legend for sleep parameters, psychiatric measures, and fatigue, as well as the code for change in symptom severity, are the same as those expressed in Fig. 1. (1) indicates that patient experienced leg pain. C Case 3 sleep diary entries for 3 days prior to and up to 32 days after commencing LT. The dark squares represent the time spent sleeping while the unfilled blocks represent the time awake. The two columns on the left represent the day in the program and the actual date, from left to right, respectively. The numbers across the top of the chart represent the hours of the day (black numbers on white) versus the night (white numbers on black). Morning LT (grey arrow) commenced 3 days after entering the program. Note that sleep was not recorded during the 10 days marked as overseas travel, prior to returning to her home domicile. Note also that the longer that LT was applied the fewer the number of awakenings and the shorter their duration as sleep became more consolidated

Case report 4

A 64-year-old Caucasian male patient presented with a sleep pattern reporting a 6–7-hour block of continuous sleep per night. The time of falling off to sleep varied from 11:00 to 06:00 hours with wake time fluctuating from 06:00 to 11:30 hours (Fig. 4C). He reported 1–2 awakenings per night and being unable to fall back to sleep if awakened. Moderate fatigue, depression, brain fog, anxiety, and irritability were noted (Fig. 4B), which he attributed to his irregular sleep pattern and inconsistent time of sleep onset. Past history revealed elevated Prostate Specific Antigen (PSA) with acute adenocarcinoma, for which he was a candidate for radical prostatectomy. Inguinal hernia repair and rotor cuff injury occurred 14 and 2 years earlier, respectively. The patient was not on any long-term medication. He was diagnosed with COVID 46 weeks earlier with acute symptoms (Fig. 4A) that included rhinorrhea, headache, hypersomnia, back pain, and anosmia that resolved in 3–4 weeks. While fatigue, depression brain fog, mood, and anxiety were severe on presentation, by week 8 post-treatment they showed a marked improvement, lasting for at least 16 weeks after commencing LT (Fig. 4B). Olfaction showed slight impairment on presentation with moderate improvement for the treatment period.

Fig. 4.

A The principle acute (black bars) and extended long-COVID (solid grey bars) depict symptoms occurring after case 4 contracted COVID. Diagnosis was made 46 weeks earlier and confirmed 1 month after the acute symptoms first appeared and these are listed in the right-hand column. The time at which the acute symptoms first emerged after the infection is indicated by the black arrow at the top of the diagram. The legend for this diagram is the same as that expressed in Fig. 1A. B Symptomatic expression of case 4 during the chronic phase of COVID. The status of each symptom displayed after contracting the virus and prior to attending the clinic are expressed in the left-hand column and labelled acute; 0 to 46 weeks post-COVID (p/c). The patient commenced treatment with evening LT at 54 weeks (M) and was continued until the end of the observation period at 16 weeks post-treatment (p/t). Sleep and related parameters are expressed in the top row of boxes and measured at the times listed weeks p/c. Sleep and psychiatric symptoms are listed in the first and second rows for the three periods of observation over the 16-week period. The legend for sleep parameters, psychiatric measures, and fatigue, as well as the code for change in symptom severity, are the same as those expressed in Fig. 1. (1) indicates the patient experienced back pain. C Case 4 sleep diary entries for 2 days prior to and up to 26 days after commencing LT. The dark squares represent the time spent sleeping while the unfilled blocks represent the time awake. The two columns on the left represent the day in the program and the actual date, from left to right, respectively. The numbers across the top of the chart represent the hours of the day (black numbers on white) versus the night (white numbers on black). Morning LT (grey arrow) commenced 2 days after entering the program. Note also that the longer that LT was applied the more the periods of sleep became aligned during the dark phase of the photoperiod

Case report 5

A 45-year-old Caucasian female patient presented with broken sleep achieving a total sleep time of 3–4 hours per night, frequent awakenings, and difficulty in falling back to sleep (Fig. 5C). She reported severe fatigue, anxiety, and intermittent body pain. Moderate depression, brain fog, irritability, internal tremor, and loss of appetite with a slight impairment of olfaction presenting as phantosmia (Fig. 4B). Her history was generally unremarkable with a tonsillectomy performed 20 years earlier. The diagnosis of COVID was made 43 weeks prior to presentation with minor sore throat and headache lasting 3–4 days (Fig. 5A). Only over-the-counter medications were used for COVID symptoms. While fatigue, body pain, insomnia, and brain fog were present from the time of infection until entering the program, cardiac symptom including palpitations, Postural Orthostatic Tachycardia Syndrome (POTS), irregular heartbeat, and weakness were present from week 15. Neurological signs consisting of heaviness in the legs and numbness on the left side of the face and were present intermittently since the time of infection. Depression did not appear until after week 50 post-COVID. Further, 1 week after commencing LT, all symptoms improved substantially while improvement at weeks 2–5 was marked (Fig. 5B), and the patients resumed normal daily activity, with scheduled daily LT.

