Skip to main content
NCBI home page
Neurobiology of Sleep and Circadian Rhythms logoLink to Neurobiology of Sleep and Circadian Rhythms
. 2024 Dec 25;18:100110. doi: 10.1016/j.nbscr.2024.100110

Beyond sleep: Rest and activity rhythm as a marker of preclinical and mild dementia in older adults with less education

Erika Satomi a,b, Daniel Apolinário a,c, Regina Miksian Magaldi a, Alexandre Leopold Busse a, Gisele Cristina Vieira Gomes d, Elyse Ribeiro e, Pedro Rodrigues Genta f, Ronaldo Delmonte Piovezan g,h, Dalva Poyares g, Wilson Jacob-Filho a, Claudia Kimie Suemoto a,⁎
PMCID: PMC11745811  PMID: 39834590

Abstract

Background

Although sleep duration and sleep-related breathing disorders were associated with dementia previously, few studies examined the association between circadian rhythm association and cognitive status.

Objective

We aimed to investigate the association of rest and activity rhythm with cognitive performance in older people with cognitive complaints and less education.

Methods

Activity rhythm was evaluated with wrist actigraphy in 109 community-dwelling older people with cognitive complaints without diagnosed dementia. Each participant completed a neuropsychological battery and was classified as having cognitive impairment (MCI), dementia, or normal cognition. We used adjusted multinomial logistic regression and linear regression models to compare sleep and circadian non-parametric measures with cognitive groups and cognitive z-scores, respectively.

Results

The mean age of the 109 participants was 79.3 ± 6.3 years old, 74% were women, 68% were white, and the mean education was 5.6 ± 5.2 years. Daytime activity intensity was associated with better language (β = 0.178; 95% CI = 0.022, 0.334; p = 0.03) and visuospatial performance (β = 0.158; 95%CI = 0.008, 0.308; p = 0.04). Also, less fragmented rhythm was associated with better visuospatial (β = 0.172; 95%CI = 0.025, 0.320; p = 0.02) and global cognitive scores (β = 0.134; 95%CI = 0.005, 0.263; p = 0.04). More interdaily stability was associated with a lower risk of MCI and dementia (RR = 0.54; 95%CI = 0.29–0.99; p = 0.04, and RR = 0.44; 95%CI = 0.21–0.94; p = 0.03, respectively). Moreover, more daytime activity (RR = 0.40; 95%CI = 0.18–0.89; p = 0.02) and less rhythm fragmentation (RR = 0.31; 95%CI = 0.14–0.73; p = 0.007) were associated with lower risk for dementia.

Conclusion

Daytime activity intensity and fragmented rhythm during the day and night may play an important role as markers for cognitive impairment in less educated populations. Future studies with larger samples should confirm these findings.

Keywords: Alzheimer's disease, Cognition, Dementia, Non-parametric, Rest-activity rhythm

Highlights

  • •

    Higher daytime activity (M10) was linked to better language and visuospatial performance.

  • •

    Less rhythm fragmentation (IV) was associated with better visuospatial and global cognitive scores.

  • •

    Increased daytime activity was linked to lower dementia risk.

  • •

    Less rhythm fragmentation was associated with lower dementia risk.

  • •

    Stronger interdaily stability (synchronization to the light-dark cycle) was linked to a reduced risk of MCI and dementia.

1. Introduction

In 2015, the worldwide prevalence of Alzheimer's disease (AD), the most common dementia etiology, was 46.8 million with numbers projected to nearly double every 20 years. If interventions could delay both disease onset and progression by one year, there would be nearly 9.2 million fewer cases of AD by 2050 (Brookmeyer et al., 2007; Prince et al., 2007). Therefore, efforts have been made to identify the early dementia stages, including both early symptomatic disease (mild cognitive impairment) and preclinical AD (no cognitive symptoms with evidence of neuropathological lesions). The incidence of dementia in low-income countries is a growing concern as these regions face unique challenges. Limited access to healthcare resources, low levels of awareness, and insufficient infrastructure for early diagnosis and treatment contribute to the increasing dementia prevalence. Socioeconomic factors, such as poverty and lower educational attainment, exacerbate the situation. As the global population ages, addressing the disparities in dementia prevention in low-income countries is crucial for improving health outcomes and quality of life.

Circadian dysfunction and other sleep disorders are prominent features of early AD with a high prevalence, ranging from 24 to 43%, and an important cause of caregiver distress and institutionalization (Hope et al., 1998; Moran et al., 2005). Sleep disorders have been associated with cognitive decline and may be a risk factor or biological marker for AD and other neurodegenerative diseases. Individuals with AD showed reductions in slow-wave sleep and REM sleep, increased frequency of nighttime wakefulness, and breakdown of the sleep/wake circadian rhythmicity, resulting in frequent and prolonged sleep periods during the day. Also, most of these changes increase in magnitude with increasing dementia severity (Vitiello and Prinz, 1989). Lower sleep efficiency, longer sleep latency, increase in wake after sleep onset (WASO), and daytime naps longer than 1 h have been associated with cognitive impairment (Blackwell et al., 2006; Sabia et al., 2021). Particularly, long sleep durations and changes in sleep duration over time were associated with poor cognition in older women (Devore et al., 2014). In community-dwelling older adults insomnia negatively impacted cognition only among individuals with a low educational level, suggesting that low cognitive reserve may play a role in the association between sleep and cognitive impairment (Zimmerman et al., 2012).

However, disrupted rest-activity rhythm patterns have been poorly studied as risk factors for cognitive decline. Individuals with dementia have a worse fit to the 24-h circadian model, and nighttime sleep parameters are not sufficient to study the rest patterns in this population (Smagula et al., 2019). In the present study, we examined whether indicators of disruption in the rest-activity rhythm obtained by objective measurements were associated with worse cognitive performance in Brazilian older adults with suspected cognitive impairment.

