Circadian Rhythms of Sleep–Wake, Thermoregulatory and Cardiovascular Function in Progressive Supranuclear Palsy

Abstract

Background

The sleep–wake cycle and the autonomic nervous system (ANS) can be impaired in neurodegenerative diseases.

Objectives

To describe sleep–wake cycle and circadian rhythms of body core temperature (BcT) and cardiovascular parameters in Progressive Supranuclear Palsy (PSP).

Methods

We prospectively recorded 48 hour video‐polysomnography, BcT and blood pressure (BP) in 14 PSP patients (disease duration 5.9 ± 2.6 years) under controlled environmental conditions. We analyzed wake–sleep parameters and state‐dependent modulation for BcT, BP and heart rate from the last 24 h of recording. Twelve healthy younger controls were used for comparison.

Results

Patients slept less than 5 hours/night with frequent awakenings, daytime naps were shorter than 60 minutes and mainly represented by light sleep resulting in a 24 h sleep deprivation. Patients with PSP slept less compared to younger controls and normative values for age. Only two patients reported excessive daytime sleepiness. There were no significant differences in sleep parameters according to the presence of sleep disorders. Compared to controls, patients showed significantly higher BcT values during wake and sleep. Twelve patients presented an abnormal BP pattern at nighttime.

Conclusion

PSP patients experienced profound sleep deprivation across the 24 h study period. An imbalance between sympathetic and parasympathetic function with sympathetic predominance may be associated with this alteration. Whether this autonomic dysfunction is the primary drive for the sleep–wake cycle disruption or the consequence of sleep loss needs to be elucidated.

Increasing evidence has shown that a disruption in sleep–wake cycle is associated to neurodegenerative diseases. ¹ Sleep–wake cycle is regulated by a circadian and homeostatic drive, both controlled at the hypothalamic level and depending on the integrity of the autonomic nervous system (ANS). Sleep and the ANS are closely related: nuclei of the central autonomic network are connected to pathways in the brainstem and basal forebrain involved in wake–sleep cycle regulation and autonomic functions are modulated with wake and sleep stages. ²

A single study showed that patients with Progressive Supranuclear Palsy (PSP) present diminished sleep at night with no propensity to sleep during the day, leading to sleep deprivation over 24 hours suggesting a diminished homeostatic sleep drive. ³ Few studies have investigated the ANS in PSP so far, ⁴ leading to conflicting results.

We aimed to describe the sleep–wake cycle and the circadian rhythms of body core temperature (BcT) and cardiovascular parameters under controlled environmental conditions in PSP.

Methods

Participants

Fifteen patients with probable PSP (14 with Richardson's syndrome and one with predominant parkinsonism) ⁵ were prospectively recruited. All patients were recruited from the Movement Disorders clinic at our Institution and assessed by a movement disorder specialist. None of them had concomitant severe medical conditions that could have affected results.

The study was approved by the Ethics Committee of Area Vasta Emilia Centro (reference number 18141), and was performed in accordance with the Declaration of Helsinki and all patients gave written informed consent. Twelve healthy control subjects that had previously performed the same protocol and had given written informed consent to personal data processing for research purposes were used for comparison.

Study Protocol

Upon admittance, patients underwent general and neurological examination including cognitive assessment with the Mini‐Mental State Examination (MMSE) ⁶ and PSP rating scale (PSPRS). ⁷ Subjective sleep quality and daytime sleepiness were assessed by the Pittsburgh sleep quality index (PSQI) ⁸ , ⁹ (global score >5 poor sleeper) and the Epworth sleepiness scale (ESS) ¹⁰ (score > 10 excessive daytime sleepiness). Screening for Restless legs syndrome (RLS) was performed with International RLS Study Group criteria ¹¹ and severity was assessed using the RLS rating scale ¹² (mild severity score 0–10; moderate 11–20; severe 21–30; very severe 31–40). Patients and caregivers underwent a semi‐structured interview for episodes suggestive of REM Sleep Behavior Disorder (RBD). ¹³ In the afternoon patients received an enema for rectal temperature monitoring and were catheterized for the whole protocol duration. On the following morning, sleep–wake cycle, BcT, blood pressure (BP) and heart rate (HR) were monitored for 48 h under controlled environmental conditions, according to standardized procedures ¹⁴ (Supplementary).

