Circadian rest-activity rhythm pattern in the elderly with cerebral small vessel disease: Using multiple estimated methods

Abstract

Background: Disruption of circadian rest-activity rhythm (RAR) has been found in many neurological disorders. Objective: In this study, actigraphic data were collected and analyzed to identify the RAR pattern in the elderly with cerebral small vessel disease. Methods: 115 cerebral small vessel disease (CSVD) cases were recruited. The presence of lacune infarct, white matter hyperintensities, and cerebral microbleeds in magnetic resonance imaging (MRI) images were rated independently, as well as using a simple MRI score of 0–3 points. Each subject wore an Actigraph device in their nondominant hand for 4–7 days to collect raw data. RAR parameters were generated using both extended cosinor model (RAR α, RAR β, amplitude, acrophase, up-mesor, down-mesor, and pseudo-F statistic) and non-parametric methods (interdaily stability, intradaily variability, and relative amplitude). Results: Elder patients with a simple MRI score of 2–3 points showed a statistically lower amplitude compared with individuals with a simple MRI score of 0 points in the extended cosinor model. For the non-parametric method, elderly people with a simple MRI score of 1–3 points exhibited higher intradaily variability relative to those participants with a simple MRI score of 0 points. However, no differences were found regarding sleep quality among individuals with different simple MRI scores. White matter hyperintensities, lacune infarct, and cerebral microbleeds were independently associated with RAR β, RAR α, and intradaily variability, respectively. Conclusions: The RAR pattern was disturbed in elderly adults with CSVD. Abnormal RAR parameters were independently associated with CSVD MRI markers.

Keywords

actigraphy aging Alzheimer's disease cerebral small vessel disease circadian rest-activity rhythm magnetic resonance imaging

Introduction

In humans, sleep-wake activity follows a diurnal pattern known as circadian rest-activity rhythm (RAR).¹ Proper RAR is fundamental to optimal health,² whereas RAR disruption is associated with an increased risk of chronic diseases, such as cancer and diabetes.^1,3 RAR characteristics, including regularity, fragmentation, amplitude, and fitting to 24-h models, can be objectively quantified by applying various methods to several days of continuous accelerometer recordings.⁴

As one of the most common causes of permanent functional disability and mortality in elderly people,^5,6 cerebral small vessel disease (CSVD) affects deep perforating vessels in the brain and, thus, can be seen on magnetic resonance imaging (MRI).⁷ White matter hyperintensity (WMH), lacunar infarct (LI), and cerebral microbleeds (CMBs) are the most acknowledged MRI features.^8,9 Furthermore, since these MRI markers do not always occur separately, several scoring systems were invented to represent the overall burden of CSVD in clinical practice.^10,11

Besides common features of CSVD, such as cognitive impairment, gait abnormalities, and emotional changes, recent research has elucidated a potential relationship between CSVD and RAR characteristics as well as sleep quality.¹² At present, the identified RAR feature discovered in CSVD patients is still far from being consistent. For example, fragmentation (reflected by intradaily variability [IV]) RAR might be related to MRI markers for CSVD, such as white matter lesions and cerebral microbleeds.¹³ In contrast, Oosterman et al. reported that stability (reflected by interdaily stability [IS]) and amplitude (reflected by relative amplitude [RA]) rather than fragmentation were related to deep WMH.¹⁴ In this study, we sought to identify the RAR pattern of elderly persons with CSVD by analyzing actigraphy data using the extended cosinor model and non-parametric methods and clarify the relationship between RAR and CSVD MRI markers.