2. Methods

2.1. Participants

Patients were referred with suspected cognitive impairment from an outpatient geriatric clinic at the Hospital das Clínicas of the University of São Paulo Medical School (HCFMUSP), Brazil. Eligibility criteria were age 60 years or older and cognitive impairment suspicion by patients, family members, or their physicians or abnormal scores on cognitive screening tests. We excluded individuals with previous dementia diagnosis; delirium (ASSOCIATION, 2014); severe visual, auditory, language, or motor impairment; a severe or decompensated clinical condition that precluded completion of the neuropsychological assessment; and non-Portuguese speaking individuals.

Participants provided information on sociodemographic variables (age, sex, race, and education), and they were screened for symptoms of depression with the Patient Health Questionnaire-9 (PHQ-9) (Kroenke et al., 2001). We used information on comorbidities retrieved from the patient charts to calculate the Charlson Age Comorbidity Index and the number of daily medications (Charlson et al., 1987).

Informed consent was obtained before the study participation. The next of kin consented on behalf of the participant and provided information when the interviewer considered that the participant was unable to understand the study procedures. The local research ethics committee approved this study.

2.2. Cognitive assessment

Participants underwent a comprehensive protocol that included a 60-min neuropsychological battery conducted by experienced geriatricians and psychologists. Memory was evaluated using the Hopkins Verbal Learning Test and the Logical Memory subtest of the Wechsler Memory Scale-Revised (Shapiro et al., 1999; Miotto et al., 2012; Woodard and Axelrod, 1987; Bolognani et al., 2015). The forward and backward digit span of the Wechsler Adult Intelligence Scale-III (Wechsler, 1997; Figueiredo and Nascimento, 2007) and the Color Trail 1 (Rabelo et al., 2010; Williams et al., 1995) were used to evaluate attention. Language was assessed using the semantic fluency test (animals) (WG, 1980; Brucki et al., 1997) and the 30-item version of the Boston Naming Test (Fisher et al., 1999; Mansur et al., 2006). Visuospatial functions were evaluated using the Line Orientation of the Repeatable Battery for the Assessment of Neuropsychological Status (Randolph et al., 1998) and the Clock Drawing Test (Royall et al., 1998). Finally, executive functions were assessed using the Color Trails 2 and the matrix reasoning of the Wechsler Abbreviated Scale of Intelligence (Ryan et al., 2005; Yates, 2006). Composite measures of each cognitive domain were created by converting each test into a z-score based on the mean and standard deviation of the sample, averaging these z-scores, and standardizing this mean to create a domain z-score. A composite measure of global cognition was calculated by averaging the z-scores of all cognitive tests and standardizing this mean.

2.3. Functional assessment

The Clinical Dementia Rating (CDR) interview was administered to a knowledgeable informant (Hughes et al., 1982). It was obtained through semi-structured interviews and rated in six domains: memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care. CRD scale showed good global score agreement to the gold standard regardless of education and education (Maia et al., 2006).

Participants were allocated into three groups according to their performance in the neuropsychological battery and functional assessment. Participants with Z-score > −1 in all cognitive domains were classified as normal cognition. Participants with Z-score ≤ −1 in at least one cognitive domain were classified as having mild cognitive impairment when there was no functional impairment, while participants were classified as having dementia when cognitive and functional impairments were present.

2.4. Objective sleep and circadian rhythm measurements

Participants used the ActTrust actimeter (Condor Instruments, São Paulo, Brazil) to measure sleep and circadian parameters. Similar to a wristwatch device monitoring, the actigraphy was used in the non-dominant arm for 7 consecutive days. The ActTrust device has a 3-axis accelerometer, two precision temperature sensors (one for the skin and one for the environment), and a light sensor with RGB spectrum detailing. The device can monitor up to three months of continuous data. The devices were configured to register the activity data and to process it with a Proportional Integral Mode (PIM) algorithm with 60-s epochs. This algorithm filters and integrates the acceleration to obtain a measure of the user's activity. The PIM data was integrated within every hour of the day, generating twenty-four epochs of 3600 s. The median registry of actigraphy data was seven 24-h periods (SD = 0.65). We collected sleep data on 138 participants, of whom 109 had actigraphy data. Twenty-nine participants were unable to follow the instructions to use the actigraphy and were excluded from this study.

The most widely used method to characterize the rest-activity rhythm is the cosinor method, which can be used to obtain each circadian rhythm information, such as acrophase, mesor, period, and amplitude. These measures are called parametric variables. However, not all biological rhythms follow this sinusoid pattern during the 24-h circadian model, and other variables have been proposed to describe better the circadian rhythm. These measurements are referred to as nonparametric functions (Gonçalves et al., 2014). These nonparametric functions included the following variables: intradaily variability (IV), interdaily stability (IS), the least active 5-h period (L5), and the most active 10-h period (M10). The intradaily variability provides information on the rest-activity rhythm fragmentation. Large hour differences, such as daytime sleep or nighttime awakenings, increase the value of the intradaily variability. Interdaily stability (IS) yields information about rest-activity rhythm synchronization with the light-dark cycle. Higher interdaily variability values show better synchronization. Nocturnal activity is measured by the L5 measurement with lower L5 values indicating a more regular rest. Diurnal activity is measured by the M10 measurement, with higher M10 values suggesting a more active lifestyle. The difference between the nighttime and the daytime activity values shows the amplitude of the rest-activity rhythm. The higher the amplitude, the more adequate the circadian rhythm is. The relative rhythm amplitude was calculated as the difference between M10 and L5 divided by the sum of these two variables (M10+L5).

Other parameters such as the WASO (total minutes awake after sleep onset), sleep efficiency (total sleep time x 100/total minutes in bed with lights out), and total sleep time in minutes were also analyzed.

2.5. Statistical analysis

We described the sample using mean and standard deviation for continuous variables, and absolute and relative frequencies for categorical variables. Chi-square and ANOVA or t-tests were used to compare categorical and continuous variables across cognitive groups (normal, mild cognitive impairment, and dementia) and individuals who were included and excluded from this study. To investigate the association between the circadian rhythm variables and the cognitive tests, we fitted linear regression models with each cognitive domain z-scores and the global composite cognitive z-score as the dependent variables. The independent variables were the nonparametric functions and sleep z-scores. Larger values for intradaily stability and M10 were related to better circadian rhythm, while larger values for L5 and intradaily variability were related to poor rhythm. We multiplied the values of L5 and intradaily variability by −1 before calculating the z-scores so the coefficients for all metrics could be interpreted in the same direction. Linear regression assumptions were met and are presented in the Supplementary Material.