Subjects were allowed to sleep ad libitum, living in a temperature (24 ± 1°C) and humidity (40–50%) controlled room, lying in bed except when eating, in a light–dark schedule (dark period: 23:00–07:00) and were placed on a 1.800 kcal/day diet, divided into three meals (08:00, 12:00, 18:00) and three snacks (10:00, 16:00, 23:00). From midnight preceding the monitoring, subjects were instructed to avoid alcohol and caffeinated beverages and to abstain from smoking. Before starting the recording, patients also underwent a complete battery of cardiovascular autonomic testing ¹⁵ (data in supplementary materials).

Sleep–Wake Cycle Parameters

The 24 h sleep–wake cycle was visually scored in 30‐s epochs according to the American Academy of Sleep Medicine (AASM) criteria ¹⁶ as NREM sleep 1, 2 and 3 (N1, N2, N3) and REM. Light‐off period was considered from 23:00 h to 07:00 h; however, if the patient fell asleep before 23:00 h or woke up after 07:00 h the reference time was changed accordingly to include the actual sleep time. Time in bed (TIB), total sleep time (TST), sleep efficiency (SE: TST/TIB * 100, normative value of our laboratory is >85%), duration and TST percentage of NREM and REM stages were calculated over the 24 h. Reference values for each sleep stage according to age were taken from published literature. ¹⁷ Latency to sleep onset was estimated for patients that fell asleep after 23:00 h. We further computed latency to REM sleep, number of REM episodes, duration and percentage of wake after sleep onset (WASO), number of awakenings, arousal index (AI: number of arousals/hour), PLMS index (PLMSI: number of periodic limb movements/hour, cut‐off for pathological >15 ¹⁸ ), percentage of REM with tonic and phasic muscle activity, apnoea‐hypopnea index (AHI: number of obstructive and mixed apnoea and hypopnea/hour, mild 5–15, moderate 15–30 and severe >30 events/hour ¹⁹ ). All data recorded were analyzed, and for the final analysis we used data from the second night (the first night allowed the patient to adapt).

Body Core Temperature Parameters

Day‐Night Pattern of BcT

Mean values of BcT during daytime (from 07:00 h to 23:00 h) and night‐time (from 23:00 h to 07:00 h) were calculated. Nocturnal BcT decline was determined by calculating the difference between night‐time and daytime values (∆BcT).

24‐Hr Circadian Rhythm of BcT

Rhythmicity was analyzed by evaluating time series for BcT according to the single cosinor method, using a computerized procedure (Chronolab). ²⁰ This procedure elaborates the raw temperature curve obtained across the 24 h period using a specific mathematical curve‐fitting method (cosine curve), and it determines whether there is a rhythm within this period and evaluates the following parameters within their 95% confidence limits: (1) mesor (Midline Estimating Statistic of Rhythm), the mean value of the cosine function; (2) amplitude, the difference between the maximum value measured at the acrophase and the mesor of the cosine curve used to approximate the rhythm; and (3) acrophase, the interval between midnight hour (reference time) and time of highest value of the cosine function. Circadian rhythmicity is indicated by the presence of a significant cosinor fit, with the strength of the rhythm being represented by the amplitude or the percentage of variance explained by the model. The acrophase indicates the timing of peak of activity (phase), hence exploring the function of the suprachiasmatic nucleus of the hypothalamus (SCN). Conversely, mesor and amplitude are representative of the homeostatic thermoregulatory process regulated by the preoptic/anterior area of the hypothalamus. ²¹ The 24 h rhythmicity of the last 24 h of recording is reported.

State‐Dependent Analysis of BcT

The mean value of BcT in wake and each sleep stage (N1, N2, N3 and REM sleep) was determined over the 24 h period. The difference (Δ) between the mean value of BcT in every sleep stage during dark period and in wake, considered as reference value, was also calculated.