Methods

Participants

This cross-sectional observational study was conducted between November 1, 2020, and June 20, 2023. It included 115 consecutive elder CSVD cases at the inpatient of Department of Neurology of Seventh Medical Center of PLA General Hospital (Beijing, China). The current research protocol was approved by the Academic Ethics Committee of the Biological Sciences Division of the PLA General Hospital in Beijing, China. Most of the participants showed complaints about slight dizziness, mild headache, mood disturbances, cognitive disorders, etc., with strong will of MRI scan. Patients with the following conditions were excluded: major tremor, cerebral bleeding episodes, or stroke; additional factors, inducing leukoencephalopathy (caused by immune or genetic factors or demyelination); multisystem disorders, such as polyarteritis nodosa, connective tissue diseases–related nervous system vasculitis, arthritis, infection-associated vasculitis; psychotropic medications; musculoskeletal diseases; contraindications for MRI together with additional neurodegenerative diseases such as Alzheimer's disease (clinical diagnostic criteria < NIA-AA>), Parkinson's disease, and frontotemporal dementia. Age, sex, educational level, body mass index (BMI), and a series of vascular risk factors, including hypertension, diabetes, coronary artery disease, hyperlipidemia, smoking, and alcohol use, were considered possible covariates. Hypertension was defined as systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg and/or use of antihypertensive medications.¹⁵ Diabetes was defined as ongoing or newly initiated therapy with antidiabetic drugs or hemoglobin A1c of ≥6.5%.¹⁶ Hyperlipidemia was defined as one of the following points in fasting venous plasma test: total cholesterol ≥6.2 mmol/L; low density lipoprotein cholesterol ≥4.1 mmol/L; triglycerides ≥2.3 mmol/L.¹⁷ All the vascular risk factors were in binomial form.

MRI measurements

The 3.0T MRI brain scan (Discovery MR750; GE Healthcare, USA) revealed white matter lesions that reflected the severity of CSVD. This study used brain MRI (inter-slice and slice thicknesses of 1.5 and 5 mm, separately) with T1 fluid-attenuated inversion recovery (TR, 1750 ms; TE, 23 ms; T1, 780 ms; FOV, 24 cm) and T2-weighted imaging (TR, 7498 ms; TE, 105 ms; FOV, 24 cm). The evaluators had no knowledge of imaging results.

Simple MRI score

The CSVD burden was calculated according to Olama et al.¹⁸ and the simple CSVD MRI score in the current study ranged from 0 to 3 points. In detail, 1 point was awarded for each of the following: WMH graded above 2 points according to Fazekas scale and the presence of LIs (≥3) and CMBs (present). Simple MRI score was related to cognitive deficits, mortality, and neurologic dysfunction in CSVD patients.^19–21

Wrist actigraphy

In accordance with our protocol, all cases were asked to wear the ActiGraph GT3X + device (ActiGraph, Pensacola, USA) on the nondominant wrist all day (except when swimming or bathing) for 4–7 days.

ActiLife software (ActiGraph, Pensacola, FL, USA) was used to download data after each wear period. Data with sufficient wear time were chosen and analyzed.

Actigraphic data capture

This work obtained actigraphic data at the epoch of 60 s using the corresponding software. Due to the triaxial accelerometer nature of the ActiGraph GT3X + device, this work sampled diurnal VM using the formula below:

VM = \sqrt{X^{2} + Y^{2} + Z^{2}},

where X, Y, and Z are vector magnitude counts in X-, Y-, and Z-axes, respectively.

RAR data analysis

First, we used the extended cosinor model, which is derived from R code implementing the sigmoidally transformed cosine (Figure 1). RAR features were exhibited in the following variables.

Figure 1.

Illustration of extended cosinor–modeled RAR parameters from a study participant. Time (x-axis) is set in days. RAR α indicates the relative width of the active period, with higher values corresponding to more narrow active periods. RAR β indicates the steepness of the modeled rhythm, with higher values corresponding to more steep curves. Amplitude and mesor are height parameters, and up-mesor, acrophase, and down-mesor are time parameters, as described in the text.

Figure 2.

Scatter graphs demonstrating the relationship between WMH and RAR β (A), LI and RAR α (B), as well as CMBs and IV (C).

RAR α: a parameter indicating the relative width of active to rest periods (higher alpha indicates narrower active relative to resting periods). RAR β: a parameter indicating the steepness of RARs (higher values indicate steeper or “square-like” RARs). Acrophase: a parameter indicating the time when activity peaks in the modeled rhythm. Amplitude: a parameter reflecting the difference in magnitude of activity between active and rest phases. Up-mesor: the time when activity passes up through mesor, approximating the time the participant “gets going” in the morning. Down-mesor: the time when activity passes down through the mesor, approximating the time of day the participant “settles down” for the night pseudo-F: a parameter reflecting how well the modeled rhythm fits the observed data (lower values indicate poorer model fit, which suggests that the rhythm might be erratic and/or variable).