First, we adjusted the models for sociodemographic variables (age, sex, and education). Next, we adjusted the models for comorbidities (Charlson Comorbidity Index), number of daily medications, and depression (PHQ-9). Additionally, we applied multinomial logistic regression models to investigate the association of cognitive status with sleep and non-parametric variables and adjusted these models for sociodemographic variables, comorbidities, and depression. The alpha level was initially set at 0.05 in the two-sided tests. However, multiple comparisons were performed, and we also show results corrected by Bonferroni. The analyses were performed using Stata 15 (StataCorp LP, College Station, TX, USA).

3. Results

The mean age of the 109 participants was 79.3 ± 6.3 years old, 74.3% were women, 67.9% were white, and the mean education was 5.6 ± 5.2 years. Mean daily total sleep time was 451 ± 116 min daily, sleep efficiency was 88.1% ± 5.2%, and WASO was 59.5 ± 24.0 min. There were 25 (22.9%) normal cognitive participants, 63 (57.8%) with mild cognitive impairment, and 21 (19.3%) with mild dementia. The cognitive groups were different regarding age, education, comorbidities, diurnal activity (M10), and intradaily fragmentation (IV) (Table 1). Circadian rhythm variables by cognitive groups are presented in Fig. 1. The 29 individuals who did not have actigraphy data and were excluded from this study had similar sociodemographic, clinical, and cognitive profiles to the included participants.

Table 1.

Characteristic of the study sample (n = 109).

Variable Total (n = 109)
 Normal (n = 25)
 MCI (n = 63)
 Dementia (n = 21)
 p
Mean (SD) or %
Age (years) 79.33 (6.33) 76.44 (6.88) 79.88 (5.64) 81.14 (6.81) 0.02
Male 25.69% 32.0% 22.22% 28.57% 0.60
Education (years) 5.57 (5.18) 6.08 (5.33) 6.38 (5.24) 2.57 (3.74) 0.01
White race 67.89% 56.0% 69.84% 76.19% 0.25
CCI 1.21 (1.38) 0.64 (0.95) 1.31 (1.49) 1.57 (1.36) 0.04
PHQ-9 6.91 (6.25) 6.52 (6.23) 7.17 (6.53) 6.61 (5.62) 0.88
Number of medications 7.52 (3.76) 6.40 (3.93) 7.57 (3.49) 8.71 (4.09) 0.11
Most active 10h period (M10) 4266.73 (1588.21) 5070.09 (1753.61) 4194.25 (1349.68) 3527.81 (1699.16) 0.003
Least active 5h period (L5) 74.13 (64.89) 77.99 (64.82) 73.41 (57.52) 71.69 (86.10) 0.94
Intradaily variability (IV) 0.73 (0.16) 0.67 (0.12) 0.73 (0.12) 0.84 (0.26) 0.002
Interdaily stability (IS) 0.45 (0.10) 0.49 (0.09) 0.44 (0.09) 0.43 (0.12) 0.06
Relative amplitude (RA) 0.96 (0.04) 0.96 (0.03) 0.96 (0.03) 0.94 (0.08) 0.20
Total sleeping time (minutes) 451.54 (115.85) 415 (77.90) 451.55 (109.68) 495 (155.72) 0.06
Sleep efficiency (%) 88.07 (5.22) 87.96 (6.47) 88.13 (4.49) 88.06 (5.87) 0.99
WASO (minutes) 59.50 (23.98) 58.3 (27.61) 58.88 (20.32) 62.76 (29.93) 0.78
Open in a new tab

MCI: mild cognitive impairment; CCI: Charlson Comorbidiy Index; PHQ-9 Patient Health Questionanire-9; WASO: wake after sleep onset.

Cognitive groups were compared using chi-square tests for categorical variables and one-way ANOVA for continuous variables.

Fig. 1.

Fig. 1

Open in a new tab

Violin plots showing the z-scores of the distribution of circadian rhythm variables by cognitive status. MCI: mild cognitive impairment; L5: least active 5-h period; M10: most active 10-h period; IS: interdaily stability; IV: intradaily variability; RA: relative amplitude.

We found associations of the M10 measurement with language (β = 0.178; 95%CI = 0.022 to 0.334; p = 0.03), and visuospatial function (β = 0.158; 95%CI = 0.008 to 0.308; p = 0.04) in the adjusted analysis, which indicates that being more active during the day was related to better performance on these cognitive domains. Larger activity in the L5 period showed an association with worse immediate memory performance after adjustment for confounders (β = −0.176; 95%CI = −0.341 to −0.010; p = 0.04) (Table 2). The difference between nighttime and daytime activity (RA) showed no association with domain-specific or global scores. (Supplementary Table 2).

Table 2.

Association of nocturnal activity (L5) and diurnal activity (M10) with cognitive performance (n = 109).