Blood Pressure and Heart Rate

We calculated systolic BP (SBP), diastolic BP (DBP) and HR daytime (07:00–23:00 h) and night‐time (23:00–07:00 h) mean values, and their nocturnal decline (the difference between night‐time and daytime values: ∆SBP, ∆DBP, ∆HR). Mean BP values at night‐time are expected to fall by at least ≥10% compared to mean daytime values (dipping pattern). Loss of this nocturnal fall, or an increase in the mean night‐time values, are defined as non‐dipping and reverse‐dipping. ²² HR decreases during NREM sleep by 5–10%, ²³ , ²⁴ with an average difference between mean wake and sleep values of 6 bpm. ²⁵

Statistical Analysis

Statistical analyses were performed with SPSS Statistics (version 25). Normality was checked with the Shapiro–Wilk test. Comparisons were performed with Student t‐test for normally distributed data and Mann–Whitney test for not normally distributed data. Categorical variables were compared with the chi‐square test. Statistical significance was set at p ≤ 0.05.

Watson's U ² test was applied to verify whether the distribution of circular variables follows a von Mises distribution. Depending on the distribution, when appropriate, circular data analysis was performed using either the Watson‐Williams test or the Mardia‐Watson‐Wheeler test. This analysis was performed using STATA vers 17.

Results

Patients’ Characteristics and Sleep Questionnaires/Scales

One patient intentionally removed most of the recording, therefore 14 patients were included for analysis (eight females, six males, mean age 76 ± 3 years, disease duration 5.9 ± 2.6 years). PSPRS mean score was 51.3 ± 10.7 and MMSE mean score was 25.0 ± 2.7. H&Y stage expressed as median and interquartile range was 4 ± 0.75. Nine patients were treated with dopaminergic medications (six with levodopa, three with levodopa + dopamine agonist). The mean levodopa equivalent daily dose (LEDD) was 325 ± 229 mg. Two patients were treated with sedatives and six with antidepressants that for ethical reasons could not be discontinued. Eleven patients were treated with anti‐hypertensive medications that for safety reasons could only be discontinued in five patients with constant monitoring.

Mean PSQI was 10.4 ± 6.3. Nine patients had an overall score >5 (poor sleepers). Mean ESS score was 6.2 ± 3.6. Only two patients reported excessive daytime sleepiness. Six patients satisfied all the five criteria for RLS with a mean gravity score of 12 ± 2. No patient had history of RBD.

Controls were eight males and four females, significantly younger than PSP (43 ± 12 years) while BMI was comparable (26.69 ± 4.5 vs 28.5 ± 4.0 Kg/m ² ).

Sleep–Wake Cycle

Patients with PSP slept less than 5 h/night (TST 287 ± 118 min, sleep efficiency 53% ± 20) (Table 1). Night‐time sleep was interrupted by frequent awakenings (n = 37 ± 23, WASO 232 ± 137 min, arousal index 21.4 ± 7.8). For six patients, mean sleep latency of falling asleep after the light‐off was 66 ± 77 min. All sleep stages were represented in all patients but one did not reach REM stage. The mean percentage of N1 was 24.2 ± 13.7%, N2 45.1 ± 9.1%, N3 24.2 ± 14.7% and REM sleep was 9.9 ± 8.7% (normative values in the literature for this age range are N1 9.3%, N2 53.3%, N3 19.9% and REM 17.7% ¹⁷ ). There was poor representation of sleep spindles and K‐complexes in N2.

TABLE 1.