Second, non-parametric properties were determined using ActiGraph enumeration data (vector magnitude data) as follows: (1) IS; (2) IV; (3) time occurrence with associated activity counts for a 10-h period with the highest activity (M10) and 5-h period with the lowest activity (L5); (4) RA.²²

IS provides data on RAR synchronization for environmental stimulation with claimed stability, which is calculated using formula 1 below based on an average 24-h profile:

IS = \frac{n \sum_{h = 1}^{p} {(x_{h} - x)}^{2}}{p \sum_{1}^{n} {(x_{i} - x)}^{2}} .

IV offers data regarding RAR fragmentation within a 24-h profile, and it can be determined based on hourly raw data by formula 2 below:

IV = \frac{n \sum_{i = 2}^{N} {(x_{i} - x_{i - 1})}^{2}}{(n - 1) \sum_{i = 1}^{n} {(x_{i} - x)}^{2}} .

In the IS and IV formulas, N represents the total number of data points, p represents the daily data entry count, x_h represents the mean every hour, `x represents mean data, and x_i represents raw data hourly. Furthermore, IS ranges from 0 to 1, with 0 representing Gaussian noise (including normal distribution noise) and 1 representing perfect synchronization. Additionally, IV ranges from 0 to 2, with a high value indicating a fragmented rhythm.

M10 is the maximum sum for a 10-h continuous activity log, and L5 represents the smallest sum for a 5-h continuous activity log. The ratio of (M10-L5)/(M10 + L5) can be used to calculate RA.²³

Sleep quality

Aside from RAR features, sleep variables such as sleep latency (SL), wake after sleep onset (WASO), sleep efficiency (SE), average duration of awakenings (ADA), and times of awakenings (TA) could be automatically analyzed and exported. Daytime sleepiness was evaluated by the Epworth sleepiness scale (ESS). The ESS consists of 8 self-rated items, each scoring from 0 to 3, which measure a subject's habitual “likelihood of dozing or falling asleep” in common situations of daily living.²⁴

Statistics

For analyzing the differences in demographic and clinical data between groups, Dunnett for one-way ANOVA and the Bonferroni method for the chi-square test were selected to compare the demographic data. Analysis of Covariance (ANCOVA), adjusting for ages was used to compare the RAR parameters and sleep quality. A linear regression analysis adjusted for age, gender, and educational level was used to find the association between RAR variables and CSVD severity. A p < 0.05 was used to indicate statistical significance. SPSS22.0 was used for statistical analysis.

Results

Patients’ demographic characteristics are shown in Table 1. Among groups, differences in demographic features were not significant except for age (67.47 ± 5.18 years versus 69.20 ± 7.03 years versus 72.93 ± 9.32 years versus 73.09 ± 9.70 years; p = 0.014) and hypertension frequency (15.79% versus 50.00% versus 60.60% versus 55.17%; p = 0.009). For the extended cosinor model, elderly patients with simple MRI score of 2–3 points showed statistically lower amplitude compared with individuals with simple MRI score of 0 points (110.57 ± 106.56 versus 68.20 ± 49.72 versus 52.72 ± 52.85, p = 0.044). For the non-parametric method, elderly people with a simple MRI score of 1–3 points exhibited higher fragmentation (higher IV) compared to those with a simple MRI score of 0 points (0.77 ± 0.37 versus 0.97 ± 0.33 versus 1.05 ± 0.32 versus 1.11 ± 0.33, p = 0.020). It seemed that elder patients with a simple MRI score of 2–3 points showed lower RA compared with individuals with a simple MRI score of 0 points (0.78 ± 0.13 versus 0.70 ± 0.12 versus 0.67 ± 0.13, p = 0.057), whereas the trends were only marginal. We did not find obvious differences regarding other RAR parameters and sleep quality. Further details are given in Table 2.

Table 1.

Clinical and demographic characteristics of participants.

	Simple MRI score
Item	0	1	2	3	p
Demographic characteristics	N = 19	N = 34	N = 33	N = 29
Men, %	8 (42.10)	15 (40.54)	21 (63.63)	15 (51.72)	0.553
Age, mean (SD), y	67.47 (5.18)	69.20 (7.03)	72.39 (9.32)	73.09 (9.70)	0.014*^#
Education, mean (SD), y	9.79 (3.82)	10.06 (3.12)	9.26 (3.89)	10.63 (3.41)	0.488
BMI, mean (SD), cm/kg²	28.50 (8.19)	27.05 (8.07)	30.36 (10.02)	30.19 (9.69)	0.402
Hypertension, %	3 (15.79)	17 (50.00)	20 (60.60)	16 (55.17)	0.009*^#&
Diabetes, %	8 (42.10)	9 (26.47)	11 (33.33)	12 (41.37)	0.741
CAD, %	7 (36.84)	4 (11.76)	4 (12.12)	5 (17.24)	0.174
Hyperlipidemia, %	13 (69.42)	25 (73.53)	23 (69.70)	23 (79.31)	0.490
Smoking, %	5 (26.32)	10 (29.41)	12 (36.36)	10 (34.48)	0.859
Alcohol user, %	4 (21.05)	14 (41.17)	8 (24.24)	9 (31.03)	0.537