Crude
 p Model 1
 p Model 2
 p
β (IC 95%) β (IC 95%) β (IC 95%)
Immediate memory MEMO memory L5 −0.149 (−0.339; 0.040) 0.12 −0.175 (−0.340; −0.010) 0.04 −0.176 (−0.341; −0.010) 0.04
M10 0.120 (−0.071; 0.310) 0.22 0.075 (−0.084; 0.234) 0.35 0.050 (−0.116; 0.214) 0.55
Late memory L5 −0.123 (−0.313; 0.067) 0.20 −0.127 (−0.296; 0.042) 0.14 −0.126 (−0.294; 0.041) 0.14
M10 0.185 (−0.003; 0.373) 0.05 0.135 (−0.024; 0.295) 0.10 0.098 (−0.067; 0.262) 0.24
Attention L5 −0.050 (−0.242; 0.141) 0.60 0.003(-0.154; 0.160) 0.97 −0.007(-0.167; 0.154) 0.93
M10 0.107 (−0.084; 0.297) 0.27 0.073 (−0.075; 0.222) 0.33 0.078 (−0.078; 0.234) 0.32
Language L5 0.009 (−0.183; 0.200) 0.93 0.056 (−0.104; 0.217) 0.49 0.044 (−0.119; 0.207) 0.60
M10 0.172 (−0.017; 0.361) 0.07 0.147 (−0.003; 0.298) 0.06 0.178 (0.022; 0.334) 0.03
Executive function L5 −0.002 (−0.193; 0.190) 0.99 0.041 (−0.097; 0.180) 0.56 0.024 (−0.114; 0.163) 0.73
M10 0.098 (−0.092; 0.289) 0.31 0.033 (−0.099; 0.164) 0.62 0.055 (−0.080; 0.191) 0.42
Visuospatial L5 −0.048 (−0.239; 0.143) 0.62 −0.015 (−0.171; 0.141) 0.85 −0.020 (−0.177; 0.136) 0.80
M10 0.193 (0.005; 0.381) 0.04 0.155 (0.010; 0.300) 0.04 0.158 (0.008; 0.308) 0.04
Global score L5 −0.070 (−0.262; 0.121) 0.47 −0.042 (−0.177; 0.093) 0.54 −0.050 (−0.187; 0.086) 0.46
M10 0.170 (−0.019; 0.358) 0.08 0.120 (−0.006; 0.246) 0.06 0.120 (−0.012; 0.251) 0.07
Open in a new tab

L5: Least active 5h period (nocturnal activity); M10: Most active 10h period (diurnal activity).

Model 1: Linear regression model adjusted for sociodemographic variables (age, sex, and education).

Model 2: Linear regression model adjusted for sociodemographic variables, comorbidities (Charlson Comorbidity Index), number of daily medications, and depression (PHQ-9).

Less intradaily variability was associated with better visuospatial performance after adjustment for confounders (β = 0.172; 95%CI = 0.025; 0.320, p = 0.02). Less intradaily variability was also related to better global cognitive score in adjusted analysis (β = 0.134; 95%CI = 0.005, 0.263; p = 0.04) (Table 3). These findings indicated less rhythm fragmentation was associated with better global and visuospatial performance. A heatmap describing the associations between circadian rhythm and cognitive z-scores is shown in Fig. 2.

Table 3.

Association of intradaily variability (IV) and interdaily stability (IS) with cognitive performance (n = 109).

Crude
 p Model 1
 p Model 2
 p
β (IC 95%) β (IC 95%) β (IC 95%)
Immediate memory IV 0.132 (−0.058; 0.322) 0.17 0.079 (−0.080; 0.239) 0.32 0.065 (−0.098; 0.228) 0.43
IS −0.080 (−0.271; 0.110) 0.41 −0.014 (−0.174; 0.146) 0.86 −0.033 (−0.201; 0.134) 0.70
Late memory IV 0.167 (−0.022; 0.356) 0.08 0.125 (−0.035; 0.285) 0.13 −0.109 (−0.273; 0.055) 0.19
IS −0.029 (−0.221; 0.162) 0.76 0.043 (−0.119; 0.205) 0.60 0.011 (−0.157; 0.179) 0.90
Attention IV 0.157 (−0.032; 0.346) 0.10 0.138 (−0.009; 0.285) 0.07 0.137 (−0.016; 0.290) 0.08
IS −0.032 (−0.223; 0.160) 0.74 0.082 (−0.067; 0.231) 0.28 0.093 (−0.065; 0.252) 0.25
Language IV 0.156 (−0.033; 0.345) 0.11 0.129 (−0.022; 0.280) 0.09 −0.141 (−0.014; 0.297) 0.08
IS −0.040 (−0.232; 0.151) 0.68 0.062 (−0.091; 0.215) 0.43 0.093 (−0.065; 0.258) 0.24
Executive function IV 0.124 (−0.066; 0.314) 0.20 0.074 (−0.057; 0.205) 0.27 0.078 (−0.055; 0.211) 0.25
IS −0.112 (−0.303; 0.078) 0.24 −0.017 (−0.150; 0.115) 0.79 0.013 (−0.125; 0.151) 0.85
Visuospatial function IV 0.198 (0.010; 0.386) 0.04 0.167 (0.022; 0.312) 0.02 0.172 (0.025; 0.320) 0.02
IS −0.008 (−0.200; 0.183) 0.93 0.093 (−0.055; 0.241) 0.21 0.106 (−0.048; 0.260) 0.18
Global score IV 0.181 (−0.007; 0.370) 0.06 0.138 (0.012; 0.264) 0.03 0.134 (0.005; 0.263) 0.04
IS −0.059 (−0.250; 0.133) 0.54 0.048 (−0.080; 0.177) 0.46 0.056 (−0.079; 0.190) 0.42
Open in a new tab

IV: Intradaily variability (rest-activity rhythm fragmentation); IS: Interdaily stability (Rest-activity synchronization with the light-dark cycle).

Model 1: Linear regression model adjusted for sociodemographic variables (age, sex, and education).

Model 2: Linear regression model adjusted for sociodemographic variables, comorbidities (Charlson Comorbidity Index), number of daily medications, and depression (PHQ-9).

Fig. 2.

Fig. 2

Open in a new tab

Heatmap for the associations of circadian rhythm variables with global and domain-specific cognitive z-scores. Numbers are the linear regression coefficients for the associations between these variables. Significant associations (p < 0.05) were colored in green, borderline associations in yellow (0.05 ≤ p ≤ 0.10), and non-significant associations in red (p > 0.05). Exact p-values are shown in Table 3, Table 4, and Supplementary Table 2.

