Longitudinal evolution of sleep disturbances in early multiple system atrophy: a 2‐year prospective cohort study

We recruited patients with MSA from January 2018 to May 2022, referred from the Department of Neurology, West China Hospital, Sichuan University in the prospective cohort study. The disease duration of patients with MSA enrolled in the present study should be less than 3 years at baseline. Patients were excluded if the spinal cerebellar ataxia (SCA) genetic tests (SCA1, 2, 3, 6, 7), brain magnetic resonance imaging scan, or blood tests indicated other neurological disorders.

Patients underwent regular face-to-face interviews by neurologists at baseline, 1-year, and 2-year follow-ups until May 2023. A total of 231 patients were enrolled at baseline, seven patients voluntarily withdrew from the study, and four patients diagnosed with PD during the 1-year follow-up were excluded from the analysis. Finally, 220 patients completed the 1-year follow-up evaluation. Seventy-one patients could not come to the hospital due to death (n = 12), COVID-19 infection (n = 31), or wheelchair confinement (n = 28); 8 patients lost contact; 10 patients voluntarily withdrew from the study; 41 patients have not reached the time to completed the 2-year follow-up visit. Eventually, ninety patients completed the 2-year follow-up visit. All patients met the 2008 criteria for a probable diagnosis of MSA [16].

According to the predominant motor symptoms, patients were divided into two subtypes: MSA with predominant parkinsonism (MSA-P) or MSA with predominant cerebellar ataxia (MSA-C). Demographic and clinical data, including age, sex, age of onset, disease duration, and treatment, were recorded. The total daily Levodopa equivalent daily dose (LEDD) was calculated using the established methods [17]. Disease onset was defined as the initial presentation of any motor symptoms (Parkinsonism or cerebellar dysfunction) or autonomic features, except erectile dysfunction [16]. Disease duration was defined as the time between disease onset and evaluation.

The sleep disturbances evaluation included RBD, EDS, and PD-SP. The RBD-SQ, comprising a 10-item patient self-rating questionnaire covering the clinical features of RBD, is widely used as a screening tool for RBD [18]. RBD-SQ score ≥ 5 was defined as probable RBD. ESS is a simple, self-administered questionnaire for measuring daytime sleepiness [19], with a score of ≥ 10 indicating EDS. PD-SP was assessed using the PDSS-2 [20], which contains 15 individual items and was divided into three domains including disturbed sleep, motor symptoms at night, and PD symptoms at night. PD-SP was defined as a PDSS-2 score of ≥ 18. Disease severity was evaluated using the Unified Multiple System Atrophy Rating Scale (UMSARS) with part I (activities of daily living), part II (motor examination), part III (autonomic examination), and part IV (global disability) [21], and the total UMSARS score was calculated as the sum of parts I and II. The blood pressure measurements were taken at 1-, 3-, 5-, and 10-min intervals in the upright position, and were compared with the measurements in the supine position [22]. Orthostatic hypotension (OH) was defined as a reduction in the systolic blood pressure by at least 20 mmHg and/or diastolic blood pressure by at least 10 mmHg. Cognitive function was evaluated using a comprehensive and standardized cognitive screening tool (Montreal Cognitive Assessment (MoCA)) [23].