SD: standard deviation; CAD: coronary artery disease; MRI: magnetic resonance imaging.

*p < 0.05 simple MRI score = 0 versus simple MRI score = 3.

p < 0.05 simple MRI score = 0 versus simple MRI score = 2.

p < 0.05 simple MRI score = 0 versus simple MRI score = 1.

Table 2.

RAR parameters and sleep quality of participants.

	Simple MRI score
Item	0	1	2	3	p
Extended cosinor model	N = 19	N = 34	N = 33	N = 29
RAR α	−0.45 (0.12)	−0.39 (0.30)	−0.36 (0.30)	−0.39 (0.25)	0.854
RAR β	30.52 (17.76)	27.80 (25.92)	32.82 (18.66)	52.05 (57.43)	0.174
Acrophase	13.07 (2.73)	13.97 (3.63)	13.71 (2.99)	13.63 (4.42)	0.813
Amplitude	110.57 (106.56)	74.77 (80.28)	68.20 (49.72)	52.72 (52.85)	0.044*^#
Up-Mesor	5.26 (2.47)	6.49 (4.26)	6.32 (3.48)	6.09 (4.84)	0.736
Down-Mesor	20.88 (3.03)	21.44 (2.62)	21.11 (2.99)	21.18 (4.21)	0.910
Pseudo-F	965.78 (712.60)	768.07 (522.47)	767.08 (417.58)	683.86 (485.55)	0.283
Non-parametric model	N = 19	N = 34	N = 33	N = 29
IS	0.32 (0.20)	0.34 (0.16)	0.36 (0.16)	0.35 (0.14)	0.931
IV	0.77 (0.37)	0.97 (0.33)	1.05 (0.32)	1.11 (0.33)	0.041*^#&
M10	90,642 (89,525)	204,397 (541,534)	54,816 (19,840)	60,251 (54,677)	0.192
L5	7039 (2933)	11,749 (17,477)	9248 (3559)	10,218 (7444)	0.517
RA	0.78 (0.13)	0.73 (0.15)	0.70 (0.12)	0.67 (0.13)	0.098
Sleep quality	N = 19	N = 34	N = 33	N = 29
SL, mean (SD), min	8.63 (13.63)	18.98 (18.44)	18.28 (29.34)	16.21 (16.24)	0.346
SE, mean (SD), %	82.78 (7.33)	83.68 (10.64)	81.15 (12.01)	82.28 (11.80)	0.818
TTB, mean (SD), min	411.68 (27.51)	411.27 (19.63)	417.13 (25.80)	414.99 (22.01)	0.708
TST, mean (SD), min	348.52 (60.71)	347.97 (52.96)	347.84 (47.88)	340.16 (59.92)	0.554
WASO, mean (SD), min	48.64 (25.59)	49.97 (42.24)	62.85 (44.39)	57.91 (55.43)	0.582
TA, mean (SD), times	15.47 (7.01)	17.24 (10.79)	17.55 (7.36)	13.69 (8.17)	0.282
ESS, mean (SD), score	2.40 (3.60)	2.79 (2.92)	3.31 (3.61)	2.91 (3.73)	0.817

RAR: rest-activity rhythm; IS: interdaily stability; IV: intradaily variability; RA: relative amplitude; MRI: magnetic resonance imaging; SL: sleep latency; SD: standard deviation; SE: sleep efficiency; TTB: total time in bed; TST: total sleep time; WASO: wake after sleep onset; TA: times of awakenings; ESS: Epworth sleepiness scale.

*p < 0.05 simple MRI score = 0 versus simple MRI score = 3.

p < 0.05 simple MRI score = 0 versus simple MRI score = 2.

p < 0.05 simple MRI score = 0 versus simple MRI score = 1.