Sleep time, efficiency, and fragmentation (WASO) were not associated with any cognitive score (Supplementary Tables 3 and 4). Compared to the group without cognitive impairment, higher diurnal activity (M10) was associated with lower risk for dementia (RR = 0.35; 95%CI = 0.17–0.84; p = 0.02). More intense rhythm fragmentation (IV) was associated with higher dementia risk (RR = 3.21; 95% CI = 1.40 to 7.34; p = 0.006), but not with MCI risk (RR = 1.86; 95%CI = 0.92 to 3.75; p = 0.08). Finally, higher circadian stability (IS) was associated with lower risk for MCI (RR = 0.54; 95%CI = 0.30–0.98; p = 0.04) and dementia (RR = 0.44; 95%CI = 0.21–0.93; p = 0.03) (Table 4). Since we performed 40 tests to evaluate different exposures and cognitive outcomes, associations would not be statistically significant since the corrected alpha level would be 0.001, considering the Bonferroni correction.

Table 4.

Association between circadian rhythm variables and cognitive groups (n = 109).

Crude RR (CI 95%) p Model 1 RR (CI 95%) p Model 2 RR (CI 95%) p
M10 MCI 0.58 (0.36; 0.94) 0.03 0.65 (0.39; 1.09) 0.10 0.71 (0.40–1.23) 0.22
Dementia 0.32 (0.16; 0.67) 0.002 0.36 (0.17; 0.79) 0.01 0.40 (0.18–0.89) 0.02
L5 MCI 1.07 (0.68; 1.67) 0.77 0.93 (053; 1.63) 0.81 0.93 (0.52; 1.64) 0.80
Dementia 1.10 (0.61; 1.97) 0.75 1.11 (0.58; 2.13) 0.74 1.13 (0.57; 2.22) 0.73
IS MCI 0.56 (0.33; 0.94) 0.03 0.49 (0.28–0.87) 0.01 0.54 (0.29; 0.99) 0.04
Dementia 0.52 (0.28; 0.98) 0.04 0.41 (0.20; 0.85) 0.02 0.44 (0.21–0.94) 0.03
IV MCI 0.59 (0.31; 1.10) 0.10 0.50 (0.26; 0.99) 0.05 0.53 (0.26; 1.09) 0.08
Dementia 0.33 (0.16; 0.68) 0.003 0.29 (0.13; 0.65) 0.003 0.31 (0.14; 0.73) 0.007
RA MCI 0.88 (0.47; 1.63) 0.68 0.69 (0.30; 1.59) 0.38 0.77 (0.33; 1.79) 0.51
Dementia 0.64 (0.33; 1.23) 0.19 0.57 (0.23–1.38) 0.21 0.65 (0.27; 1.59) 0.35
Open in a new tab

Reference: normal cognition (n = 23); MCI: mild cognitive impairment (n = 63); dementia (n = 21).

L5: Least active 5h period (nocturnal activity); M10: Most active 10h period (diurnal activity); IV: Intradaily variability (rest-activity rhythm fragmentation); IS: Interdaily stability (Rest-activity synchronization with the light-dark cycle); RA: relative rhythm amplitude (difference between the nighttime and the daytime activity).

Reference group: individuals with normal cognitive function.

Model 1: Multinomial logistic regression model adjusted for sociodemographic variables (age, sex, and education).

Model 2: Multinomial logistic regression model adjusted for sociodemographic variables, comorbidities (Charlson Comorbidity Index), number of daily medications, and depression (PHQ-9).

4. Discussion

In this study of older people with suspected cognitive impairment and less education, not only sleep duration but also rest-activity rhythm patterns were associated with cognitive performance. A more intense diurnal activity pattern (M10) was associated with better language and visuospatial performances. Besides, greater sleep-wake rhythm (IV) fragmentation was related to poor visuospatial function, global cognitive scores, and higher dementia risk. Finally, a better synchronization to the light-dark cycle was associated with lower risk for both MCI and dementia, suggesting IS to be a potential circadian early marker of neurodegenerative diseases.

Sleep characteristics such as total sleep time, daytime sleepiness, and insomnia have been associated with cognitive decline (Devore et al., 2014; Faubel et al., 2009; Keage et al., 2012; Cricco et al., 2001). Most previous results relied only on screening cognitive tests (e.g. the mini-mental state exam) and subjective sleep information, with heterogeneous findings (Devore et al., 2014). Recently, actigraphy has been used to objectively measure sleep and circadian rhythm parameters. According to these measurements, low sleep efficiency, sleep fragmentation, and diurnal napping, but not sleep duration were associated with worse cognitive outcomes (Blackwell et al., 2006; Lim et al., 2012). Moreover, disruptions in the circadian rhythm (day/wake and night/sleep) may represent new biological markers of cognitive decline. Previous cohort studies showed that lower amplitude (strength of activity rhythm), lower mesors (mean level of activity), and lower robustness of circadian activity rhythm were associated with worse cognitive performance and greater odds of developing dementia or mild cognitive impairment after five years in community-dwelling older adults. (Rogers-Soeder et al., 2018)., (Tranah et al., 2011)

However, as not all biological rhythms follow this sinusoid pattern to fit the 24-h circadian model and nonparametric variables may be more suitable measurements for individuals at risk for dementia due to the absence of a clear 24-h sleep-wake pattern, usually with long periods of wakefulness or activity during the night and irregular bouts of sleep or inactivity throughout the day as the neurodegenerative disease progresses (Vitiello and Prinz, 1989). Our findings on the association between a more active pattern during the day and a better cognitive performance are consistent with prior systematic reviews and meta-analyses that have shown physical activity, more specifically leisure-time activity, was associated with a decreased risk of cognitive impairment (Stephen et al., 2017; Blondell et al., 2014; Lee et al., 2010). In preclinical AD, amyloid pathology was associated with intradaily variability. Our findings on the potential capacity of diurnal activity (M10) and rhythm fragmentation to discriminate early stages of dementia, but not MCI, are similar to previous findings suggesting WASO and slow wave sleep could discriminate mild AD but not MCI patients from controls with 90% accuracy. Interestingly, our results suggest that better interdaily stability measured during a period greater than 24 h was associated with a lower MCI risk.