All continuous data are presented as mean and standard deviation, and categorical variables as counts (percentages). Mann–Whitney U test was used to compare the continuous variables between different groups, since most of the data were not normally distributed. Chi-squared test was used to compare the categorical variables between different groups. When comparing the severity or frequency of the three types of sleep disturbances between patients with MSA-P and MSA-C, the logistic regression analysis was used to correct for confounding factors (age and LEDD) [24]. The 2-year change in the PDSS-2 score \(=\frac{(\mathrm{PDSS}-2\mathrm{\ score\ at }2-\mathrm{year\ follow\ up }-\mathrm{ baseline\ PDSS}-2\mathrm{\ score})}{\left(\mathrm{date\ follow}-\mathrm{up\ in\ months}-\mathrm{date\ baseline\ in\ months}\right)/12}\), represents the progression of PD-SP. The 2-year change in the ESS score \(=\frac{(\mathrm{ESS\ score\ at\ }2-\mathrm{year\ follow\ up }-\mathrm{\ baseline\ ESS\ score})}{\left(\mathrm{date\ follow}-\mathrm{up\ in\ months}-\mathrm{date\ baseline\ in\ months}\right)/12}\), represents the progression of EDS. The 2-year change in the RBDSQ score \(=\frac{(\mathrm{RBDSQ\ score\ at }2-\mathrm{year\ follow\ up }-\mathrm{ baseline\ RBDSQ\ score})}{\left(\mathrm{date\ follow}-\mathrm{up\ in\ months}-\mathrm{date\ baseline\ in\ months}\right)/12}\). A generalized estimating equation (GEE) model with an exchangeable working correlation structure was used to investigate significant longitudinal sleep disturbance changes (score and frequency) in MSA after adjusting for age and LEDD, in this longitudinal observational study [4]. The Bonferroni correction was used for multiple comparisons. Additionally, we used GEE with multiple linear regression analysis to explore the factors associated with the score of PDSS-2, ESS, and RBDSQ. The models included all patients in the cohort and allowed for an association between repeated measurements acquired in the same patients. The independent variables included the following repeated measures: age, sex (male = 1, female = 0), diagnosis subtype (MSA-P = 1, MSA-C = 0), UMSARS total score, OH, and MoCA score. All data analyses were performed using the IBM SPSS Statistics software (version 26.0). P value < 0.05 was considered statistically significant.

In the current study, we recruited 231 patients with probable MSA at baseline and 220 patients (51.8% male, 48.6% MSA-P) were enrolled in analysis. Among the total of 220 patients with MSA, all completed 1-year follow-up assessment, and 90 patients completed the 2-year follow-up visit. There were no significant differences in the baseline clinical and demographic features between patients with MSA who were lost to follow-up and those who were not except for disease duration (Additional file 1: Table S1). The mean age of patients with MSA was 59.32 ± 8.09 years, and the mean disease duration was 1.68 ± 0.75 years at baseline. The score of PDSS-2 (11.13 ± 6.52 vs. 8.14 ± 5.27, p < 0.001) and ESS (6.13 ± 5.06 vs. 3.57 ± 3.87, p < 0.001) was higher in patients with MSA-P than in those with MSA-C after adjusting for age and LEDD. The frequency of PD-SP (20.6% vs. 8.8%, p = 0.016) and EDS (25.2% vs. 10.6%, p = 0.006) was also higher in patients with MSA-P than in those with MSA-C after adjusting for age and LEDD. Additionally, the score of RBDSQ and the frequency of RBD were not significantly different between patients with MSA-P and MSA-C. The score and frequency of sleep disturbances were not significantly different between male and female patients with MSA. The 2-year change in the PDSS-2, ESS, or RBDSQ score was not significantly different between MSA-P and MSA-C (all p > 0.05). The 2-year change in the ESS score was higher in male MSA than in female (p = 0.025), while the 2-year change in the PDSS-2 or RBDSQ score was not significantly different between male and female patients (all p > 0.05) (Table 1).

In the total MSA group, the score of PDSS-2, ESS, and RBDSQ was significantly increased over the 2-year follow-up (all p < 0.05), even after adjusting for age and LEDD. In the MSA-P group, the score of PDSS-2, ESS, and RBDSQ was significantly increased over the 2-year follow-up (all p < 0.05). After adjusting for age and LEDD, the score of ESS and RBDSQ remained significantly increased (all p < 0.05), while the score of PDSS-2 did not (p = 0.234). In the MSA-C group, the score of PDSS-2 and ESS was significantly increased over the 2-year follow-up (all p < 0.05), even after adjusting for age and LEDD. The score of RBDSQ was not significantly increased over the 2-year follow-up (all p > 0.05). In the male patients in the MSA group, the score of ESS was significantly increased over the 2-year follow-up (all p < 0.001), while the score of PDSS-2 and RBDSQ was not (all p > 0.05). In the female patients with MSA, the score of PDSS-2 and ESS was significantly increased over the 2-year follow-up (all p < 0.001), while the score of RBDSQ was not (p = 0.144) (Table 2 and Fig. 1).

The comparison of the score of the PDSS-2 domain at baseline and 1- and 2-year follow-ups was shown in Additional file 1: Table S2. We found that the score of disturbed sleep was significantly increased over the 2-year follow-up (all p < 0.05) in the total MSA, MSA-C, and male patients with MSA groups, even after adjusting for age and LEDD. However, after adjusting for age and LEDD, the score of motor symptoms at night and PD symptoms at night was not significantly increased over the 2-year follow-up in all groups (all p > 0.05).