Furthermore, the association between RAR parameters and CSVD was explored using linear regression analysis. As given in Table 3, for the extended cosinor model, WMH was significantly related to higher RAR β (standardized β = 0.387, p = 0.002), and LI was significantly related to higher RAR α (standardized β = 0.297, p = 0.045). The trends were still significant after adjusting for age, gender, education, BMI, and vascular risk factors. As shown in Table 4, for the non-parametric method, elderly persons with CMBs were more likely to have higher fragmentation reflected by higher IV (standardized β = 0.359, p = 0.015), and the statistical association was unchanged after controlling for age, gender, education, BMI, and vascular risk factors.

Table 3.

Association between CSVD and extended cosinor model RAR parameters.

	WMH		LI		Microbleeds
	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2
RAR α
standardized β value	−0.006	0.014	0.297	0.322	−0.246	−0.241
p	0.958	0.910	0.045*	0.032*	0.081	0.112
RAR β
standardized β value	0.387	0.408	−0.087	−0.155	−0.148	−0.186
p	0.002*	0.001*	0.536	0.270	0.272	0.165
Amplitude
standardized β value	−0.216	−0.217	−0.146	−0.170	0.162	0.196
p	0.074	0.085	0.316	0.255	0.240	0.167
Pseudo-F
standardized β value	−0.105	−0.069	−0.117	−0.127	0.093	0.510
p	0.393	0.593	0.430	0.409	0.104	0.478

RAR: rest-activity rhythm; WMH: white matter hyperintensity; LI: lacunar infarct; MRI: magnetic resonance imaging; CSVD: cerebral small vessel disease.

Model 1 represents the association between CSVD and extended cosinor model RAR parameters without adjustment.

Model 2 represents the association between CSVD and extended cosinor model RAR parameters adjusted for age, gender, education, BMI, and vascular risk factors.

*p < 0.05

Table 4.

Association between CSVD and non-parametric RAR parameters.

	WMH		LI		Microbleeds
	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2
IS
standardized β value	−0.201	−0.006	−0.110	−0.187	−0.073	−0.081
p	0.554	0.987	0.461	0.208	0.606	0.567
IV
standardized β value	−0.335	0.286	−0.087	−0.155	0.359	0.320
p	−0.300	0.352	0.536	0.270	0.015*	0.006*
M10
standardized β value	−0.148	−0.127	−0.085	−0.129	0.028	0.003
p	0.235	0.303	0.505	0.323	0.825	0.948
L5
standardized β value	0.017	−0.037	−0.055	−0.127	−0.065	−0.035
p	0.882	0.750	0.638	0.998	0.578	0.780
RA
standardized β value	0.108	0.225	0.015	0.916	0.105	0.077
p	0.741	0.522	−0.006	0.969	0.441	0.558

IS: interdaily stability; IV: intradaily variability; RA: relative amplitude; RAR: rest-activity rhythm; WMH: white matter hyperintensity; LI: lacunar infarct; MRI: magnetic resonance imaging; CSVD: cerebral small vessel disease.

Model 1 represents the association between CSVD and non-parametric RAR parameters without adjustment.

Model 2 represents the association between CSVD and non-parametric RAR parameters adjusted for age, gender, education, BMI, and vascular risk factors.

*p < 0.05

Moreover, we explored the relationship between sleep quality and CSVD. WMH, LI, and CMBs were not related to SE, TTB, WASO, and ESS. Corresponding data are given in Table 5.

Table 5.

Association between CSVD and sleep quality parameters.

	WMH		LI		Microbleeds
	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2
SE
standardized β value	−0.149	−0.116	0.257	0.227	0.119	0.288
p	0.176	0.304	0.052	0.084	0.134	0.251
TTB
standardized β value	−0.003	−0.041	−0.087	−0.155	−0.089	−0.105
p	0.745	0.727	0.536	0.270	0.471	0.413
WASO
standardized β value	0.162	0.079	−0.174	−0.141	0.051	0.053
p	0.144	0.479	0.187	0.272	0.677	0.660
ESS
standardized β value	0.119	0.134	−0.055	−0.127	−0.022	−0.016
p	0.288	0.251	0.638	0.998	0.861	0.899

SE: sleep efficiency; TTB: total time in bed; WASO: wake after sleep onset; ESS: Epworth sleepiness scale; CSVD: cerebral small vessel disease; WMH: white matter hyperintensity; LI: lacunar infarct.