Brainstem regions, including the reticular formation of the pons and medulla, seem to be affected at the early stages of AD. These same regions have an important role in regulating the sleep/wake cycle. Therefore, sleep/wake variables may serve as biological markers in patients at risk for dementia (Vitiello and Prinz, 1989). Particularly, fragmentation of nighttime rest might portend cerebral amyloid-beta deposition. Recent studies found that cognitively normal adults with cerebral amyloid-beta deposition were more likely to nap frequently, had lower sleep efficiency, and poorer sleep quality (Spira et al., 2017; Musiek et al., 2018) In fact, animal studies revealed that there is a diurnal rhythm of amyloid-beta in the brain interstitial fluid, with higher levels when mice are awake and lower levels during sleep (Kang et al., 2009). Human studies showed a similar pattern of amyloid-beta variation in the cerebrospinal fluid. Moreover, amyloid deposition decreased diurnal rhythm oscillations in amyloid-beta, while immunotherapy could restore the circadian amyloid metabolism (Musiek et al., 2015).

Increasing evidence points out that rest-activity changes are present in the early stages of cognitive impairment and seem to be promising early and non-invasive dementia biomarkers. The suprachiasmatic nucleus is the main circadian pacemaker, and its functional deterioration might play a role in the association between AD pathophysiology and the development of sleep/wake abnormalities (Van Erum et al., 2018). The detection of sleep/wake abnormalities could prevent or hasten cognitive decline and might be even more relevant in individuals with less education, who have low cognitive reserve (Zimmerman et al., 2012).

Our study presents some limitations. Although the neuropsychological battery of tests was carefully chosen to evaluate individuals with less education and the neurocognitive battery validation was performed in healthy older adults with heterogeneous education levels, some misclassification might have occurred. Moreover, this was a cross-sectional study and longitudinal analyses should be performed in the future to determine the associations between circadian rhythm and cognitive decline. Although we adjusted the multivariate models for several confounders, residual confounding may be present. For instance, information about physical activity levels is missing. Another important limitation was the possibility of type 1 error since multiple tests were performed. Indeed, no results would be considered statistically significant if the Bonferroni correction was applied. The small sample could also lead to type 2 error, mainly when circadian rhythm variables were related to cognitive groups using multinomial regression models. Considering an alpha level of 5%, the power was only 6% for the association between L5 and cognitive groups, which had an R2 of 0.006. It would be necessary to have a sample of 13,076 participants to see an association between these two variables. For the association between RA and cognitive groups, which had a power of 22%, a sample size of 603 participants would be required. On the other hand, the association between IV and cognitive groups was significant and had an estimated power of 71%. While associations of small effect sizes were not significant, those with larger sizes were, suggesting potential markers of circadian rhythm disruption that would have more clinical importance. Therefore, our findings were novel and should be confirmed in larger samples. Finally, actigraphy accuracy can be affected by chronic conditions and low activity levels due to other reasons (Conley et al., 2019).

On the other hand, among the strengths of our study, different from single-night polysomnography, actigraphy measurements during multiple consecutive days performed in the home environment respecting habitual bedtimes and other nighttime lifestyle-related factors are more likely to reflect the sleep pattern of the participants (Martin and Hakim, 2011). Due to the scarcity of human resources and diagnostic tools in developing countries, the use of devices such as actigraphs could aid in cognitive screening and early diagnosis.

In conclusion, robust diurnal activity patterns and lower fragmentation of rest-activity patterns were associated with worse cognitive performance. Longitudinal studies are important to determine if changes in rest-activity patterns could be biological markers of cognitive decline or even modifiable risk factors.

CRediT authorship contribution statement

Erika Satomi: Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Daniel Apolinário: Writing – review & editing, Project administration, Methodology. Regina Miksian Magaldi: Writing – review & editing, Project administration, Methodology. Alexandre Leopold Busse: Writing – review & editing, Project administration, Methodology. Gisele Cristina Vieira Gomes: Writing – review & editing, Methodology. Elyse Ribeiro: Writing – review & editing, Methodology. Pedro Rodrigues Genta: Writing – review & editing, Methodology. Ronaldo Delmonte Piovezan: Writing – review & editing. Dalva Poyares: Writing – review & editing, Writing – review & editing. Wilson Jacob-Filho: Writing – review & editing, Project administration. Claudia Kimie Suemoto: Writing – original draft, Supervision, Resources, Project administration, Methodology, Funding acquisition, Formal analysis, Conceptualization.

Funding statement

The authors have no funding to report.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors have no conflict of interest related to this study.

Handling Editor: Mark R. Opp

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.nbscr.2024.100110.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Multimedia component 1
mmc1.docx (90.6KB, docx)

Data availability

Data is available upon request to the corresponding author (CKS).