Night‐time sleep parameters of PSP patients

	PSP	Controls	Literature a
TST (min)	287 ± 118	393 ± 74* (p 0.014)	346.0 (326.7–365.4)
Sleep efficiency (%)	53 ± 20	80 ± 13* (p < 0.01)	77.5% (73.0–81.9)
Sleep latency (min)	66 ± 77	28 ± 34 (p 0.268)	N/A
REM latency (min)	151 ± 122	85 ± 42 (p 0.099)	N/A
WASO (min)	232 ± 137	88 ± 70 (p < 0.01)	77.1 (57.3–96.9)
N1 (%)	24.2 ± 13.7	7.6 ± 6.7* (p < 0.001)	9.3% (7.0–11.6)
N2 (%)	41.6 ± 15.3	43.8 ± 8.9 (p 0.730)	53.3% (50.0–56.7)
N3 (%)	24.2 ± 14.7	23.8 ± 6.1 (p 0.932)	19.9% (17.8–22.1)
REM (%)	9.9 ± 8.7	24.8 ± 4.9* (p < 0.001)	17.7% (16.9–18.5)

Abbreviations: N1, NREM sleep stage 1; N2, NREM sleep stage 2; N3, NREM sleep Stage 3; RWA, REM sleep without atonia; TST, total sleep time; WASO, wake after sleep onset.

^a Expressed as mean with 95% CIs.

^* p ≤ 0.05 (PSP vs controls).

Compared to healthy but younger controls from our laboratory, patients with PSP had significantly lower TST (p 0.014) and REM% (p < 0.001) and significantly higher N1% (p < 0.001).

REM onset latency was prolonged (>90 min) in nine patients. RBD (REM sleep without atonia, RWA, associated with motor activity and vocalizations) was recorded in two patients. A further nine patients presented with increased phasic activity during night REM (mean % of phasic RWA 33.7 ± 39.0). There were no significant differences in REM sleep %, duration and % of RWA between patients with and without antidepressant treatment at the time of the recording (p > 0.05 in all comparisons).

Eight patients had PLMSI >15 (mean PLMSI 56.0 ± 44.2). There were no significant differences in TST, SE, WASO, sleep stage duration, arousal index and number of awakenings according to PLMSI (p > 0.05 in all comparisons). Similarly, there were no significant differences in these parameters and PLMSI between patients with and without RLS (p > 0.05 in all comparisons).

AHI was available for eight patients and was below 5 in four patients and within 5–15 for the remaining four patients (mean AHI 8.1 ± 5.3).

During the light period, all patients had one or more naps. Four patients slept more than 60 min, the others had naps with a mean duration of 32 ± 29 min. There were no significant differences in nighttime parameters of TST, sleep stage duration, arousal index and number of awakenings between patients that slept more or less than 60 min during daytime (p > 0.05 in all comparisons). Daytime sleep was mainly represented by light sleep (N1% 54 ± 25, N2% 33.6 ± 20.9, N3% 11.9 ± 16.2, REM 0.3 ± 1.1). Nine controls had daytime naps with mean duration 55 ± 36 min.

TST for PSP patients over the 24 h (dark and light periods) was 338 ± 132 min.

Body Core Temperature

Ten patients completed the study protocol for BcT (two removed most of the recording set, two refused BcT recording).

BcT parameters are reported in Table 2 and Figure 1. All PSP patients presented the expected nocturnal decline in BcT. Patients presented significantly higher BcT values during day and night compared to controls, however BcT Δ was comparable.

TABLE 2.

Body core temperature parameters in PSP patients and controls

	PSP	Controls	p
BcT day (°C)	37.34 ± 0.20*	36.98 ± 0.16	<0.001
BcT night (°C)	37.06 ± 0.33*	36.61 ± 0.20	0.003
BcT Δ (°C)	−0.28 ± 0.26	−0.38 ± 0.20	0.497
Mesor (°C)	37.22 ± 0.21*	36.85 ± 0.10	<0.001
Amplitude (°C)	0.28 ± 0.16	0.39 ± 0.14	0.089
Acrophase (°) (hh:mm)	269.8 ± 47.2 17:59 ± 3:09	249.61 ± 32.31 16:02 ± 2:00	0.140
Wake_BcT (°C)	37.30 ± 0.19*	37.00 ± 0.13	<0.001
N1_BcT (°C)	37.03 ± 0.31*	36.56 ± 0.19	0.003
N2_BcT (°C)	37.05 ± 0.33*	36.50 ± 0.21	0.002
N3_BcT (°C)	37.01 ± 0.32*	36.57 ± 0.19	0.007
R_BcT (°C)	37.00 ± 0.37*	36.45 ± 0.22	0.001
N1_Δ (°C)	−0.27 ± 0.23	−0.44 ± 0.22	0.107
N2_Δ (°C)	−0.25 ± 0.27*	−0.50 ± 0.23	0.028
N3_Δ (°C)	−0.31 ± 0.21	−0.43 ± 0.20	0.190
R_Δ (°C)	−0.30 ± 0.25*	−0.55 ± 0.26	0.043