Fig. 5.

A The principle acute (black bars) and extended long-COVID (solid grey bars) depict symptoms occurring after case 4 contracted COVID. Diagnosis was confirmed within 6 weeks after the acute symptoms first appeared and these are listed in the right-hand column. The time at which the acute symptoms first emerged after infection is indicated by the black arrow at the top of the diagram. The legend for this diagram is the same as that expressed in Fig. 1A. B Symptomatic expression of case 5 during the chronic phase of COVID. The status of each symptom displayed after contracting the virus and prior to attending the clinic are expressed in the left-hand column and labelled acute; 0 to 43 weeks post-COVID (p/c). The patient commenced treatment with morning LT at 44 weeks (M) and was continued for 5 weeks until the end of the observation period post-treatment (p/t). Sleep and related parameters are expressed in the top rows of boxes and measured at the times listed weeks post-COVID. Sleep and psychiatric symptoms are listed in the first and second rows for the four periods of observation over the 5-week period. The legend for sleep parameters, psychiatric measures, and fatigue, as well as the code for change in symptom severity, are the same as those expressed in Fig. 1. (1) indicates the patient experienced generalized body pain. C Case 5 sleep diary entries for 4 days prior to and up to 26 days after commencing LT. The dark squares represent the time spent sleeping while the unfilled blocks represent the time awake. The two columns on the left represent the day in the program and the actual date, from left to right, respectively. The numbers across the top of the chart represent the hours of the day (black numbers on white) versus the night (white numbers on black). Morning LT (grey arrow) commenced 2 days after entering the program. Note also that the longer that LT was applied the fewer the number of awakenings and the shorter their duration as sleep became more consolidated

In total, five Caucasian patients were the subject of study, with three women and two men, diagnosed with COVID-19 and referred from general practices to assist with their sleep disorders and depression. Patients were not selected for the study but included the first five patients who were admitted to the clinic consecutively, with a previous diagnosis of long-COVID. To protect patient identity during publication, each case was coded with restricted access to code translation. After the first consultation each patient was instructed on how to complete the sleep diary and this served as a baseline. The dose of light administered was 2–4000 LUX for 30 minutes to 1 hour of exposure when the device was positioned within 30° of deviation from the midsagittal plane at a distance of 0.8 to 1.0 m. The patient received both verbal and written instruction on how to set up the light source to optimize phototransduction. LT was well tolerated, with the only adverse event observed during screening was case 1, where it was administered in the evening, at the incorrect time within the 24-hour cycle and sleep deteriorated. However, this was corrected by shifting the time of light administration to the morning and sleep repaired, as established in previous work [23, 24, cf. 25–27].

Details of the sleep features where obtained using three different methods employed for cross-verification and this permitted detailed adherence monitoring. Sleep diary: This involved entering the time asleep, time of awakening, and the number and duration of awakenings during each consecutive 24-hour period, commencing the day after the first consultation. The diary was filled out in the morning, after awakening and the partner/spouse/carer assisted in verifying the accuracy of each daily entry. The time of light use was also recorded and made accessible to the clinician during that time. The second method involved the patient in eliciting a spontaneous response on questions administered during a structured interview regarding the patient’s sleep including, (i) on average, the total number of hours they slept during the dark phase of the light/dark cycle, (ii) the number of awakenings, and (iii) if awoken, did they fall back using the ratings of rarely (1), sometimes (2), often (3), or always (4). Each patient also reported the number and duration of day naps. The third method was to complete a 24-hour circadian record. The clinician administered this during the structured interview to determine the average, (i) time to bed, (ii) time to fall asleep, (iii) time to wake, and (iv) time to arise each day, since the last appointment. Other parameters including time of LT, the timing of drug dosing and if and when a symptom free period occurred during each 24-hour period. The clinician was careful to also specify the time during the 24-hours cycle during which each answer applied. State versus trait features of each parameter and each patient was deciphered compared with when they last attended the clinic with a spouse/partner/carer being present to verify their responses. The patient and spouse/partner/carer was also asked to assist in recalling the number, severity, and duration of acute and chronic symptoms at the onset of COVID occurring weeks earlier. The long-COVID symptoms formally assessed in the clinic included sleep, REM Sleep Behavior Disorder (RSBD), fatigue, depression, cognition (brain fog), anxiety, irritability and mood swings, internal tremor [28], psychotic symptoms, bradyphrenia, taste and smell, visual anomalies, pain, appetite, and body weight. Note that clinical assessment of these parameters has been practiced extensively in this clinic and reported previously for patients with PD [10, 19, 23]. Care was taken to record the patient’s recall immediately after COVID diagnosis, during the progression of the disease, and over the longer term, including the period of treatment.