References

  1. ASSOCIATION. AP . 5ed ed. Porto Alegre: Artmed; 2014. DSM-5: manual diagnóstico e estatístico de transtornos mentais; p. 992. [Google Scholar]
  2. Blackwell T., Yaffe K., Ancoli-Israel S., Schneider J.L., Cauley J.A., Hillier T.A., et al. Poor sleep is associated with impaired cognitive function in older women: the study of osteoporotic fractures. J. Gerontol. A Biol. Sci. Med. Sci. 2006;61(4):405–410. doi: 10.1093/gerona/61.4.405. [DOI] [PubMed] [Google Scholar]
  3. Blondell S.J., Hammersley-Mather R., Veerman J.L. Does physical activity prevent cognitive decline and dementia?: a systematic review and meta-analysis of longitudinal studies. BMC Publ. Health. 2014;14:510. doi: 10.1186/1471-2458-14-510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bolognani S.A.P., Miranda M.C., Martins M., Rzezak P., Bueno O.F.A., de Camargo C.H.P., et al. Development of alternative versions of the Logical Memory subtest of the WMS-R for use in Brazil. Dement. Neuropsychol. 2015;9(2):136–148. doi: 10.1590/1980-57642015DN92000008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brookmeyer R., Johnson E., Ziegler-Graham K., Arrighi H. 2007. Forecasting the Global Burden of Alzheimer's Disease. [DOI] [PubMed] [Google Scholar]
  6. Brucki S.M., Malheiros S.M., Okamoto I.H., Bertolucci P.H. [Normative data on the verbal fluency test in the animal category in our milieu] Arq. Neuropsiquiatr. 1997;55(1):56–61. doi: 10.1590/s0004-282x1997000100009. [DOI] [PubMed] [Google Scholar]
  7. Charlson M.E., Pompei P., Ales K.L., MacKenzie C.R. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J. Chron. Dis. 1987;40(5):373–383. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
  8. Conley S., Knies A., Batten J., Ash G., Miner B., Hwang Y., et al. Agreement between actigraphic and polysomnographic measures of sleep in adults with and without chronic conditions: a systematic review and meta-analysis. Sleep Med. Rev. 2019;46:151–160. doi: 10.1016/j.smrv.2019.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cricco M., Simonsick E.M., Foley D.J. The impact of insomnia on cognitive functioning in older adults. J. Am. Geriatr. Soc. 2001;49(9):1185–1189. doi: 10.1046/j.1532-5415.2001.49235.x. [DOI] [PubMed] [Google Scholar]
  10. Devore E.E., Grodstein F., Duffy J.F., Stampfer M.J., Czeisler C.A., Schernhammer E.S. Sleep duration in midlife and later life in relation to cognition. J. Am. Geriatr. Soc. 2014;62(6):1073–1081. doi: 10.1111/jgs.12790. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Faubel R., López-García E., Guallar-Castillón P., Graciani A., Banegas J.R., Rodríguez-Artalejo F. Usual sleep duration and cognitive function in older adults in Spain. J. Sleep Res. 2009;18(4):427–435. doi: 10.1111/j.1365-2869.2009.00759.x. [DOI] [PubMed] [Google Scholar]
  12. Figueiredo Vd, Nascimento E. Desempenhos nas duas tarefas do subteste dígitos do WISC-III e do WAIS-III <i style="">. Psicol. Teor. Pesqui. 2007:313–318. [Google Scholar]
  13. Fisher N.J., Tierney M.C., Snow W.G., Szalai J.P. Odd/Even short forms of the Boston Naming Test: preliminary geriatric norms. Clin. Neuropsychol. 1999;13(3):359–364. doi: 10.1076/clin.13.3.359.1742. [DOI] [PubMed] [Google Scholar]
  14. Gonçalves B.S., Cavalcanti P.R., Tavares G.R., Campos T.F., Araujo J.F. Nonparametric methods in actigraphy: an update. Sleep Sci. 2014;7(3):158–164. doi: 10.1016/j.slsci.2014.09.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hope T., Keene J., Gedling K., Fairburn C.G., Jacoby R. Predictors of institutionalization for people with dementia living at home with a carer. Int. J. Geriatr. Psychiatr. 1998;13(10):682–690. doi: 10.1002/(sici)1099-1166(1998100)13:10<682::aid-gps847>3.0.co;2-y. [DOI] [PubMed] [Google Scholar]
  16. Hughes C.P., Berg L., Danziger W.L., Coben L.A., Martin R.L. A new clinical scale for the staging of dementia. Br. J. Psychiatry. 1982;140:566–572. doi: 10.1192/bjp.140.6.566. [DOI] [PubMed] [Google Scholar]
  17. Kang J.E., Lim M.M., Bateman R.J., Lee J.J., Smyth L.P., Cirrito J.R., et al. Amyloid-beta dynamics are regulated by orexin and the sleep-wake cycle. Science. 2009;326(5955):1005–1007. doi: 10.1126/science.1180962. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Keage H.A., Banks S., Yang K.L., Morgan K., Brayne C., Matthews F.E. What sleep characteristics predict cognitive decline in the elderly? Sleep Med. 2012;13(7):886–892. doi: 10.1016/j.sleep.2012.02.003. [DOI] [PubMed] [Google Scholar]
  19. Kroenke K., Spitzer R.L., Williams J.B. The PHQ-9: validity of a brief depression severity measure. J. Gen. Intern. Med. 2001;16(9):606–613. doi: 10.1046/j.1525-1497.2001.016009606.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lee Y., Back J.H., Kim J., Kim S.H., Na D.L., Cheong H.K., et al. Systematic review of health behavioral risks and cognitive health in older adults. Int. Psychogeriatr. 2010;22(2):174–187. doi: 10.1017/S1041610209991189. [DOI] [PubMed] [Google Scholar]
  21. Lim A.S., Yu L., Costa M.D., Leurgans S.E., Buchman A.S., Bennett D.A., et al. Increased fragmentation of rest-activity patterns is associated with a characteristic pattern of cognitive impairment in older individuals. Sleep. 2012;35(5) doi: 10.5665/sleep.1820. 633-40B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Maia A.L., Godinho C., Ferreira E.D., Almeida V., Schuh A., Kaye J., et al. [Application of the Brazilian version of the CDR scale in samples of dementia patients] Arq. Neuropsiquiatr. 2006;64(2B):485–489. doi: 10.1590/s0004-282x2006000300025. [DOI] [PubMed] [Google Scholar]
  23. Mansur L.L., Radanovic M., Araújo GeC., Taquemori L.Y., Greco L.L. [Boston naming test: performance of Brazilian population from São Paulo] Pro Fono. 2006;18(1):13–20. doi: 10.1590/s0104-56872006000100003. [DOI] [PubMed] [Google Scholar]