Model 1 represents the association between CSVD and sleep quality parameters without adjustment.

Model 2 represents the association between CSVD and sleep quality parameters adjusted for age, gender, education, BMI, and vascular risk factors.

Discussion

We found that elderly adults with CSVD displayed a disturbed RAR pattern. Using the extended cosinor model, the amplitude was the most obvious parameter disturbed in CSVD patients. Additionally, using a non-parametric method, RA, which is similar to amplitude derived from the former model, was the most disturbed parameter captured. All these findings might be caused by the diurnal low physical activity in CSVD patients, which was reported in our previous research.²⁵

Interestingly, we did not find any markedly affected parameter related to RAR shape and time in elderly patients with CSVD. However, IV, a non-parametric parameter, was marginally disrupted in our study, which was according to the results of Zuurbier et al.¹³ Being different from the extended cosinor model, the non-parametric approach is characterized as “without making any assumptions about the shape of RAR and time schedule”.²⁶ Thus, we proposed that CSVD patients showed a fragmented RAR feature; however, this kind of RAR disturbance found in CSVD patients was not in a characteristic shape or exact time schedule.

Furthermore, different CSVD MRI markers were associated with circadian RAR. WMH was negatively associated with RAR β, which was beyond our proposal. Most previous studies reported that lower RAR β (decreased steepness of the RAR curve) was found in patients with higher depressive symptom severity27 or higher risk of stroke.²⁸ However, there is opposite evidence. For example, higher RAR β was found in elderly people with more severe perceived physical fatigability compared to healthy controls.²⁹ This could be explained by the fact that “the demands of caregiving may have been the predominant driver determining the beta of the RAR”.²⁷ Subjects in our study were collected from a ward of neurology, where most severe CSVD patients received high-level caregiving, which might be a possible reason. Research with a larger sample is still needed in the future to clarify the relationship between WMH and RAR patterns.

By assessing the autopsy of CSVD, Sommers et al.³⁰ failed to find the association between microinfarct and non-parametric RAR parameters. In another study on MRI markers of CSVD, Zuurbier and colleagues did not report the association between LI and non-parametric RAR parameters either.¹³ In the current study, LI was positively and independently associated with RAR α. Thus, elderly people with more LI showed narrower active periods of RAR. According to the results of Zhao et al.,³¹ LI was associated with excessive daytime sleepiness (reflected by ESS score). Although their study did not include the sleep quality of participants with CSVD, the authors assumed that LI might impair the sleep-wake cycle. However, our results did not show an association between LI and objective sleep quality or ESS score. We assumed that other factors, such as apathy and depression, might mediate higher RAR α found with more LI, because both apathy and depression were commonly encountered symptoms in LI.^25,32

In line with previous findings,¹³ CMBs were positively associated with IV. Besides, there was a lack of an association between objective sleep quality and CSVD. CMBs, especially those located in the cortex, are related to amyloid angiopathy.³³ Our results could suggest that amyloid metabolism in the brain was a potential cause of fragmentation in CSVD patients. A similar conclusion was drawn in a study of Alzheimer's disease.³ Considering that CSVD can interrupt frontal-subcortical connections, we agreed with the hypothesis that subclinical brain damage might affect circadian rhythms.

As a cost-effective, objective, and comfortable device, wearable actimetry is popularly used in clinical research and large-scale cohorts such as UK Biobank.^34,35 Circadian RAR measurement based on actigraphyic device is thus becoming more and more widely recognized. In recent years, abnormal circadian RAR has been evidenced as a new emerging predictor in aged adult, for prodromal Alzheimer's disease, Parkinson disease, even long-term all-cause mortality.^3,34,36 More in detail, low amplitude (derived from extended cosinor model) or relative amplitude (derived from non-parametric model) was reported to be a risk factor for adverse health outcomes. Other parameters, reflecting high fragmentation (IV derived from non-parametric model), delayed acrophase (derived from non-parametric model), as well as low robustness (pseudo-F derived from extended cosinor model), were also found to be contributors for delirium and dementia in elderly.³⁵ Based on the actigraphyic data, interventions should be done by researchers of chronobiology, chrono-exercise, and chrono-nutrition, to clarify the effect of improving RAR on health and longevity in future.