Note: Data are expressed as mean ± standard deviation.

Abbreviations: BcT, body core temperature; BcT Δ, BcT night − BcT day; N1, NREM sleep stage 1; N2, NREM sleep stage 2; N3, NREM sleep Stage 3; N1_Δ, BcT N1 – BcT wake; N2_Δ, BcT N2 – BcT wake; N3_Δ, BcT N3 – BcT wake; R_Δ, BcT R – BcT wake.

^* p ≤ 0.05.

Figure 1.

Day and night‐time BcT in PSP and controls. Legend: BcT = body core temperature (°C).

A circadian rhythmicity of BcT was detected in all PSP patients. Mesor was significantly higher in PSP than controls. Mean acrophase for PSP patients appeared to be postponed by 2 h compared to controls but this was not statistically significant. Amplitude was similar in the two groups.

The mean values of BcT during wake, N1, N2, N3 and REM were significantly higher in PSP than controls. There were no significant differences in BcT Δ in N1 and N3.

Blood Pressure and Heart Rate

A summary of mean BP and HR values during day‐ and night‐time is reported in Table 3. The mean daytime BP met the criteria for hypertension in only two patients, and was only systolic in both cases. Two patients presented a dipping BP pattern, seven patients a reduced dipping or non‐dipping pattern and the remaining five a reverse dipping pattern. HR decreased during night‐time by more than 6 bpm (or 10%) in six patients. Therefore the expected “dipping pattern” for HR was maintained in more patients compared to BP.

TABLE 3.

Day‐ and night‐time profile of blood pressure and heart rate in PSP (standard arm‐cuff)

	PSP (n = 14)
SBP day (mmHg)	126.5 ± 14.6
SBP night (mmHg)	123.2 ± 14.8
SBP Δ (mmHg)	−3.3 ± 10.1
DBP day (mmHg)	73.1 ± 6.8
DBP night (mmHg)	70.9 ± 6.9
DBP Δ (mmHg)	−2.3 ± 5.4
HR day (bpm)	74.9 ± 5.9
HR night (bpm)	69.4 ± 7.0
HR Δ (bpm)	−5.4 ± 4.7
BP pattern at night‐time
Dipping (n, %)	2 (14%)
Reduced/non‐dipping (n, %)	7 (50%)
Reverse dipping (n, %)	5 (36%)

Note: data are expressed as mean ± standard deviation unless specified.

Abbreviations: DBP, diastolic blood pressure; HR, heart rate; SBP, systolic blood pressure; Δ, night – day mean values.

Discussion

In the present study we aimed at describing sleep–wake cycle over 24 h under controlled environmental conditions in PSP patients and investigating circadian rhythmicity and state dependent changes of cardiovascular parameters and BcT.

We observed that PSP showed a significant reduction of sleep during night‐time which was not associated with a compensatory increase of daytime sleep. This led to a profound sleep deprivation across the 24 h.

Even though patients were resting all day in a room without being engaged in any activity, daytime naps were short and mainly represented by light sleep. Even when considering the total amount of sleep/24 h, median TST was still lower than the recommended amount of sleep for the elderly. ²⁶ While a reduction in nighttime sleep duration could be expected with aging, the same is not true for daytime naps that should become longer as a compensating mechanism for sleep loss overnight, to maintain a stable total sleep requirement across 24 h. ²⁷

Interestingly, the majority of our patients did not complain of excessive daytime sleepiness.