  24. Martin J.L., Hakim A.D. Wrist actigraphy. Chest. 2011;139(6):1514–1527. doi: 10.1378/chest.10-1872. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Miotto E.C., Campanholo K.R., Rodrigues M.M., Serrao V.T., Lucia M.C., Scaff M. Hopkins verbal learning test-revised and brief visuospatial memory test-revised: preliminary normative data for the Brazilian population. Arq. Neuropsiquiatr. 2012;70(12):962–965. doi: 10.1590/s0004-282x2012001200014. [DOI] [PubMed] [Google Scholar]
  26. Moran M., Lynch C.A., Walsh C., Coen R., Coakley D., Lawlor B.A. Sleep disturbance in mild to moderate Alzheimer's disease. Sleep Med. 2005;6(4):347–352. doi: 10.1016/j.sleep.2004.12.005. [DOI] [PubMed] [Google Scholar]
  27. Musiek E.S., Xiong D.D., Holtzman D.M. Sleep, circadian rhythms, and the pathogenesis of Alzheimer disease. Exp. Mol. Med. 2015;47:e148. doi: 10.1038/emm.2014.121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Musiek E.S., Bhimasani M., Zangrilli M.A., Morris J.C., Holtzman D.M., Ju Y.S. Circadian rest-activity pattern changes in aging and preclinical alzheimer disease. JAMA Neurol. 2018;75(5):582–590. doi: 10.1001/jamaneurol.2017.4719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Prince M., Ferri C.P., Acosta D., Albanese E., Arizaga R., Dewey M., et al. The protocols for the 10/66 dementia research group population-based research programme. BMC Publ. Health. 2007;7:165. doi: 10.1186/1471-2458-7-165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Rabelo I., Pacanaro S., Rossetti M., Leme I., Castro N., Güntert C., et al. Color trails test: a Brazilian normative sample. Psychol. Neurosci. 2010:93–99. [Google Scholar]
  31. Randolph C., Tierney M.C., Mohr E., Chase T.N. The repeatable battery for the assessment of neuropsychological status (RBANS): preliminary clinical validity. J. Clin. Exp. Neuropsychol. 1998;20(3):310–319. doi: 10.1076/jcen.20.3.310.823. [DOI] [PubMed] [Google Scholar]
  32. Rogers-Soeder T.S., Blackwell T., Yaffe K., Ancoli-Israel S., Redline S., Cauley J.A., et al. Rest-activity rhythms and cognitive decline in older men: the osteoporotic fractures in men sleep study. J. Am. Geriatr. Soc. 2018;66(11):2136–2143. doi: 10.1111/jgs.15555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Royall D.R., Cordes J.A., Polk M. CLOX: an executive clock drawing task. J. Neurol. Neurosurg. Psychiatry. 1998;64(5):588–594. doi: 10.1136/jnnp.64.5.588. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Ryan J.J., Carruthers C.A., Miller L.J., Souheaver G.T., Gontkovsky S.T., Zehr M.D. The WASI matrix reasoning subtest: performance in traumatic brain injury, stroke, and dementia. Int. J. Neurosci. 2005;115(1):129–136. doi: 10.1080/00207450490512704. [DOI] [PubMed] [Google Scholar]
  35. Sabia S., Fayosse A., Dumurgier J., van Hees V.T., Paquet C., Sommerlad A., et al. Association of sleep duration in middle and old age with incidence of dementia. Nat. Commun. 2021;12(1):2289. doi: 10.1038/s41467-021-22354-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Shapiro A.M., Benedict R.H., Schretlen D., Brandt J. Construct and concurrent validity of the Hopkins verbal learning test-revised. Clin. Neuropsychol. 1999;13(3):348–358. doi: 10.1076/clin.13.3.348.1749. [DOI] [PubMed] [Google Scholar]
  37. Smagula S.F., Gujral S., Capps C.S., Krafty R.T. A systematic review of evidence for a role of rest-activity rhythms in dementia. Front. Psychiatr. 2019;10:778. doi: 10.3389/fpsyt.2019.00778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Spira A.P., Stone K.L., Redline S., Ensrud K.E., Ancoli-Israel S., Cauley J.A., et al. Actigraphic sleep duration and fragmentation in older women: associations with performance across cognitive domains. Sleep. 2017;40(8) doi: 10.1093/sleep/zsx073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Stephen R., Hongisto K., Solomon A., Lönnroos E. Physical activity and alzheimer's disease: a systematic review. J. Gerontol. A Biol. Sci. Med. Sci. 2017;72(6):733–739. doi: 10.1093/gerona/glw251. [DOI] [PubMed] [Google Scholar]
  40. Tranah G.J., Blackwell T., Stone K.L., Ancoli-Israel S., Paudel M.L., Ensrud K.E., et al. Circadian activity rhythms and risk of incident dementia and mild cognitive impairment in older women. Ann. Neurol. 2011;70(5):722–732. doi: 10.1002/ana.22468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Van Erum J., Van Dam D., De Deyn P.P. Sleep and Alzheimer's disease: a pivotal role for the suprachiasmatic nucleus. Sleep Med. Rev. 2018;40:17–27. doi: 10.1016/j.smrv.2017.07.005. [DOI] [PubMed] [Google Scholar]
  42. Vitiello M.V., Prinz P.N. Alzheimer's disease. Sleep and sleep/wake patterns. Clin. Geriatr. Med. 1989;5(2):289–299. [PubMed] [Google Scholar]
  43. Wechsler D. The Psychological Corporation; San Antonio, TX: 1997. Psychological Assessment. [Google Scholar]
  44. WG R. Verbal fluency in aging and dementia. J. Clin. Neuropsychol. 1980:135–146. [Google Scholar]
  45. Williams J., Rickert V., Hogan J., Zolten A.J., Satz P., D'Elia L.F., et al. Children's color trails. Arch. Clin. Neuropsychol. 1995;10(3):211–223. [PubMed] [Google Scholar]
  46. Woodard J., Axelrod B. Wechsler memory scale - revised. Psychol. Assess. 1987:445–449. [PubMed] [Google Scholar]
  47. Yates D. Apresentação da Escala de Inteligência Wechsler abreviada: (WASI) Avaliação Psicológica. 2006:227–233. [Google Scholar]
  48. Zimmerman M.E., Bigal M.E., Katz M.J., Brickman A.M., Lipton R.B. Sleep onset/maintenance difficulties and cognitive function in nondemented older adults: the role of cognitive reserve. J. Int. Neuropsychol. Soc. 2012;18(3):461–470. doi: 10.1017/S1355617711001901. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1
mmc1.docx (90.6KB, docx)

Data Availability Statement

Data is available upon request to the corresponding author (CKS).

Articles from Neurobiology of Sleep and Circadian Rhythms are provided here courtesy of Elsevier

RESOURCES