There are two major strengths in the present work. First, the two most commonly used RAR analyzing methods, the extended cosinor model and the non-parametric method, were used to describe RAR features. Second, a simple CSVD score was selected to discover whether RAR was disrupted more severely in elderly adults with a higher CSVD burden. This work also had some limitations. First, the sample size in this study was small. Therefore, a larger sample size is required. Second, we cannot elucidate the temporal relationship between RAR disruption and CSVD since this was a cross-sectional study. Third, enlarged perivascular space, which is correlated with sleep disorder and was not evaluated in this study, should be added in future research.

In summary, the RAR pattern was changed in elderly persons with CSVD. WMH, LI, and CMBs were independently associated with RAR β, RAR α, and IV, respectively.

Footnotes

Acknowledgments

We thank Prof Stephen Smagula for technological support.

ORCID iD

Yonghua Huang

Author contributions

Hongyi Zhao (Formal analysis; Writing—original draft); Haiyang Zhang (Formal analysis; Software); Yu Ding (Data curation); Hong Li (Writing—review & editing); Yonghua Huang (Conceptualization; Funding acquisition; Project administration)

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The present work was funded by Wu Jieping Foundation (grant no.: 320.6750.18456).

Declaration of conflicting interests

Yonghua Huang is an Editorial Board Member of this journal but was not involved in the peer-review process of this article nor had access to any information regarding its peer-review.

The remaining authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

The data that support the findings of this study are available on request from the corresponding author Yonghua Huang.

References

Somers

Lopez-Jimenez

, et al. Demographic characteristics associated with circadian rest-activity rhythm patterns: a cross-sectional study. Int J Behav Nutr Phys Act 2021; 18: 107.

Zhang

Manza

Tomasi

, et al. Dopamine D1 and D2 receptors are distinctly associated with rest-activity rhythms and drug reward. J Clin Invest 2021; 131: e149722.

Musiek

Bhimasani

Zangrilli

, et al. Circadian rest-activity pattern changes in aging and preclinical Alzheimer disease. JAMA Neurol 2018; 75: 582–590.

Smagula

Chahine

Metti

, et al. Regional gray matter volume links rest-activity rhythm fragmentation with past cognitive decline. Am J Geriatr Psychiatry 2020; 28: 248–251.

Kang

Woo

Kim

, et al. Independent effects of amyloid and vascular markers on long-term functional outcomes: an 8-year longitudinal study of subcortical vascular cognitive impairment. Eur J Neurol 2022; 29: 413–421.

Regenhardt

Das

, et al. Advances in understanding the pathophysiology of lacunar stroke: a review. JAMA Neurol 2018; 75: 1273–1281.

Zhào

Wei

Liu

, et al. Cognitive frailty among elderly Chinese patients with cerebral small vessel disease: a structural MRI study. Front Med (Lausanne) 2020; 7: 397.

Wardlaw

Smith

Dichgans

. Mechanisms of sporadic cerebral small vessel disease: insights from neuroimaging. Lancet Neurol 2013; 12: 532.

Staals

Booth

Morris

, et al. Total MRI load of cerebral small vessel disease and cognitive ability in older people. Neurobiol Aging 2015; 36: 2806–2811.

10.

Fan

Shen

, et al. Total cerebral small vessel disease burden on MRI correlates with cognitive impairment in outpatients with amnestic disorders. Front Neuro 2021; 12: 747115.

11.

Staals

Makin

Doubal

, et al. Stroke subtype, vascular risk factors, and total MRI brain small-vessel disease burden. Neurology 2014; 83: 1228–1234.

12.

Wang

Chen

Liao

, et al. Non breathing-related sleep fragmentation and imaging markers in patients with atherosclerotic cerebral small vessel disease (CSVD): a cross-sectional case-control study. BMC Neurol 2020; 20: 98.

13.

Zuurbier

Ikram

Luik

, et al. Cerebral small vessel disease is related to disturbed 24-h activity rhythms: a population-based study. Eur J Neurol 2015; 22: 1482–1487.

14.

Oosterman

Van Harten

. Vogels R, et al. Distortions in rest-activity rhythm in aging relate to white matter hyperintensities. Neurobiol Aging 2008; 29: 1265–1271.

15.

Petrea

O'Donnell

Beiser

, et al. Mid to late life hypertension trends and cerebral small vessel disease in the framingham heart study. Hypertension 2020; 76: 707–714.