Previous studies assessing subjective excessive daytime sleepiness found similar scores ²⁸ , ²⁹ , ³⁰ with the exception of Gama et al who reported excessive daytime sleepiness in a higher percentage of cases (43%). ³¹

Concerning the possible contribution of sleep disorders, we did not observe a significant impact of either RLS and PLMSI nor OSA on sleep parameters.

Overall our data suggest a central origin or a homeostatic impairment of the sleep–wake cycle disruption in PSP, independent from total night sleep duration and other sleep disorders.

Previous studies evaluated sleep with Video‐EEG Polysomnography (VPSG) and daytime propensity to sleep with MSLT (which however is not validated for this purpose in PSP). These also supported a central origin underlying sleep–wake cycle disruption in PSP. Walsh and colleagues demonstrated a longer latency to fall asleep during daytime despite poor night‐time sleep in PSP, suggesting an impaired sensitivity to the homeostatic sleep drive with an increased arousal across the 24 h. ³ On the contrary, four out of 15 PSP patients in the study by Arnulf et al had shorter sleep latencies at MSLT and the same patients tended to sleep longer during night‐time which suggested that sleepiness in PSP was caused by primary central hypersomnia. ²⁸ Finally, Montplasir et al studied three patients with MSLT and reported high intersubject variability (mean sleep latencies were 1.89, 9.53 and 20 min). However, they selected only patients who reported daytime sleepiness. The authors suggested that daytime somnolence is not a characteristic feature of PSP.

At VPSG, nocturnal SE was markedly lower than normative values from our sleep laboratory, TST significantly reduced and WASO significantly increased with frequent awakenings, which reflected the main patients’ complaint of poor sleep. Previous studies addressing sleep disturbances by means of questionnaires and interviews found that approximately 77–100% of patients reported poor sleep. ³² , ³³ , ³⁴ , ³⁵ , ³⁶ Compared to studies using the PSQI, ³⁰ we found a higher mean score and higher percentage of poor sleepers. This difference could be attributed to the younger population of patients in previous studies and the use of a different cut off to define a poor sleep score.

Analysis of sleep structure showed that PSP patients, compared to age‐matched reference values, presented higher proportion of N1 and lower N2, with poor representation of sleep spindles and K‐complexes. N3 was increased in PSP, which is likely to be related to increased EEG slowing and lack of sleep spindles and K‐complexes. Indeed patients with neurodegenerative diseases can present an abnormal slowing of EEG and NREM characterized by continuous and diffuse slow‐wave EEG activity without K‐complexes and sleep spindles. Undifferentiated NREM sleep has been proposed as a better term to classify certain NREM periods in patients with neurodegenerative conditions and it has been proposed that sleep scoring in neurodegenerative diseases should be tailored to address these different features. ³⁷

These findings are in line with the few previous studies that assessed night‐time sleep by means of PSG ³ , ²⁸ , ³⁸ , ³⁹ , ⁴⁰ ; however N3 was found reduced in one study. ³

RBD was VPSG recorded in two patients. In previous studies RBD prevalence ranged from 13% to 35%, ²⁸ , ⁴¹ , ⁴² therefore even though RBD is considered a typical marker of synucleinopathies (nearly 30–50% of Parkinson's disease (PD) and > 70% of Dementia with Lewy bodies or Multiple system atrophy patients have RBD), it can be found even in tauopathies but at a lower frequency. ⁴³ , ⁴⁴ A further nine patients presented increased phasic activity during REM stage. The clinical significance of this in the absence of a history suggestive of RBD remains uncertain.

To evaluate the hypothesis that patients with PSP have a hyperarousal state, we also investigated circadian rhythms and the state dependent modulation of thermoregulatory and cardiovascular autonomic function.

PSP maintained BcT circadian rhythmicity (acrophase), whereas the higher mesor suggests an impairment of the homeostatic control of thermoregulation. We observed higher mean values of BcT during wakefulness, N1, N2, N3 and REM in PSP vs controls, with preserved physiological state‐dependent modulation.