All clinical symptoms monitored in the present study were rated on a Likert scale that has been used extensively [10, 19, 23, 25] and run parallel with validated instruments in controlled trials [10]. The scale employed is a 10-point scale with 10 being the most severe, translatable to slight (1–3), moderate (4–7), or marked (8–10) in severity. This method also included input from the carer and the patient in the context of professional clinical evaluation. In consideration of the reasonable level of agreement between these results and those published previously, we propose that our method of clinical evaluation is reasonably valid and reliable. Nevertheless, the inherent problems associated with clinical assessment used in the present study are duly recognized and we expect that many of the findings reported here will be interpreted conservatively until further randomized, blinded, placebo-controlled trials are implemented using more formal, validated instruments.

Discussion

The present findings demonstrate that LT used for treating the prodromal symptoms of PD also improves sleep in patients with long-COVID. A remarkable improvement in other long symptoms was also seen in our sample including, depression, fatigue, brain fog, and other long-COVID symptoms intimating parallel circadian involvement in post-encephalitic Parkinsonism and prodromal PD (Figs. 1, 2, 4, 5) [2–6, 10, 19, 23–29]. While it has been suggested that a post-encephalic increase in the incidence of PD may be consequential to the COVID pandemic [4, 30, 31], other work supports such a contention on the basis that neurological disease may eventuate from long-term perturbation of circadian function [18]. This position is advanced further by the present findings whereby LT regimens used for treating prodromal PD are also effective in treating similar debilitating symptoms of long-COVID [18, 22, 23, cf. 32–34].

A parallel between the geographical/climatic incidence of PD and light supports the contention that ambient light may be an important factor in COVID survival and proliferation as darker, winter climates have a higher incidence of COVID, with warmer climates reporting lower rates of infection [35]. Higher latitudes with darker climates [36, 37] and occupations and lifestyles (Fig. 1) requiring light exposure at night [38–41] provide further evidence linking circadian function, long-COVID, and prodromal PD. While other applications of light have been used for treating skin conditions during the COVID pandemic [42–44], this is the first report suggesting that light may play an important role in the etiology and treatment of long-COVID by virtue of its effect upon circadian function.

Another parallel drawn between the symptoms of long-COVID [45, 46] and prodromal PD relates to possible circadian misalignment (Fig. 4) [20, 21, 27], as the time of light presentation would be critical for realigning circadian phase. For most patients with PD with permanent DA deficiency, LT is most effective when applied in the evening [10, 23, compare 26, 27], based on studies suggesting they are phase advanced [10, 19, 23, 47, 48], while patients with seasonal affective disorder are phase delayed and require light in the morning [49]. Under the assumption that patients with long-COVID are also phase advanced, we administered light at night in a shift worker that developed Long-COVID (Fig. 1). However, this caused further fragmentation of sleep, prompting application of LT in the morning, whereby sleep improved. Shift workers that contract COVID have been defined as being predisposed to developing long-COVID [50, 51]. A difference in predisposition is shown in the response where PD was diagnosed shortly after this patient contracted COVID (Fig. 2). Like other patients with PD, a therapeutic response was best achieved with evening light [10, 19, 23]. Even in the presence of reduced sensitivity of patients with PD to the positive effects of treatment [52], the improvement in symptoms in case 2 after evening light is consistent with that of other patients with PD treated with LT [10, 19, 23], and highlights the need for careful evaluation of each patient before prescribing the appropriate parameters of LT exposure.

Patients that experience prolonged durations of acute and/or chronic symptoms after COVID lasting for months to years before commencing LT represent a group needing special consideration during treatment (Figs. 3 and 5). These patients responded well to morning LT, but their high levels of anxiety were often associated with physical disability, which in turn exacerbated their anxiety. The rate of sleep repair remained incrementally progressive, extending over many weeks and progressing in a manner similar to that seen in patients with PD [10, 22, 23, 25]. By comparison, they required an extended LT duration compared with the other patients with long-COVID who were less severe or were symptomatic for shorter periods before commencing LT. These patients are still under study, and the minimal duration of LT treatment is yet to be determined.