16.

Nash

Fandler-Höfler

Ambler

, et al. Associations of cerebral small vessel disease and chronic kidney disease in patients with acute intracerebral hemorrhage: a cross-sectional study. Neurology 2024; 103: e209540.

17.

Joint Committee on the Chinese Guidelines for Lipid Management. [Chinese guidelines for lipid management (2023)]. Zhonghua Xin Xue Guan Bing Za Zhi 2023; 51: 221–255.

18.

Amin Al Olama

Wason

JMS

Tuladhar

, et al. Simple MRI score aids prediction of dementia in cerebral small vessel disease. Neurology 2020; 94: e1294–e1302.

19.

Feng

Wen

Xin

, et al. Altered spontaneous brain activity related to neurologic dysfunction in patients with cerebral small vessel disease. Front Aging Neurosci 2021; 13: 731585.

20.

Lee

Chou

Lee

, et al. Cerebral small vessel disease phenotype and 5-year mortality in asymptomatic middle-to-old aged individuals. Sci Rep 2021; 11: 23149.

21.

Zhang

, et al. Increased brain iron detection by voxel-based quantitative susceptibility mapping in type 2 diabetes mellitus patients with an executive function decline. Front Neurosci 2021; 14: 606182.

22.

Van Someren

Swaab

Colenda

, et al. Bright light therapy: improved sensitivity to its effects on rest-activity rhythms in Alzheimer patients by application of nonparametric methods. Chronobiol Int 1999; 16: 505–518.

23.

Witting

Kwa

Eikelenboom

, et al. Alterations in the circadian rest-activity rhythm in aging and Alzheimer’s disease. Biol Psychiatry 1999; 27: 563–572.

24.

Buysse

Hall

Strollo

, et al. Relationships between the Pittsburgh Sleep Quality Index (PSQI), Epworth Sleepiness Scale (ESS), and clinical/polysomnographic measures in a community sample. J Clin Sleep Med 2008; 4: 563–571.

25.

Cai

Zhào

, et al. Detecting apathy in patients with cerebral small vessel disease. Front Aging Neurosci 2022; 14: 933958.

26.

Smagula

. Opportunities for clinical applications of rest-activity rhythms in detecting and preventing mood disorders. Curr Opin Psychiatry 2016; 29: 389–396.

27.

Smagula

Krafty

Taylor

, et al. Rest-activity rhythm and sleep characteristics associated with depression symptom severity in strained dementia caregivers. J Sleep Res 2017; 26: 718–725.

28.

Paudel

Taylor

Ancoli-Israel

, et al. Rest/activity rhythms and cardiovascular disease in older men. Chronobiol Int 2011; 28: 258–266.

29.

Graves

Qiao

Moored

, et al. Profiles of accelerometry-derived physical activity are related to perceived physical fatigability in older adults. Sensors (Basel) 2021; 21: 1718.

30.

Sommer

Schneider

, et al. Disrupted rest-activity rhythms and cerebral small vessel disease pathology in older adults. Stroke 2021; 52: 2427–2431.

31.

Zhao

Kong

Wang

, et al. Association of excessive daytime sleepiness with cerebral small vessel disease in community-dwelling older adults. Nat Sci Sleep 2022; 14: 765–773.

32.

Tully

Alpérovitch

Soumaré

, et al. Association between cerebral small vessel disease with antidepressant use and depression: 3C Dijon Magnetic Resonance Imaging Study. Stroke 2020; 51: 402–408.

33.

Kövari

Herrmann

, et al. The relationship between cerebral amyloid angiopathy and cortical microinfarcts in brain ageing and Alzheimer’s disease. Neuropathol Appl Neurobiol 2013; 39: 498–509.

34.

Feng

Yang

, et al. Association between accelerometer-measured amplitude of rest-activity rhythm and future health risk: a prospective cohort study of the UK Biobank. Lancet Healthy Longev 2023; 4: e200–e210.

35.

Gao

Gaykova

, et al. Circadian rest-activity rhythms, delirium risk, and progression to dementia. Ann Neurol 2023; 93: 1145–1157.

36.

Leng

Blackwell

Cawthon

, et al. Association of circadian abnormalities in older adults with an increased risk of developing Parkinson disease. JAMA Neurol 2020; 77: 1270–1278.