Only one previous study explored BcT circadian rhythm in PSP. The authors found normal circadian function in all patients but a lower amplitude compared to PD. ⁴⁵ Interestingly PSP patients in our study showed a similar BcT profile to multiple system atrophy patients, ⁴⁶ who had significantly higher mesor and BcT during N1‐2, N3‐4 and REM compared to PD and controls.

Our controls were significantly younger than patients, however evidence suggest that BcT may decline slightly with age, therefore this age difference is unlikely to explain our results. ⁴⁷

The primary area for thermoregulation is the preoptic/anterior hypothalamic area of the hypothalamus. This contains thermosensitive neurons that trigger heat‐loss or heat‐gain responses through sympathetically mediated responses including skin vasoconstriction and vasodilation, sweating and shivering. ⁴⁸ Thermoregulation is activated in response to internal and external stimuli but also following a circadian rhythm from the suprachiasmatic nucleus responsible for circadian variations of BcT. ⁴⁹ Heat loss and BcT reduction require an inhibition of the sympathetic tone, which therefore leads to vasodilatation and increased blood flow to the periphery and lower metabolic rate. The higher mean BcT values observed in our patients may reflect an impairment of the central control leading to reduced inhibition of the sympathetic drive.

BcT reduction at nighttime is essential for sleep initiation and maintenance. The higher mean BcT values observed in patients at night may therefore impair sleep structure as a consequence.

Alternatively, the BcT abnormalities observed in PSP may be secondary to the sleep/wake cycle disturbances described earlier. Total sleep time reduction and frequent awakenings may prevent the required sympathetic inhibition to produce peripheral vasodilatation necessary for heat loss and BcT fall, thus maintaining mean BcT values above the expected range. There is a bidirectional relationship between sleep and thermoregulation ⁵⁰ and further studies are needed to elucidate the pathophysiology behind our results.

Regarding circadian rhythms of cardiovascular autonomic function, 24 h BP profile in PSP was abnormal in 86% of patients as only two presented the expected dipping pattern, whereas the rest had either a reduced/non‐dipping or a reverse dipping pattern.

Central commands are primarily responsible for the changes in BP and HR that occur during transition from wakefulness to sleep. During NREM sleep there is usually a progressive reduction of BP and HR and increased baroreflex sensitivity. This may in part reflect reduced input from sympathoexcitatory hypothalamic neurons ⁵¹ . A non‐dipping profile of BP and HR similarly could be secondary to reduced sympathetic inhibition.

On the other hand, non‐dipping profile may be due to essential hypertension as the majority of patients included in this study were diagnosed with this condition. However, we discontinued antihypertensive medications only in half of the patients, and an abnormal pattern was observed also in those still receiving treatment.

A recent multi‐center study found that patients with PSP have a higher prevalence of cardiovascular disease and polypharmacotherapy compared to the age‐matched general population. Whether there is a causal effect or a common predisposition still has to be elucidated. Nevertheless this study highlights the complexity of PSP patients and the challenges in studying them given the high degree of comorbidities and concomitant medications. ⁵²

Absence of the physiological BP fall may be also related to cardiovascular autonomic failure which presents with neurogenic orthostatic hypotension and supine hypertension. ²² However all of our patients had cardiovascular autonomic testing that excluded neurogenic orthostatic hypotension.

Sleep disorders could also contribute to abnormal nocturnal BP profile. ⁵³ In particular OSAS leads to sympathetic overactivity and has been associated with nocturnal BP increase. ⁵³ , ⁵⁴ Indeed six out of 12 of our patients with an abnormal pattern had OSAS, but the degree of OSAS was moderate–severe only in two. A non‐dipping pattern is related to apnoea severity and significantly correlates with AHI. ⁵⁵ Therefore, it is unlikely that the mild OSA observed in our patients explains the abnormal BP nocturnal profile.