In the total MSA group, the most common sleep disturbance was RBD, followed by EDS and PD-SP over the 2-year follow-up. The frequency of EDS was significantly increased from 17.7 to 37.8% over the 2-year follow-up (p < 0.001), even after adjusting for age and LEDD. The frequency of PD-SP was significantly increased from 14.5 to 26.7% over the 2-year follow-up (p = 0.001), while the finding did not remain after adjusting for age and LEDD (p = 0.108). In the MSA-P group, the frequency of EDS was significantly increased from 25.2 to 47.8% over the 2-year follow-up (p < 0.001), even after adjusting for age and LEDD. The frequency of PD-SP was significantly increased from 20.6 to 37.0% over the 2-year follow-up (p = 0.026), while the finding did not remain after adjusting for age and LEDD (p = 0.232). In the MSA-C group, the frequency of EDS significantly increased from 10.6 to 27.3% over the 2-year follow-up (all p < 0.05), even after adjusting for age and LEDD. The frequency of PD-SP was significantly increased from 8.8 to 15.9% over the 2-year follow-up (p = 0.041), while the finding did not remain after adjusting for age and LEDD (p = 0.160). In the male patients in the MSA group, the frequency of EDS was significantly increased from 21.1 to 50.0% over the 2-year follow-up (p < 0.001), even after adjusting for age and LEDD. The frequency of PD-SP was not significantly increased over the 2-year follow-up (all p > 0.05). In the female patients in the MSA group, the frequency of PD-SP (from 17.9 to 34.8%) and EDS (from 14.2 to 26.1%) was significantly increased over the 2-year follow-up (all p < 0.05), even after adjusting for age and LEDD. However, the frequency of RBD was not significantly increased over the 2-year follow-up in any group (all p > 0.05) (Table 2 and Fig. 1).

In the total MSA group, the frequency of patients with MSA did not report any sleep disturbances at baseline, 1-year, and 2-year follow-up was 36.8%, 26.8%, and 22.2%, respectively. The frequency of overlap in two types of sleep disturbances at baseline, 1-year, and 2-year follow-up was 17.3%, 21.4%, and 25.5%, respectively. The frequency of coexistence of all three sleep disturbances at baseline, 1-year, and 2-year follow-up was 1.8%, 6.8%, and 13.3%, respectively (Fig. 2).

In the MSA-P group, the frequency of overlap in two types of sleep disturbances at baseline, 1-year, and 2-year follow-up was 22.4%, 28.0%, and 34.7%, respectively. The frequency of coexistence of all three sleep disturbances at baseline, 1-year, and 2-year follow-up was 3.7%, 10.3%, and 19.6%, respectively. In the MSA-C group, the frequency of overlap in two types of sleep disturbances at baseline, 1-year, and 2-year follow-up was 12.4%, 15.1%, and 15.9%, respectively. The frequency of coexistence of all three sleep disturbances at baseline, 1-year, and 2-year follow-up was 0.0%, 3.5%, and 6.8%, respectively (Fig. 3).

In the male patients with MSA, the frequency of overlap in two types of sleep disturbances increased from 17.6 to 25.0% over the 2-year follow-up. The frequency of coexistence of all three sleep disturbances increased from 1.8 to 15.9% over the 2-year follow-up. In the female patients with MSA, the frequency of overlap in two types of sleep disturbances increased from 17.0 to 26.0% over the 2-year follow-up. The frequency of coexistence of all three sleep disturbances increased from 1.9 to 10.9% over the 2-year follow-up (Fig. 4).

The GEE model showed that the sex (B =  − 2.063, p = 0.006), diagnosis subtype (B = 3.146, p < 0.001), and UMSARS total score (B = 0.080, p < 0.001) were associated with the score of PDSS-2; that sex (B = 1.537, p = 0.012), diagnosis subtype (B = 3.066, p < 0.001), UMSARS total score (B = 0.085, p < 0.001), and MoCA score (B =  − 0.282, p < 0.001) were associated with the score of ESS; and that the UMSARS total score (B = 0.027, p = 0.014) were associated with the score of RBDSQ (Additional file 1: Table S3).