It has been suggested that phase shifting and the antiinflammatory properties of melatonin make it a good candidate for treating long-COVID [53–55]. This strategy is based on the finding that COVID-induced inflammation can reduce serotonin in serum and brain [56]. However, the post-encephalitic brain exhibits a 5× hypersensitivity to psychotropic and neuropsychiatric drugs [33, 57], suggesting that any drug being repurposed for long-COVID, including melatonin and selective serotonin reuptake inhibitors (SSRIs) [58], could be problematic [59, 60]. Alternatively, LT is minimally invasive and elevates serotonin and its precursor tryptophan [61], suggesting that this may be the mechanism underlying the therapeutic effect seen in the present findings. It has been suggested further that melatonin could be employed in a dual role as an antiinflammatory and to repair disrupted circadian phase [62] in patients with long-COVID [53–55]. However, with time of melatonin administration being critical in shifting circadian phase [9, 21], its ad hoc administration as an antiinflammatory may abrogate or even worsen the intended therapeutic effect on circadian function (Fig. 1). Additional evidence that melatonin possesses pro-Parkinsonian properties and may exacerbate symptoms of long-COVID is consistent with the antagonistic effect of LT on melatonin in PD [6, 23, 25], and with the already elevated plasma melatonin levels seen in PD and its models [63, 64]. The demonstrated benefits of melatonin receptor antagonists in experimental rodent and primate models of PD [65, 66] support this further. In the presence of the unresolved dilemma that post-encephalitic patients may develop PD [15, 17] and the demonstrated hypersensitivity to psychotropic, anti-viral, and Parkinsonian drugs often present in these patients [33, 57, 67, 68], a more conservative approach to drug repurposing should be encouraged [58]. Even in the presence of antifibrotic and antiinflammatory compounds currently under development that show promise for treating long-COVID [53–55], LT remains the ideal candidate until better adjunctive treatments are developed.

The parallel between long-COVID and prodromal PD described herein is based on three converging theoretical points. (i) The symptoms of long-COVID and prodromal PD are strikingly similar and are indicative of compromised circadian function [5, 6, 25, 26, 69]. Those symptoms include sleep disturbance, fatigue, depression, anxiety, and cognitive change [19, 27, 29, 70]. (ii) The occurrence of post-encephalitic Parkinsonism was endemic after the 1919 pandemic with similar symptomatic features to those of long-COVID [8, 22, 33]. (iii) With LT being effective in repairing the circadian-based features of idiopathic PD [10, 23–27], we hypothesized that post-encephalitic circadian features might be equally sensitive to the therapeutic effects of LT [71]. From this perspective, we discovered that LT is highly effective in repairing the fractured circadian underlay of long-COVID as it is in PD [8]. Indeed, the greatest diagnostic challenge of the present study was to determine the critical time of LT (Fig. 2), whereby long-COVID and PD developed simultaneously. However, the 4-day stabilization period of sleep pattern utilized previously [10, 19] and its implemented here serves to define the most effective time of LT application in patients with PD and long-COVID alike. In this regard, this preliminary study and the underlying hypothesis, provide a new direction [71] for exploring a noninvasive therapy [19, 26, 47], a diagnostic tool [70], and a potential prophylactic [18, 34] for long-COVID.

Conclusions

To date there is no treatment yet identified that has a significant impact on the most debilitating symptoms of long-COVID. In consideration of the similarities between the prodromal symptoms of PD and long-COVID, LT application of once daily for the purpose of mitigating sleep disturbance and depression was undertaken. Not only did sleep and its architecture improve after LT but fatigue, depression, anxiety, cognitive function (brain fog), and other symptoms also improved markedly within 1 week after starting LT and continued as long as LT was maintained. Some patients were able to discontinue LT within a few weeks after commencing without relapse. These results represent preliminary findings regarding a novel, noninvasive means of improving many debilitating symptoms of long-COVID, intimating circadian involvement in post-encephalitic brain. This work provides a new direction for effectively treating long-COVID.

Author contributions

GLW was involved in study conception and patient treatment and manuscript preparation. TE and SS were involved in treatment development and manuscript preparation. All authors read and approved the final manuscript.

Funding

No specific funding source served to support this project.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

While this was a retrospective study and ethics approval was not required, informed consent was obtained prior to commencing treatment. A copy of written consent is available for review by the Editor-in-Chief of this journal.

Consent for publication

Written informed consent was obtained from the patients for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.

Competing interests

The authors declare that they have no competing interests.