RLS and PLMS have been associated with hypertension due to increased sympathetic activity. ⁵⁶ , ⁵⁷ RLS patients have increased frequency of a non‐dipping BP profile, however PLMSI (cut off at 15) did not influence non‐dipping profile frequency. ⁵⁸ This may suggest that PLMSI alone may not determine a non dipping BP profile. Studying the specific contribution of each disorder is difficult as in clinical practice RLS, PLMS and OSA frequently overlap.

A nocturnal non‐dipping profile of HR was observed in 60%, and a sympathetic predominance may underlie such results similarly to BP. Interestingly, abnormalities in circadian rhythm of HR were less frequent than those observed for BP, suggesting that different mechanisms may underlie its control.

Very few neuropathological studies investigated autonomic structures in PSP. One study demonstrated abnormalities in several autonomic nuclei located in the brainstem which are mainly related to cardiovascular function. ⁵⁹ The intermediolateral column in the spinal cord was found to be spared ⁶⁰ or moderately affected (39% nerve cell loss) by the neurodegenerative process in PSP. ⁶¹

Tau pathology was described in the SCN of PSP patients. ⁶² However in our study circadian rhythmicity was maintained, therefore the SCN does not appear to be affected. Ideally, having clinical and neuropathological data from the same individuals would provide stronger evidence in this regard. The role of melatonin should also be further investigated. The same neuropathological study found preserved pineal gland tissue however no study so far investigated melatonin profile and its association to sleep and BcT parameters in PSP patients. This may be useful from a therapeutic point of view.

Limitations of this study include the small number of patients and reduced cooperation in a few cases given the complexity of the tests. We did not have neuropathological confirmation. However the specificity for PSP pathology of new diagnostic criteria is 91.4% ⁶³ and in addition, none had red flags and all patients had brain MRI with spectroscopy consistent with PSP. The control group had different demographic characteristics from the patients. Seven patients were taking anti‐hypertensive medications (beta‐blocker, ACE‐inhibitor, angiotensin II receptor antagonist, alpha‐adrenoreceptor blocker, calcium channel blocker, diuretic) that were not discontinued for safety reasons. Nevertheless we found our results of interest considering that the majority of patients presented an abnormal BP profile at night despite being on anti‐hypertensive medications.

In summary, our study is the first that assessed circadian rhythms of sleep–wake cycle, BcT and cardiovascular parameters under controlled environmental conditions for 48 h in PSP. We demonstrated that PSP patients present with severe sleep deprivation across the 24 h. An imbalance between sympathetic and parasympathetic function with sympathetic predominance may be associated with this alteration. Further studies would be needed to understand whether autonomic dysfunction is the primary drive for this sleep–wake cycle disruption or the consequence of sleep loss. Understanding the underlying pathophysiology provides grounds for potential treatment options. For example, it would be possible to explore the effect of agents that modulate sympathetic function (eg, clonidine) on the sleep–wake cycle of patients with PSP. Given the role of circadian rhythm disruption in promoting the progression of neurodegenerative processes, further studies may lead to novel therapies for PSP and other neurodegenerative conditions.

Author roles

(1) Research project: A. Conception, B. Organization, C. Execution;

(2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique;

(3) Manuscript Preparation: A. Writing of the first draft, B. Review and Critique;

F.B.: 1B, 1C, 2A, 2B, 2C, 3A.

L.S.: 1B, 1C, 3B.

P.A.: 1C, 2B, 3B.

A.C.: 1C.

P.G.: 3B.

P.C.: 1A, 3B.

F.P.: 3B.

G.C.B.: 1A, 1B, 3B.

Disclosures

Ethical Compliance Statement: The study was approved by the Ethics Committee of Area Vasta Emilia Centro (reference number 18141). Written informed consent was obtained and documented for each patient enrolled in the study. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Sources and Conflict of Interest: No specific funding was received for this work. The authors declare that there are no conflicts of interest relevant to this work.

Financial Disclosures for the previous 12 months: The authors declare that there are no additional disclosures to report.

Supporting information

Data S1. Circadian rhythm recording standardized procedures. Cardiovascular reflex test results of PSP patients. Clinical data of PSP patients.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.